United Kingdom House of Lords Decisions

HOUSE OF LORDS

OPINIONS OF THE LORDS OF APPEAL FOR JUDGMENT

IN THE CAUSE

Secretary of State for the Home Department (Appellant) v. JJ and others (FC) (Respondents)

[2007] UKHL 45

LORD BINGHAM OF CORNHILL

My Lords,

1.   In a judgment given on 28 June 2006 Sullivan J held that obligations imposed on the respondents in control orders made by the Secretary of State under the Prevention of Terrorism Act 2005 deprived the respondents of their liberty in breach of article 5 of the European Convention on Human Rights and that the orders should be quashed: [2006] EWHC 1623 (Admin). The Court of Appeal (Lord Phillips of Worth Matravers CJ, Sir Anthony Clarke MR and Sir Igor Judge P) dismissed the Secretary of State's appeal against that decision on 1 August 2006: [2006] EWCA Civ 1141, [2007] QB 446. On this appeal to the House the Secretary of State challenges both limbs of the decisions below, contending that the obligations imposed on the respondents did not deprive them of their liberty and that, if they did, the orders should have been modified and not quashed.
2.   This is one of four appeals heard by the House together. The facts of the four appeals are different. Some issues are common to more than one appeal, and some are not. Separate judgments were given below at first instance and (in three of the appeals) by the Court of Appeal. It is convenient to give separate judgments in this appeal, in the appeal involving E, and in the appeals involving AF and MB, making such cross-reference as is necessary to avoid repetition.
3.   There are six respondents, to whom I shall refer as "the controlled persons" save where it is necessary to distinguish between them. Five of the controlled persons are Iraqi nationals. The sixth (LL, who has absconded and is not represented in this appeal) is either an Iraqi or an Iranian national. Three have leave to remain in this country and three have temporary admission. All are suspected by the Secretary of State to have been involved in terrorism-related activities and are assessed to pose a threat to the public within the United Kingdom or overseas. None has been charged with or prosecuted for any offence related to terrorism.
4.   The Prevention of Terrorism Act 2005 was enacted on 11 March 2005. It repealed Part 4 of the Anti-terrorism, Crime and Security Act 2001, including section 23, which the House had found to be incompatible with articles 5 and 14 of the Convention in A and others v Secretary of State for the Home Department [2004] UKHL 56, [2005] 2 AC 68. The purpose of the 2005 Act, as expressed in the long title, was "to provide for the making against individuals involved in terrorism-related activity of orders imposing obligations on them for purposes connected with preventing or restricting their further involvement in such activity." At the forefront of his argument the Secretary of State stresses the grave threat presented to the public by the criminal activity of terrorists; the imperative duty of democratic governments to do what can lawfully be done to protect the public against that threat; and the balance inherent in the European Convention between the rights of individuals and the rights of the community as a whole. These considerations provide the important backdrop to these appeals, but they need not be elaborated since they are not controversial.
5.   As will be seen in paragraph 7 below, the 2005 Act is drafted with express reference to article 5 of the European Convention. Article 5 provides that "Everyone has the right to liberty and security of person". The article continues: "No one shall be deprived of his liberty save in the following cases and in accordance with a procedure prescribed by law". There follows a list ((a) to (f)) of cases in which a person may be deprived of his liberty in accordance with a procedure prescribed by law. The cases listed are those in which any democratic state is likely to exercise a power to detain: on sentence following conviction, breach of a court order, arrest on suspicion of crime, infectious disease, mental illness, unlawful entry, pending action to deport or extradite, and so on. This list, as the European Court of Human Rights has repeatedly emphasised, is exhaustive and is to be narrowly interpreted (see, for instance, Engel v The Netherlands (No 1) (1976) 1 EHRR 647, para 57; Kurt v Turkey (1998) 27 EHRR 373, para 122; Mancini v Italy (App no 44955/98, 12 December 2001), [2001] ECHR 502, para 23. This reflects the importance attached by the Convention to the right to liberty and security. Thus a person may not be deprived of his liberty unless his case falls within one of the listed classes of case. That proposition, however, is subject to one qualification. By article 15 of the Convention, given domestic effect by sections 14 and 16 of the Human Rights Act 1998, a state party to the Convention may derogate from article 5, subject to certain formalities, "[1] in time of war or other public emergency threatening the life of the nation … to the extent strictly required by the exigencies of the situation, provided that such measures are not inconsistent with its other obligations under international law". It is common ground that none of the cases subject to this appeal fall within any of the categories listed in (a) to (f) of article 5 of the Convention, and the United Kingdom has not derogated from its obligation to comply with that article. It necessarily follows that if, as the controlled persons (with the support of Justice) contend and the Secretary of State strongly denies, the effect of the obligations imposed on the controlled persons under the control orders is to deprive them of their liberty, such orders are inconsistent with article 5 of the Convention.

The 2005 Act

6.   The core of the 2005 Act is found in section 1. Subsection (1) defines a control order as meaning "an order made against an individual that imposes obligations on him for purposes connected with protecting members of the public from a risk of terrorism". Subsection (4) specifies the obligations which a control order "may include, in particular". It is not therefore an exclusive list. But it is a detailed list, containing sixteen potential obligations running from (a) to (p). It is unnecessary to recite the full list. Among the listed obligations are : "(d) a restriction on his association or communications with specified persons or with other persons generally; (e) a restriction in respect of his place of residence or on the persons to whom he gives access to his place of residence; (f) a prohibition on his being at specified places or within a specified area at specified times or on specified days; (g) a prohibition or restriction on his movements to, from or within the United Kingdom, a specified part of the United Kingdom or a specified place or area within the United Kingdom; (j) a requirement on him to give access to specified persons to his place of residence or to other premises to which he has power to grant access; (k) a requirement on him to allow specified persons to search that place or any such premises for the purpose of ascertaining whether obligations imposed by or under the order have been, are being or are about to be contravened; …". A person who, without reasonable excuse, contravenes an obligation imposed on him by a control order is guilty of an offence punishable, on conviction on indictment to imprisonment for a term of up to five years (s.9(1) and (4)(a)).
7.   The Act draws a categorical distinction between what it calls a "derogating control order" and what it calls a "non-derogating control order". The former is defined in section 15(1) to mean "a control order imposing obligations that are or include derogating obligations" and a "derogating obligation" is defined in section 1(10) to mean "an obligation on an individual which - (a) is incompatible with his right to liberty under Article 5 of the Human Rights Convention; but (b) is of a description of obligations which, for the purposes of the designation of a designated derogation, is set out in the designation order." A "non-derogating control order" is defined in section 15(1) to mean "a control order made by the Secretary of State": it is one that does not consist of or include derogating obligations. Thus the premise of the Act is that control orders made under section 1 of the Act and including obligations within the scope of section 1(4) may, or of course may not, be incompatible with the controlled person's right to liberty under article 5 of the Convention.
8.   The power to make a control order against an individual, in the case of an order imposing obligations that are or include derogating obligations, is exercisable by the court on an application by the Secretary of State (s.1(2)(b)); save where the order imposes obligations that are incompatible with the individual's right to liberty under article 5, the power is exercisable by the Secretary of State (s.1(2)(a)), with the permission of the court (s.3(1)(a)) save where the urgency of the case requires an order to be made without permission (s.3(1)(b)). In each case there is a preliminary hearing by the court, but the procedure differs (section 4(1) applies to derogating control orders, section 3(1)(a), (2), (3), (5) and (6) to non-derogating control orders). The threshold conditions for making an order are different. At the preliminary hearing, the court may make a derogating control order against the individual in question under section 4(3) if it appears to the court
"(a)  that there is material which (if not disproved) is capable of being relied on by the court as establishing that the individual is or has been involved in terrorism-related activity;

(b) that there are reasonable grounds for believing that the imposition of obligations on that individual is necessary for purposes connected with protecting members of the public from a risk of terrorism;

(c) that the risk arises out of, or is associated with, a public emergency in respect of which there is a designated derogation from the whole or a part of Article 5 of the Human Rights Convention; and

(d) that the obligations that there are reasonable grounds for believing should be imposed on the individual are or include derogating obligations of a description set out for the purposes of the designated derogation in the designation order."

By contrast, under section 2(1) the Secretary of State may make a non-derogating control order against an individual if he "(a) has reasonable grounds for suspecting that the individual is or has been involved in terrorism-related activity; and (b) considers that it is necessary, for purposes connected with protecting members of the public from a risk of terrorism, to make a control order imposing obligations on that individual." At the preliminary hearing before such an order is made, or immediately after in case of urgency, the court's function is to consider whether the Secretary of State's decision is "obviously flawed" (s.3(2),(3)).

9.   On the full hearing the function of the court is again different. In the case of a derogating control order the test reflects that set out in section 4(3) quoted above: the court may confirm the order, with or without modifications, only if (section 4(7))
"(a)  it is satisfied, on the balance of probabilities, that the controlled person is an individual who is or has been involved in terrorism-related activity;

(b) it considers that the imposition of obligations on the controlled person is necessary for purposes connected with protecting members of the public from a risk of terrorism;

(c) it appears to the court that the risk is one arising out of, or is associated with, a public emergency in respect of which there is a designated derogation from the whole or a part of Article 5 of the Human Rights Convention; and

(d) the obligations to be imposed by the order or (as the case may be) by the order as modified are or include derogating obligations of a description set out for the purposes of the designated derogation in the derogation order."

In the case of a non-derogating order the function of the court is to decide, applying the principles applicable on an application for judicial review, whether any relevant decision of the Secretary of State was "flawed" (s.3(10)(11)).

10.   A derogating control order has effect for six months unless revoked or renewed (s.4(8)), provided the derogation remains in force and the designation order was not made more than twelve months earlier (s.6(1)), and may be revoked or modified by the court (s.7(5)-(7)). A non-derogating control order has effect for a period of twelve months (s.2(4)), renewable indefinitely for twelve months at a time if the Secretary of State considers that the conditions for making it continue to obtain (s.2(6)). It may be revoked or modified by the Secretary of State (s.7(1)-(2), but he may not make any modification which converts a non-derogating control order into a derogating control order (s.7(3)). A power of arrest exists in relation to derogating but not non-derogating control orders (s.5).
11.   In some respects the Act does not distinguish between the two types of order. Thus the duty on the Secretary of State and the chief officer of police in relation to prosecution, considered in more detail in the case of E, is the same in the two cases (s.8), as are the criminal consequences of contravening an obligation (s.9). The procedural provisions laid down in the Schedule to the Act apply to both types of control order proceedings (s.11), although the rules made pursuant to the rule-making power conferred by the Act distinguish between derogating control orders (Part II of Part 76 of the Civil Procedure Rules) and non-derogating control orders (Part III). No appeal lies to the Court of Appeal from any determination of the court in control order proceedings, except on a question of law (s.11(3)).

Deprivation of liberty

12.   In ordinary parlance a person is taken to be deprived of his or her liberty when locked up in a prison cell or its equivalent. This common sense approach is, unsurprisingly, reflected in the Convention jurisprudence. Thus in Engel, above, para.58, the European Court has recognised that "In proclaiming the 'right to liberty', paragraph 1 of Article 5 is contemplating individual liberty in its classic sense, that is to say the physical liberty of the person", a ruling repeated in Guzzardi v ItalyGuzzardi v Italy (1980) 3 EHRR 333, para.92. It has also referred to "classic detention in prison or strict arrest" (Guzzardi, para.95). Further, the court has recognised the distinction between deprivation of liberty and restriction of movement and freedom of a person to choose his residence. The latter are the subject of article 2 of Protocol 4 to the Convention, a provision which the United Kingdom has not ratified but which is accepted as relevant in interpreting the scope of the prohibition in article 5.
13.   It is, however, common ground between the parties that the prohibition in article 5 on depriving a person of his liberty has an autonomous meaning: that is, it has a Council of Europe-wide meaning for purposes of the Convention, whatever it might or might not be thought to mean in any member state. For guidance on the autonomous Convention meaning to be given to the expression, national courts must look to the jurisprudence of the Commission and the European Court in Strasbourg, which United Kingdom courts are required by section 2(1) of the Human Rights Act 1998 to take into account. But that jurisprudence must be used in the same way as other authority is to be used, as laying down principles and not mandating solutions to particular cases. It is, as observed in R(Gillan) v Commissioner of Police of the Metropolis [2006] UKHL 12, [2006] 2 AC 307, para 23, perilous to transpose the outcome of one case to another where the facts are different. The case law shows that the prohibition in article 5 has fallen to be considered in a very wide range of factual situations. It is to the principles laid down by the court in Engel and Guzzardi particularly, reiterated by the court on many occasions (see, for instance, Ashingdane v United Kingdom (1985) 7 EHRR 528, para 41, Amuur v France (1996) 22 EHRR 533, para 42), that national courts must look for guidance.
14.   A series of Strasbourg decisions establishes that 24-hour house arrest has been regarded as tantamount to imprisonment and so as depriving the subject of his or her liberty : see, for example, Mancini v Italy, above, para 17; Vachev v Bulgaria (App no 42987/98, 8 October 2004), [2004] ECHR 325, para 64; NC v Italy (App no 24952/94, 11 January 2001, para 33, [2001] ECHR 12 ; Nikolova v Bulgaria (No 2) (App no 40896/98, 30 December 2004), para 60, [2004] ECHR 462 In Trijonis v Lithuania (App no 2333/02, 17 March 2005), [2005] ECHR 875 the applicant's complaint in relation to a period of 24-hour home arrest was held to be admissible. In Pekov v Bulgaria (App no 50358/99, 30 June 2006), para 73, [2006] ECHR 299, it was argued by the Government that the house arrest of the applicant did not deprive him of his liberty since the monitoring authorities were far away, so that he could leave his house with impunity, but this was not an argument which the court accepted. The decision of the High Court of Justiciary in McDonald v Dickson 2003 SLT 467, para 17, that the appellant had not been deprived of his liberty during six days of 22-hour house arrest because he had not been subject to any physical confinement or restraint, cannot, in my respectful opinion, be reconciled with this authority.
15.   Continuous house arrest may reasonably be regarded as resembling, save as to the place of confinement, conventional modes of imprisonment or detention. But the court has made clear (Guzzardi, para 95) that deprivation of liberty may take numerous forms other than classic detention in prison or strict arrest. The variety of such forms is being increased by developments in legal standards and attitudes, and the Convention must be interpreted in the light of notions prevailing in democratic states (ibid). What has to be considered is the concrete situation of the particular individual (Engel, para 59; Guzzardi, para 92; HL v United Kingdom (2004) 40 EHRR 761, para 89). Thus the task of a court is to assess the impact of the measures in question on a person in the situation of the person subject to them. The Strasbourg court has been true to this guiding principle. Thus in Engel, para 59, the court recognised that "A disciplinary penalty or measure which on analysis would unquestionably be deemed a deprivation of liberty were it to be applied to a civilian may not possess this characteristic when imposed upon a serviceman." In Ashingdane the applicant had been transferred from a high security mental hospital to an ordinary psychiatric hospital but was, it seems, still held to be detained and so deprived of his liberty (albeit legitimately) during the latest phase of his stay in the psychiatric hospital when he was on an open ward, was free to make regular unescorted visits to his family, was going home every weekend from Thursday to Sunday and was free to leave the hospital as he pleased on Monday to Wednesday provided only that he returned to his ward at night (see pp 536, 543-544).
16.   Thus the court has insisted that account should be taken of a whole range of factors such as the nature, duration, effects and manner of execution or implementation of the penalty or measure in question (Engel, para 59; Guzzardi, paras 92, 94). There may be no deprivation of liberty if a single feature of an individual's situation is taken on its own but the combination of measures considered together may have that result (Guzzardi, para 95). Consistently with this approach, account was taken in Guzzardi of a number of aspects of the applicant's stay on the island of Asinara : the locality; the possibilities of movement; his accommodation; the availability of medical attention; the presence of his family; the possibilities of attending worship; the possibilities of obtaining work; the possibilities for cultural and recreational activities; and communications with the outside (pp 342-345). In the result, the court on the facts attached weight (para 95) to the small area of the island open to him, the dilapidated accommodation, the lack of available social intercourse, the strictness of the almost constant supervision, a nine-hour overnight curfew, the obligation on him to report to the authorities twice a day and inform them of any person he wished to telephone, the need for consent to visit Sardinia on the mainland, the liability to punishment by arrest for breach of any obligation and the sixteen month period during which he was subject to these restrictions. Some of these matters plainly fall within the purview of other articles of the Convention. Because account must be taken of an individual's whole situation it seems to me inappropriate to draw a sharp distinction between a period of confinement which will, and one which will not, amount to a deprivation of liberty, important though the period of daily confinement will be in any overall assessment.
17.   The Strasbourg court has realistically recognised that "The difference between deprivation of and restriction upon liberty is nonetheless merely one of degree or intensity, and not one of nature or substance" (Guzzardi, para 93). There is no bright line separating the two. The court acknowledges (ibid) the difficulty attending the process of classification in borderline cases, suggesting that in such cases the decision is one of pure opinion or what may, rather more aptly, be called judgment.
18.   In assessing the impact of the measures in question on a person in the situation of the person subject to them, the court has assessed the effect of the measures on the life the person would have been living otherwise. Thus no deprivation of liberty was held to result from light arrest of serving soldiers (Engel, para 61) since they continued to perform their duties and remained more or less within the ordinary framework of their army life. The decisions of the court on curfews during the night hours is consistent with that approach. The curfew from 9pm to 7am imposed in Raimondo v Italy (1994) 18 EHRR 237 and the obligation imposed on him not to leave home without informing the police did not prevent him living a normal life and did not deprive him of his liberty. In Labita v Italy (App no 26772/95, 6 April 2000), [2000] ECHR 161, the applicant made no complaint of deprivation of liberty during a period when he was subject to a curfew from 8pm to 6am. In Ciancimino v Italy (1991) 70 DR 103 the applicant was obliged to live in a nominated commune which he was not permitted to leave, was obliged to report to the police daily at 11am and was subject to a curfew from 8pm to 7am, but this did not amount to a deprivation of liberty. The same result followed in Trijonis, above, in which from 11 January 2001 until 6 May 2002, the applicant was permitted to be at his work-place during week-days, subject to a curfew at his home from 7pm to 7am on week-days and for the whole day at the weekend. The court pointed out, contrasting the case with Guzzardi, that the applicant was allowed to spend time at work as well as at home during this period.
19.   It is not, I think, suggested that the Strasbourg court has had to rule on any case at all closely comparable with the present. It is inappropriate to seek to align this case with the least dissimilar of the reported cases. The task of the English courts is to seek to give fair effect, on the facts of this case, to the principles which the Strasbourg court has laid down.

The obligations imposed on the controlled persons

20.   The obligations imposed on the controlled persons by the non-derogating control orders made by the Secretary of State in each of their respective cases were in more or less standard form. Lord Carlile of Berriew QC, the independent reviewer appointed under section 14 of the Act, annexed to his First Report a pro forma of the schedule of obligations "imposed on most but not quite all of the controlees so far" (report, para 42), and Sullivan J annexed to his judgment a list, in almost identical terms, of the obligations imposed on the controlled persons in this case. An obligation was imposed under almost all the heads specifically identified in the paragraphs of section 1(4) of the Act, and some under heads not so identified. The general effect of the obligations was helpfully summarised by the Court of Appeal in paragraph 4 of its judgment :
"4.  The obligations imposed by the control orders are set out in annex I to Sullivan J's judgment. They are essentially identical. Each respondent is required to remain within his 'residence' at all times, save for a period of six hours between 10 am and 4 pm. In the case of GG the specified residence is a one-bedroom flat provided by the local authority in which he lived before his detention. In the case of the other five respondents the specified residences are one-bedroom flats provided by the National Asylum Support Service. During the curfew period the respondents are confined in their small flats and are not even allowed into the common parts of the buildings in which these flats are situated. Visitors must be authorised by the Home Office, to which name, address, date of birth and photographic identity must be supplied. The residences are subject to spot searches by the police. During the six hours when they are permitted to leave their residences, the respondents are confined to restricted urban areas, the largest of which is 72 square kilometres. These deliberately do not extend, save in the case of GG, to any area in which they lived before. Each area contains a mosque, a hospital, primary health care facilities, shops and entertainment and sporting facilities. The respondents are prohibited from meeting anyone by pre-arrangement who has not been given the same Home Office clearance as a visitor to the residence."

It may be added that the controlled persons were required to wear an electronic tag and to report to a monitoring company on first leaving their flat after a curfew period and on returning to it before a curfew period. They were forbidden to use or possess any communications equipment of any kind save for one fixed telephone line in their flat maintained by the monitoring company. They could attend a mosque of their choice if it was in their permitted area and approved in advance by the Home Office. Some of the controlled persons are not permitted, because of their immigration status, to work; those who are permitted have not done so in the six hour period between 10am and 4pm. They received benefits of £30-£35 per week, mostly in vouchers, but in JJ's case £57.45. A request by JJ to study English at a college outside his area was refused.

21.   In the course of a careful and detailed judgment Sullivan J reviewed the authorities mentioned above, and other authorities: in particular Secretary of State for the Home Department v Mental Health Review Tribunal (PH) [2002] EWCA Civ 1868 (Court of Appeal, 19 December 2002), where attention is drawn to the significance of the purpose for which restrictions are imposed, distinguishing between those which are for the benefit of the subject and those which are for some other purpose. He regarded the orders made in these cases, although in force for only twelve months at a time, as of indefinite duration (para 48). He confined himself to facts which were agreed or were apparent on the face of the control orders (paras 57-58). He took as his starting point the confinement of the controlled persons for 18 hours each day of the week in a small flat where (save in the case of GG) they had not previously lived in a significantly different location (paras 60-62). He noted that the controlled persons were all single men, and accepted that the requirement to supply the name, address, date of birth and photographic identification to obtain prior Home Office approval of anyone wishing to visit the flat for social purposes during curfew hours deterred all but the most courageous of visitors (para 66). He expressed his conclusion in paragraph 73 of his judgment :
"73.  Drawing these threads together, and bearing in mind the type, duration, effects and manner of implementation of the obligations in these control orders, I am left in no doubt whatsoever that the cumulative effect of the obligations has been to deprive the respondents of their liberty in breach of Article 5 of the Convention. I do not consider that this is a borderline case. The collective impact of the obligations in Annex I could not sensibly be described as a mere restriction upon the respondents' liberty of movement. In terms of the length of the curfew period (18 hours), the extent of the obligations, and their intrusive impact on the respondents' ability to lead anything resembling a normal life, whether inside their residences within the curfew period, or for the 6-hour period outside it, these control orders go far beyond the restrictions in those cases where the European Court of Human Rights has concluded that there has been a restriction upon but not a deprivation of liberty."

He regarded the controlled persons' concrete situation (para 74) as the antithesis of liberty and more akin to detention in an open prison.

22.   At the outset of his judgment (para 3) the judge had noted Lord Carlile's description of the proforma obligations as "On any view … extremely restrictive … They fall not very far short of house arrest, and certainly inhibit normal life considerably." He found reassurance for his conclusion in the observations of the House of Lords and House of Commons Joint Committee on Human Rights in their Twelfth Report of Session 2005-2006 (HL Paper 122, HC 915), para 38, addressing the proforma obligations without reference to any specific case :
"In our view, those obligations are so restrictive of liberty as to amount to a deprivation of liberty for the purposes of Article 5(1) ECHR. It therefore seems to us that the control order legislation itself is such as to make it likely that the power to impose non-derogating control orders will be exercised in a way which is incompatible with Article 5(1) in the absence of a derogation from that Article."

The judge also noted (para 82) the recognition by Mr Alvaro Gil-Robles, the Council of Europe Commissioner for Human Rights, in paragraph 17 his report (8 June 2005) of a visit to the United Kingdom, of the difficulty under the 2005 Act of distinguishing between derogating and non-derogating obligations :

"The Act does not, however, as noted, provide for any clear cut off point. This is understandable as it would be difficult to provide a clear limit, in particular where there might be many combinations of a variety of different restrictions which are imposable. House arrest would, for instance, clearly, fall within the scope of Article 5(1) ECHR. However, there might be, a strict combination of other restrictions on movement, contacts and residence, falling just short of this. The question of whether the restrictions imposed by the non-derogating control order amount to a deprivation of liberty falling within the scope of Article 5(1) [ECHR] must inevitably be determined on a case-by-case basis …"

23.   On his appeal to the Court of Appeal the Secretary of State contended, as he was bound to do, that the judge had erred in law. He identified (para 7 of the Court of Appeal judgment) five errors of principle : that the judge had identified liberty too broadly, as freedom to do as one wishes; that he had wrongly had regard to the extent to which the obligations interfered with "normal life"; that he had wrongly had regard to restrictions on human rights protected by other specific articles of the Convention; that he had extended the meaning of liberty beyond that laid down in Guzzardi; and that he had concentrated excessively on the individual features of the idiosyncratic cases. The Court of Appeal reviewed these criticisms seriatim, but found no merit in any of them. The judge had clearly and correctly taken the confinement of the controlled persons to a small flat for 18 hours a day as his starting point (para 11). He had properly had regard to other features of a régime at the heart of which was physical confinement (para 19). At the end of the day the judge had to make a value judgment as to whether, having regard to the "the type, duration, effects and manner of implementation" of the control orders, they effected a deprivation of liberty (para 22). The judge's appraisal of the likely duration of the orders, although based on a false premise, was realistic (ibid). The Court of Appeal shared the judge's view that the facts of these cases clearly fell on the wrong side of the dividing line and amounted to a deprivation of liberty contrary to article 5 (para 23).
24.   The Secretary of State's argument on appeal, presented with skill and moderation, repeated, no doubt inevitably, the contentions advanced to and rejected by the Court of Appeal. It is unnecessary to rehearse them since they cannot in my opinion survive a careful reading of the judge's judgment and I would reject them for the reasons which the Court of Appeal gave. No legal error in the reasoning of the judge or the Court of Appeal is shown, and it is not for the House to make a value judgment of its own. I would, however, add that on the agreed facts of these individual cases I would have reached the same conclusion. The effect of the 18-hour curfew, coupled with the effective exclusion of social visitors, meant that the controlled persons were in practice in solitary confinement for this lengthy period every day for an indefinite duration, with very little opportunity for contact with the outside world, with means insufficient to permit provision of significant facilities for self-entertainment and with knowledge that their flats were liable to be entered and searched at any time. The area open to them during their six non-curfew hours was unobjectionable in size, much larger than that open to Mr Guzzardi. But they were (save for GG) located in an unfamiliar area where they had no family, friends or contacts, and which was no doubt chosen for that reason. The requirement to obtain prior Home Office clearance of any social meeting outside the flat in practice isolated the controlled persons during the non-curfew hours also. Their lives were wholly regulated by the Home Office, as a prisoner's would be, although breaches were much more severely punishable. The judge's analogy with detention in an open prison was apt, save that the controlled persons did not enjoy the association with others and the access to entertainment facilities which a prisoner in an open prison would expect to enjoy.

Remedy

25.   The Secretary of State submitted that if, contrary to his submission, the effect of these control orders was to deprive the controlled persons of their liberty in breach of article 5, the courts below were wrong to quash the orders. They should instead have quashed one or more obligations imposed by the orders or directed the Secretary of State to modify them. This argument depended on section 3(12) of the Act, which provides that if the court determines, at a hearing pursuant to directions given at a preliminary hearing, that a decision of the Secretary of State was flawed, its only powers are "(a) power to quash the order; (b) power to quash one or more obligations imposed by the order; and (c) power to give directions to the Secretary of State for the revocation of the order or for the modification of the obligations it imposes." Here, it was said, the court should not have quashed the whole orders, which had detrimental practical results (as when LL absconded before a new order could be made).
26.   Sullivan J did not accept this argument, holding (para 92) that since the Secretary of State had no power to make the order, there was nothing to revoke. The Court of Appeal questioned (para 26) whether the Secretary of State's decision was flawed within the meaning of section 3(10) and (12), but found (para 27) the judge's reasons for quashing the orders compelling. This was also the conclusion reached by Beatson J in Secretary of State for the Home Department v E [2007] EWHC 233 (Admin) (16 February 2007), para 310. Ouseley J in Secretary of State for the Home Department v AF [2007] EWHC 651 (Admin) (30 March 2007), para 89, having found the control order under review to deprive the controlled person of his liberty, similarly held the order to be a nullity.
27.   This conclusion is in my opinion irresistible. As recorded in paragraph 8 above, section 1(2) of the Act provides that the court on the application of the Secretary of State has power to make an order imposing obligations that are or include derogating obligations, while the power to make a control order is exercisable by the Secretary of State "except in the case of an order imposing obligations that are incompatible with the individual's right to liberty under article 5" of the Convention. Thus the Secretary of State has no power to make an order that imposes any obligation incompatible with article 5. An administrative order made without power to make it is, on well-known principles, a nullity: see the recent decision of the Privy Council in Dr Astley McLaughlin v Attorney General of the Cayman Islands [2007] UKPC 50. The defects in the orders cannot be cured by amending specific obligations, since what the Secretary of State made was a series of orders, applicable to the individuals named, and these are what he had no power to make. It is true that, because public law remedies are generally discretionary, the court may in special circumstances decline to quash an order, despite finding it to be a nullity: ibid, para 16. But no such circumstances exist here, and it would be contrary to principle to decline to quash an order, made without power to make it, which had unlawfully deprived a person of his liberty.
28.   This conclusion make it unnecessary to decide, in this case, whether control order proceedings involve the determination of a criminal charge within the meaning of article 6(1) of the Convention, a question discussed in paragraphs 13 to 24 of my opinion in MB and AF.
29.   I would dismiss this appeal with costs.

LORD HOFFMANN

My Lords,

30.   The questions in these appeals are whether the terms of certain control orders made by the Secretary of State under the Prevention of Terrorism Act 2005 are compatible with article 5.1 of the European Convention on Human Rights and whether the procedure by which they were made is compatible with article 6.
31.   The long title of the 2005 Act is "an Act to provide for the making against individuals involved in terrorism-related activity of orders imposing obligations on them for purposes connected with preventing or restricting their further involvement in such activity…" Section 1(1) defined a "control order" as "an order against an individual that imposes obligations on him for purposes connected with protecting members of the public from a risk of terrorism." The power to make the orders is contained in section 2(1):
"The Secretary of State may make a control order against an individual if he-

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

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

32.   Section 1(4) sets out, in 16 lettered paragraphs, the various kinds of obligations which a control order may impose. Since there is no dispute that all the obligations in these appeals fell within the statutory powers, it may be more useful, instead of quoting section 1(4), to set out the terms of one of the most restrictive orders in issue, which was made against an individual referred to as LL. He lives in a one-bedroom flat provided by the National Asylum Support Service (NASS) in an inner London borough, having arrived in the United Kingdom as an asylum seeker.
33.   The order provides that he shall:
(1)  at all times wear an electronic monitoring tag;

(2)  remain in his flat at all times except from 10 am until 4 pm;

(3)  report to the monitoring company twice a day;

(4)  allow the police any time to search his flat;

(5)  not receive any private visitors except with the prior consent of the Home Office, supplying them with the visitor's name, address and photograph;

(6)  Not meet anyone by prior arrangement outside his flat except with Home Office consent and not attend any pre-arranged gatherings except to attend at one mosque approved by the Home Office;

(7)  Not associate or communicate with 5 named individuals against whom control orders have also been made;

(8)  Not use a mobile telephone or internet connection;

(9)  Not go outside a designated area substantially the size of an Inner London borough;

(10)  Surrender his passport;

(11)  Not maintain more than one bank account of which details have been notified to the Home Office;

(12)  Not transfer money or goods abroad without the consent of the Home Office.

34.   The purpose of these obligations is to make it easier for the security services to keep a close watch on what LL is doing and inhibit his participation in terrorist conspiracies. They are plainly a substantial interference with his privacy and freedom of movement. They engage article 8 of the Convention ("Everyone has the right to respect for his private and family life, his home and his correspondence") and would engage article 2 of Protocol 4 ("Everyone…shall…have the right to liberty of movement") if the United Kingdom had ratified that Protocol. They may well engage articles 9, 10 and 11 of the Convention (freedom of religion, freedom of expression and freedom of association) as well. In these appeals, however, no complaint is made on any of these grounds. The reason is that all these rights are qualified. They are subject to "such…restrictions…as are prescribed by law and are necessary in a democratic society, in the interests of national security…" And there can be no doubt that the protection of the state and its people against terrorism is necessary in a democratic society. If, therefore, complaint were made under any of these qualified rights, the court would have to consider whether the particular restrictions could be justified as necessary and proportionate for the purpose of protecting the public.
35.   But your Lordships have not been invited to carry out any such exercise. Instead, LL and the others allege that the orders infringe the rights under article 5.1, which says that "No one shall be deprived of his liberty", subject to various exceptions such as imprisonment for a criminal offence, none of which apply here. The point about the right not to be deprived of one's liberty under article 5.1 is that, subject to the exceptions, it is unqualified. Such is the revulsion against detention without charge or trial, such is this country's attachment to habeas corpus, that the right to liberty ordinarily trumps even the interests of national security. Only in time of war or "public emergency threatening the life of the nation" may the government derogate from the Convention, suspend habeas corpus and imprison people without trial.
36.   There has been no derogation and the question is therefore quite simply whether the effect of the obligations imposed under the control order is to deprive LL and the others of their liberty. It is in my opinion clear from the unqualified nature of the right to liberty and its place in the scheme of the other qualified Convention rights that it deals with literal physical restraint. The right is not infringed by restrictions on liberty in a broader sense, such as restrictions on the right to communicate, associate or pray with others, each of which is protected by a separate qualified right, or with restrictions on movement, which (so far as it is protected at all) is dealt with in article 2 of Protocol 4. So much was stated by the European Court of Human Rights in Engel v The Netherlands (No 1) (1976) 1 EHRR 647. The Court said (at p.669) that the article contemplates "individual liberty in the classic sense, that is to say the physical liberty of the person". The paradigm case of deprivation of liberty is being in prison, in the custody of a gaoler.
37.   Why is deprivation of liberty regarded as so quintessential a human right that it trumps even the interests of national security? In my opinion, because it amounts to a complete deprivation of human autonomy and dignity. The prisoner has no freedom of choice about anything. He cannot leave the place to which he has been assigned. He may eat only when and what his gaoler permits. The only human beings whom he may see or speak to are his gaolers and those whom they allow to visit. He is entirely subject to the will of others.
38.   That is the paradigm case. Obviously, however, one may have some degree of deviation from the standard case without it ceasing to be to a deprivation of liberty. The question of what amounts to a deprivation of liberty was discussed by the European Court of Human Rights in Guzzardi v Italy (1980) 3 EHRR 333. Mr Guzzardi, suspected of association with organised crime, was sent for three years to live under "special supervision" on the small island of Asinara, off the coast of Sardinia, which was then mainly used as a high security prison. (It is now a nature reserve). About 2.5 km 2 of the island lay outside the prison and was available for residence by people under special supervision like Mr Guzzardi. Virtually the only people living on that small piece of land were other internal exiles and carabinieri.
39.   The Court decided by a majority of 11 votes to 7 that Mr Guzzardi had been deprived of his liberty. It is clear that both majority and minority regarded the case as very near the borderline. They agreed that the question was one of degree and, as the majority said, that "account must be taken of a whole range of criteria such as the type, duration, effects and manner of implementation of the measure in question". It is of course helpful to know that the question is one of degree and the matters which should be taken into account in answering it. But one also needs to be told what the question is. What is the criterion for deciding whether someone has been deprived of liberty or not? In the majority judgment that is not easy to discover. The nearest one gets is the statement (in paragraph 95) that "in certain respects the treatment complained of resembles detention in an 'open prison' or committal to a disciplinary unit." That suggests that the question was whether Mr Guzzardi's situation approximated sufficiently closely to the paradigm case of imprisonment. In his dissenting judgment, Judge Matscher said at paragraph 3 that deprivation of liberty was:
"a concept of some complexity, having a core which cannot be the subject of argument but which is surrounded by a 'grey zone' where it is extremely difficult to draw the line."

40.   That is the same idea that I have tried to express by saying that imprisonment is the paradigm case but that the concept may include situations which lack certain features of the paradigm case. There is a similar statement of principle in the dissenting judgment of Sir Gerald Fitzmaurice, in paragraph 6, where he says that, taking into account the separate treatment of freedom of movement in article 2 of Protocol 4:
"The resulting picture is that article 5…guaranteed the individual against illegitimate imprisonment, or confinement so close as to amount to the same thing - in sum against deprivation of liberty stricto sensu."

41.   I do not think that the majority would have disagreed with this statement of principle. It was approved by the Court of Appeal in R (Gillan) v Commissioner of Police of the Metropolis [2005] QB 388, 406 and the Court of Appeal's approach was approved by the House of Lords: see [2006] 2 AC 307, 343. The conclusion of the majority in Guzzardi that his situation was comparable with being in an open prison or a disciplinary unit suggests that they would have agreed with Sir Gerald's criterion - "confinement so close as to amount to the same thing" but thought that "on balance" the case fell on the wrong side of that line.
42.   It is therefore clear that the absence of certain features of the standard case of imprisonment - for example, locked doors or institutional surroundings - are not essential to the concept of deprivation of liberty. One may be deprived of liberty by being placed in an open prison where the doors are not locked but one will be punished if one leaves without permission. Or one may be imprisoned under house arrest in one's own home: see Pekov v Bulgaria (30 June 2006), [2006] ECHR 299 But that does not mean that these features are irrelevant in the assessment of whether one has been deprived of liberty. For example, to be placed under actual physical constraint for any length of time is, for that period, a deprivation of liberty. So in Gillan's case [2006] 2 AC 307, 343, where the appellants had been stopped and searched, Lord Bingham of Cornhill said that they had been kept waiting rather than deprived of their liberty and distinguished the case of a person who is "arrested, handcuffed, confined or removed to any different place." These amount to a deprivation of liberty. So, for example, in X v Austria (1979) 18 DR 154 the Commission expressed the view that to detain someone forcibly, even for a short time for the purpose of taking blood for a test, was a deprivation of his liberty.
43.   However, when neither physical restraint nor removal from one's home is present, the Court takes a broader view. It does not confine its attention only to those times at which the person's liberty is most restricted (for example, when he is subject to a curfew) but asks in more general terms, as in Guzzardi's case, whether his situation approximates sufficiently closely to being in prison. Thus in Trijonis v Lithuania (17 March 2005), [2005] ECHR 875 the applicant was placed under "home arrest" which required him to stay at home all week-end and between 7 pm and 7 am on work days. The court said that his movements had been restricted but he had not been deprived of his liberty. It did not say that he was deprived of his liberty at the week-ends, even though he could not then leave his house.
44.   My Lords, these cases seem to me to provide a clear enough statement and illustrations of the principle. In order to preserve the key distinction between the unqualified right to liberty and the qualified rights of freedom of movement, communication, association and so forth, it is essential not to give an over-expansive interpretation to the concept of deprivation of liberty. I remain of the opinion which I expressed in A v Secretary of State for the Home Department [2005] 2 AC 68, 129-132 that the power to derogate in peace time is a narrow one and that politically or religiously motivated violence, even threatening serious loss of life, does not necessarily "threaten the life of the nation" within the meaning of the Convention. The liberty of the subject and the right to habeas corpus are too precious to be sacrificed for any reason other than to safeguard the survival of the state. But one can only maintain this position if one confines the concept of deprivation of liberty to actual imprisonment or something which is for practical purposes little different from imprisonment. Otherwise the law would place too great a restriction on the powers of the state to deal with serious terrorist threats to the lives of its citizens. In the case of anything less than actual deprivation of liberty, the other rights which are undoubtedly engaged are in my opinion adequately protected by the requirement that any interference with them must be necessary and proportionate in the interests of national security.
45.   If one applies these principles to the facts of the present case, the answer seems to me to be clear. I find it impossible to say that a person in the position of LL is for practical purposes in prison. To describe him in such a way would be an extravagant metaphor. A person who lives in his own flat, has a telephone and whatever other conveniences he can afford, buys, prepares and cooks his own food, and is free on any day between 10 am and 4 pm to go at his own choice to walk the streets, visit the shops, places of entertainment, sports facilities and parks of a London borough, use public transport, mingle with the people and attend his place of worship, is not in prison or anything that can be called an approximation to prison. True, his freedom of movement, communication and association is greatly restricted compared with an ordinary person. But that is not the comparison which the law requires to be made. The question is rather whether he can be compared with someone in prison and in my opinion he cannot.
46.   Sullivan J and the Court of Appeal came to a different conclusion. But there is no indication in either judgment that they applied what in my opinion is the correct test. Lord Phillips of Worth Matravers CJ said (at [2007] QB 446, 457) that confining someone to his flat for 18 hours a day "makes most serious inroads on liberty". So it does. He went on to agree with Sullivan J that one had to take into account that even when he was outside his flat, LL's freedom was restricted as to how far he could go, whom he could arrange to meet and so on. That is true. The Lord Chief Justice said (at p 460) that "at the end of the day", Sullivan J had to make "a value judgment as to whether, having regard to 'the type, duration, effects and manner of implementation' of the control orders they effected a deprivation of liberty." But that formulation offers no guidance as to what would count as a deprivation of liberty. It simply says that the judge must take everything into account and decide the question, without saying what the question is. For these reasons I consider that the judge and the Court of Appeal not so much misdirected themselves as gave themselves no directions at all. If they had asked themselves whether the person in question could realistically be regarded as being for practical purposes in prison, I do not see how they could have arrived at the conclusion which they did.
47.   If I had considered that the combined effect of the obligations imposed by the control orders was a deprivation of liberty, I would have had to decide whether the control order should simply be quashed or whether different obligations which did not have the same effect could be substituted.
48.   The procedure for making a control order (except in cases of urgency) is that the Secretary of State must apply to the court for permission: see section 3(1)(a) of the 2005 Act. If the court gives permission it must give directions for a hearing "as soon as reasonably practicable after it is made": section 3(2)(c). In these cases, Sullivan J gave permission and directions for the hearing over which he then presided. Section 3(10) provides that at such a hearing the function of the court is to determine whether any of the following decisions of the Secretary of State was "flawed":
(a)  his decision that the requirements of section 2(1)(a) and (b) were satisfied for the making of the order; and

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

49.   Section 3(11) says that in deciding what constitutes a flawed decision, the court must apply the principles of judicial review. So the question for the court will be whether the decisions of the Secretary of State as to the matters mentioned in section 3(10) were unlawful on one of the normal grounds for judicial review. In the cases before the House, there is no challenge to the Secretary of State's decision that the requirements of section 2(1)(a) and (b), which I have already quoted, were satisfied. The basis of the challenge is that the cumulative effect of the imposition of the obligations infringed the Convention right under article 5.1. If that was the case, the imposition of the obligations would have been unlawful because contrary to the duty of the Secretary of State under section 6(1) of the Human Rights Act 1998.
50.   Section 3(12) then provides that if the court determines that a decision of the Secretary of State was flawed, "its only powers are":
(a)  power to quash the order;

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

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

51.   Section 3(10) makes it clear that the Secretary of State's decision to impose each of the obligations is to be considered as a separate decision, although of course in determining whether it is flawed, the court may have to consider its cumulative effect in conjunction with the decisions to impose the other obligations. By section 3(12), if the court thinks that the decision to impose a particular obligation was unlawful, it may quash that obligation or direct the Secretary of State to modify it. The plain meaning of these provisions seems to be that if, for example (like my noble and learned friend Lord Brown of Eaton-under-Heywood), the court considers, that the obligations infringe article 5.1 because they require someone to remain indoors 18 hours a day but that they would be perfectly lawful if they only required him to remain indoors for 16 hours a day, the court is not obliged to quash the order. It can simply direct the Secretary of State to modify that particular obligation.
52.   Sullivan J decided nevertheless to quash the orders. He said:
"I have no doubt that the proper course is to quash these control orders under paragraph (a) and that it would not be appropriate to direct the Secretary of State to revoke the orders or to modify the obligations imposed by them. A direction to revoke or to modify carries with it the implication that there is in existence an order which was lawfully made by the Secretary of State, but which has been found to be flawed for some reason. The short answer to the Secretary of State's submission that he should be directed to modify these orders is that since he had no power to make them in the first place, there is simply nothing to revoke. The orders were made 'without jurisdiction' in the narrow pre-Anisminic ([1969] 2 AC 147) sense of lack of jurisdiction. Each order would therefore have been described as a 'nullity', when the distinction between jurisdictional and non-jurisdictional error of law was still of consequence."

53.   I am afraid that I must respectfully disagree with this reasoning. If the order of the Secretary of State is found to be flawed on principles of judicial review, that means that it was not lawfully made. I do not understand how some unlawful orders can be more lawful than others or that it makes sense to invoke distinctions which English law abandoned forty years ago in order to create different categories of unlawfulness. The power to direct the Secretary of State to revoke or modify the order does not imply that the order was lawfully made. On the contrary, the power arises only if the order is found to have been flawed, that is to say, not lawfully made. Thus the grounds on which the judge refused to consider the exercise of the powers conferred by section 3(12)(b) and (c) would simply write them out of the statute. But there seems to me no conceptual reason why Parliament should not say that if the exercise of a power is found to have been unlawful, the court shall have power to modify the order or direct the Secretary of State to modify it so as to make it lawful. The judge's failure to accept that he had these powers means that in my opinion he did not properly exercise his discretion.
54.   The Court of Appeal said that the reasons which Sullivan J gave for quashing the orders were "compelling". In addition to the conceptual reason which he gave, Sullivan J also said that quashing the orders would be fair because then the people against whom they were made could not be prosecuted for contravening them. That may in some circumstances be a good reason, but I do not think it will always be so. Ordinarily, people who challenge the validity of orders made against them are not free simply to ignore them. They must obey them until they are set aside. The decision to make the order may have been flawed for some reason which has nothing to do with the obligation which has been contravened. The Court of Appeal added that the Secretary of State was in a better position than the court to devise a "new package of obligations". But that seems to me to carry little weight if the Secretary of State says that he will be content with the modification which the court thinks necessary to make the orders lawful, such as a small reduction in the curfew period.
55.   If, therefore, I had thought that the decision to impose any of the obligations under the orders was flawed, I would have remitted the case to the judge to reconsider the exercise of his powers under section 3(12). But because I think that none of them was flawed, the question does not arise.
56.   As for the question of compliance with article 6, I think that for the reasons I have given in Secretary of State for the Home Department v AF and Secretary of State for the Home Department v MB [2007] UKHL 46 proceedings concerning control orders are not criminal proceedings and that the special advocate procedure complies with the requirements of article 6 for civil proceedings. I would therefore allow the appeals of the Secretary of State.

BARONESS HALE OF RICHMOND

My Lords,

57.   What does it mean to be deprived of one's liberty? Not, we are all agreed, to be deprived of the freedom to live one's life as one pleases. It means to be deprived of one's physical liberty: Engel v The Netherlands (No 1)(1976) 1 EHRR 647, para 58. And what does this mean? It must mean being forced or obliged to be at a particular place where one does not choose to be: eg X v Austria (1979) 18 DR 154. But even that is not always enough, because merely being required to live at a particular address or to keep within a particular geographical area does not, without more, amount to a deprivation of liberty. There must be a greater degree of control over one's physical liberty than that. But how much? As the Judge said, the Strasbourg jurisprudence does not enable us to narrow the gap between "24-hour house arrest seven days per week (equals deprivation of liberty) and a curfew/house arrest of up to 12 hours per day on weekdays and for the whole of the weekend (equals restriction on movement)": [2006] EWHC 1623 (Admin), para 33, referring to the cases cited by my noble and learned friend Lord Bingham of Cornhill, at paras 14 and 18 above.
58.   The Strasbourg jurisprudence does tell us that "deprivation of liberty may . . . take numerous other forms" than "classic detention in prison or strict arrest imposed on a serviceman": Guzzardi v Italy (1980) 3 EHRR 333, para 95. We must look at the "concrete situation" of the individual concerned and take account of "a whole range of criteria such as the type, duration, effects and manner of implementation of the measure in question": ibid, para 92; also Engel, para 59; HL v United Kingdom (2004) 40 EHRR 761, para 89. However, the "difference between deprivation of and restriction upon liberty is nonetheless merely one of degree or intensity, and not one of nature or substance": Guzzardi, para 93. It also appears that restrictions designed, at least in part, for the benefit of the person concerned are less likely to be considered a deprivation of liberty than are restrictions designed for the protection of society: see Secretary of State for the Home Department v Mental Health Review Tribunal and PH [2002] EWCA Civ 1868, paras 16-17, citing Nielsen v Denmark (1988) 11 EHRR 175 and HM v Switzerland (2002) 38 EHRR 314; Davis v Secretary of State for the Home Department [2004] EWHC 3113 (Admin).
59.   The Judge took as his starting point the requirement to remain in the "residence" for 18 hours each day, between 4.00 pm and 10.00 am: see para 60. This is classic detention or confinement. It is secured by electronic tagging, a requirement to clock out on leaving and clock in on returning, and by arrest and imprisonment for disobedience, rather than by lock and key. But that makes no difference: "to determine whether a person is deprived of his or her liberty the Court must look upon the actual circumstances of the regime to which he or she was subject, as a matter of law and in fact"; having the opportunity to breach the requirements of that regime does not take it outside article 5: Pekov v Bulgaria (App no 50358/99, 30 June 2006), para 73, [2006] ECHR 299
60.   Having taken the 18 hour curfew as his starting point, the Judge went on to consider the concrete situation of the controlled persons, at first during the curfew hours and then during their six hours of comparative freedom. If we do the same, we can see the extent to which the regime controlled their lives and cut them off from normal society. With one exception, they were required to move from the places where they had previously lived to prescribed addresses in a different area. These addresses were one bed-roomed flats where they lived alone. They were not allowed into the communal areas during the curfew hours. They were allowed one landline telephone apart from the dedicated line supplied by the monitoring company. They were not allowed access to the internet nor were they supplied with any other means of making their isolation more bearable. No-one was to be allowed in at any time, apart from their own lawyers, the emergency services or healthcare or social work professionals in an emergency, and anyone required to be given access under the terms of the tenancy. They were also required to allow the police to enter and search at any time, to remove or inspect anything, and to install equipment to ensure compliance with the order. Any other visitor required the prior approval of the Home Office, which had to be supplied with the name, address, date of birth and a photograph. Not surprisingly, there had been few requests for approval. The cases against the controlled persons rested largely on their links with one another and with other people with links to known terrorist individuals or organisations. Who - apart from someone with a professional reason to do so or a close family member (and these people have no family here) - would want to be seen to be associating with them?
61.   Undoubtedly, these people were deprived of their liberty during the curfew hours. Did the fact that they were allowed out for up to six hours a day make any difference? The areas to which they were restricted consisted, save in one case, of large parts of some major cities, including parks, recreational facilities, libraries, shops, and healthcare services. They had the freedom to choose what to do and what to buy with the small allowances with which they were provided (mainly in vouchers). But that freedom was also severely curtailed. Without prior Home Office agreement, they were not allowed to meet anyone by prior arrangement, apart from their lawyers or health or welfare workers at an agreed establishment; nor were they allowed to attend any pre-arranged meetings or gatherings, apart from attending group prayers at a mosque. And the areas to which they were confined were deliberately designed to cut them off from their old haunts and acquaintances. Even supposing that the Home Office would have been willing to allow them to register for regular educational classes or group recreational activities, the hours of 10.00 am to 4.00 pm do not fit in with any ordinary pattern of morning, afternoon or evening activity. Nor, in practice, would those whose immigration status allowed them to work be able to seek even part time employment.
62.   It is in this context that the Judge talked of a 'normal' life. He was not starting from a normal life and seeing how far the control order regime differed from this. He was starting from the 18 hour curfew and assessing how far they were nonetheless able to pursue a normal life. The reality is that every aspect of their lives was severely controlled. They were allowed out each day to go for a long and solitary walk, to attend prayers at their nominated mosques, and to buy such limited supplies as they could afford. This would not prevent detention in a psychiatric hospital under the Mental Health Act 1983 from being a deprivation of liberty: see Ashingdane v United Kingdom (1985) 7 EHRR 528. It is not surprising that the Judge concluded that "The respondents' 'concrete situation' is the antithesis of liberty, and is more akin to detention in an open prison, where the prisoner is 'likely to be released from prison regularly in order to work, take town visits and temporary release on resettlement or facility licence': see paragraph 5.37 of Prisoners and the Law, 3rd edn, by Creighton and others." Indeed, in several respects a prisoner might be better off: para 74.
63.   In common with the Court of Appeal and with my noble and learned friends, Lord Bingham of Cornhill and Lord Brown of Eaton-under-Heywood, therefore, I consider that the Judge applied the right test and, for what it is worth, reached a conclusion on the facts with which I would agree. It is necessary to focus on the actual lives these people were required by law to lead, how far they were confined to one place, how much they were cut off from society, how closely their lives were controlled. The Judge was entitled to conclude that the concrete situation in which they found themselves did deprive them of their liberty within the meaning of article 5 of the Convention. As such situations may be many and various I would hesitate to suggest, in the abstract, what length of curfew would fall on the other side of the line.
64.   As to remedy, the 2005 Act draws a clear and principled distinction between control orders which do, and control orders which do not, amount to a deprivation of liberty. It recognises that people should not be deprived of their liberty on the basis of reasonable suspicion alone: involvement in terrorist-related activity must be proved: 2005 Act, s 4(7)(a). It also recognises that only a court may deprive people of their liberty: 2005 Act, s 1(2)(b). The Home Secretary has no power to make such an order. For the reasons given by Lord Bingham, the Judge had no choice but to quash these orders. To his speech, I am merely the chorus. I too would dismiss these appeals.

LORD CARSWELL

My Lords,

65.   The tension between the opposing imperatives of protecting the safety of the public and protecting individual human rights has increased steadily in the past few years, and finding an acceptable resolution has grown progressively more difficult. The Government has taken steps designed to discharge its duty of protecting the public against terrorism and these have been the subject of regular challenges by those adversely affected by them. Parliament is not free to legislate as it chooses in this sphere: its ability to do so is limited by the provisions of the Human Rights Act 1998 and the necessity for legislation to be compatible and steps taken to be compliant with the requirements of the European Convention for the Protection of Human Rights and Fundamental Freedoms ("the Convention"). The duty falls upon the courts of undertaking the difficult task of adjudicating upon that compatibility and compliance.
66.   The dangers to the public posed by terrorist action give rise to very serious concern, shared by all responsible citizens. The subjects of the present appeals are persons about whom the authorities possess information which, if correct, would mean that they pose a very significant potential danger to the safety of the public. In each case, notwithstanding the extent and nature of the information, the evidence capable of being adduced in criminal prosecutions is regarded as being insufficient to obtain convictions of criminal offences. They cannot be deported or extradited, because of the constraints of article 3 of the Convention, as it is claimed that they would face torture or inhuman treatment if returned to their own countries. In order to meet this situation, Parliament, rather than leave them at large, enacted section 23 of the Anti-terrorism, Crime and Security Act 2001, providing for the detention of such persons despite the fact that their removal or departure from the United Kingdom was prevented. For this to be done there had to be a derogation from article 5(1)(f) of the Convention, which was effected by the Human Rights Act 1998 (Designated Derogation) Order 2001 (SI 2001/3644).
67.   That expedient foundered when the House held in A v Secretary of State for the Home Department [2005] 2 AC 68, on a challenge by some of the detained persons, that section 23 of the 2001 Act was discriminatory and disproportionate, with the consequence that it could not be regarded as strictly required for the purposes of article 15 of the Convention, the derogation provision. The House accordingly declared section 23 to be incompatible with the Convention rights under articles 5 and 14 and quashed the 2001 Order.
68.   In consequence of this decision the Government turned to the idea of keeping such terrorist suspects under supervision by means of control orders and Parliament brought them into effect by enacting the Prevention of Terrorism Act 2005 ("the 2005 Act"). Their object, as the long title of the Act states, is to impose obligations upon them "for purposes connected with preventing or restricting their further involvement" in terrorism-related activity. They operate, as my noble and learned friend Lord Hoffmann has stated (para 34) by making it easier for the security services to keep a close watch on what they are doing and inhibiting their participation in terrorist conspiracies. The material provisions of the 2005 Act and the content of the control orders in question have been set out in detail in the opinion of my noble and learned friend Lord Bingham of Cornhill and I gratefully adopt these without repeating them.
69.   In the appeal to which this opinion relates, the issue argued was whether control orders constituted a deprivation of liberty and so a breach of article 5(1) of the Convention. The word "liberty" has a range of meanings. In a narrower sense it may mean physical freedom to move, so that deprivation of liberty would be physical incarceration or restraint. In a wider sense it may mean the freedom to behave as one chooses, for example, liberty of speech. For the reasons which I shall give, and in agreement with those set out by Lord Hoffmann, I am of opinion that in the phrase "deprived of his liberty" in article 5(1) the word should be interpreted in the narrower sense which I have defined.
70.   Lord Hoffmann has pointed out, but I would emphasise it again, that the challenge to the control orders is being made only under article 5 of the Convention and not under article 2 of Protocol No 4, which provides for the right to liberty of movement, since the United Kingdom has not ratified that Protocol. It is of great importance to draw a clear distinction between the two articles. The existence of article 2 of Protocol No 4 shows in my opinion that the framers of this provision were conscious of the limited extent of article 5 of the Convention and saw the need for a separate provision to cover restriction of movement. I think that its existence also supports the view that the ambit of article 5 should be kept clear and distinct from that of article 2 of Protocol No 4, and that there is no need or room for a purposive construction of article 5 which would extend it in the direction of applying to restrictions of movement.
71.   The Court of Appeal [2007] QB 446 expressed the view in paragraphs 12-13 of its judgment in JJ's case that Sullivan J in the Administrative Court correctly interpreted "liberty" in accordance with the direction of the European Court of Human Rights in its judgment in Guzzardi v Italy (1980) 3 EHRR 333 (to which I shall return in more detail later) and rejected Mr Sales' contention on behalf of the Secretary of State that the learned judge had taken too broad a meaning and had considered the extent to which the restrictions contained in the control orders interfered with "normal life". I am unable to agree with this view, and examination of the terms of the judgment of Sullivan J will show that he did just what Mr Sales attributed to him.
72.   Having summarised in paragraph 15 some of the principles propounded in Guzzardi, the judge then said in paragraph 54:
"54 The extent to which the individual is subject to supervision, the extent to which he can make social contacts, the extent to which he has access to public facilities, and whether he is free to make telephone calls or otherwise to communicate with whomsoever he wishes, are all aspects of a broader question: to what extent is the individual subject to the obligations able to lead a life of his choice, which for convenience may be described as a 'normal' life? If one asks the question 'deprived of liberty to do what?', the answer must be: deprived of the freedom to lead one's life as one chooses (within the law). That freedom is the antithesis of a life which is subject to the kinds of control to which a prisoner, whose 'liberty to do anything is governed by the prison regime' is subject: see per Lord Jauncey at page 176H of R v Deputy Governor of Parkhurst Prison, ex parte Hague [1992]1 AC 58."

    He went on to refer in paragraph 63 to the extent to which the restrictions enabled the persons affected to "lead a 'normal' life". Again, he stated at paragraph 77:

"77 In accordance with the principles established in Guzzardi, I have considered the cumulative impact of the obligations and therefore the extent to which they restrict the respondents' liberty in the six hours when they are allowed out of their residences, as well as the effect of the 18-hour curfew and the obligations imposed on the respondents whilst they have to remain within their residences during that period. If I had to assess the impact of the obligations individually, I would consider that house arrest for 18 hours each day, even if it was the only obligation (apart from obligations such as reporting and tagging to ensure that it was strictly observed) would be more realistically described as a deprivation of liberty, and not as a restriction on liberty, if it prevented the individual from pursuing a normal "in at home/out at work" life cycle: cf Trijonis."

73.   It seems to me abundantly clear that Sullivan J's view of the case was governed by his comparison of the life led by the respondents in the case before him with a normal life. In this I consider that he was wrong. I think that the Court of Appeal also failed to consider the correct factors in upholding Sullivan J's judgment, as appears from paragraph 14 of its judgment, where it refers with apparent approval to his consideration of "the extent to which restrictions would prevent an individual from pursuing the life of his choice, whatever that choice might be."
74.   Although many decisions of the European Court of Human Rights and the Commission were cited to your Lordships, the only one which contained any sustained discussion of the governing principle was Guzzardi v Italy (1980) 3 EHRR 333. It was analysed repeatedly in counsel's written submission and oral arguments, and your Lordships have all commented on it in greater or lesser detail. The principles discussed in it have been repeated and applied in subsequent Strasbourg cases, and the decision itself has been distinguished in several, but without critical analysis, and I think it of assistance now to make a further examination of the terms of the judgments.
75.   Mr Guzzardi, as your Lordships have rehearsed, who was suspected (and later convicted) of serious Mafia-linked crimes, was, instead of being remanded in custody, sent to live under "special supervision" on a portion of the small island of Asinara off Sardinia, most of which was occupied by a prison complex to which entry by persons under special supervision was forbidden. They had to live in the small hamlet of Cala Reale in somewhat primitive accommodation. They could not go to the neighbouring village of Cala d'Oliva and were effectively cut off from much human contact apart from persons in compulsory residence and those supervising them. He was subject to a curfew between 10 pm and 7 am, and a number of restrictions were placed on his activities and association with other people. The portion of the island to which he was confined, some 2.5 square kilometres in area, was described by Guzzardi himself as a pezzo or pezzetto di terra ("a scrap of land") and he also described the island as a "veritable concentration camp".
76.   The Court considered the issue whether the case fell within article 5 in paragraphs 92-95 of its judgment, much of which bears repetition in order to follow its reasoning:
"92.  The Court recalls that in proclaiming the 'right to liberty', paragraph 1 of Article 5 is contemplating the physical liberty of the person; its aim is to ensure that no one should be dispossessed of this liberty in an arbitrary fashion. As was pointed out by those appearing before the Court, the paragraph is not concerned with mere restrictions on liberty of movement; such restrictions are governed by Article 2 of Protocol No 4 which has not been ratified by Italy. In order to determine whether someone has been 'deprived of his liberty' within the meaning of Article 5, the starting point must be his concrete situation and account must be taken of a whole range of criteria such as the type, duration, effects and manner of implementation of the measure in question.

93. The difference between deprivation of and restriction upon liberty is nonetheless merely one of degree or intensity, and not one of nature or substance. Although the process of classification into one or other of these categories sometimes proves to be no easy task in that some borderline cases are a matter of pure opinion, the Court cannot avoid making the selection upon which the applicability or inapplicability of Article 5 depends.

94. As provided for under the 1956 Act (see paras 48-49 above), special supervision accompanied by an order for compulsory residence in a specified district does not of itself come within the scope of Article 5. The Commission acknowledged this: it focused its attention on Mr Guzzardi's 'actual position' at Cala Reale (see paras 5, 94, 99, etc, of the report) and pointed out that on 5 October 1977 it had declared inadmissible application No 7960/77 lodged by the same individual with regard to his living conditions at Force. It does not follow that 'deprivation of liberty' may never result from the manner of implementation of such a measure, and in the present case the manner of implementation is the sole issue that falls to be considered (see para 88 above).

95. The Government's reasoning (see para 91 above) is not without weight. It demonstrates very clearly the extent of the difference between the applicant's treatment on Asinara and classic detention in prison or strict arrest imposed on a serviceman. Deprivation of liberty may, however, take numerous other forms. Their variety is being increased by developments in legal standards and in attitudes; and the Convention is to be interpreted in the light of the notions currently prevailing in democratic States.

Whilst the area around which the applicant could move far exceeded the dimensions of a cell and was not bounded by any physical barrier, it covered no more than a tiny fraction of an island to which access was difficult and about nine-tenths of which was occupied by a prison. Mr Guzzardi was housed in part of the hamlet of Cala Reale which consisted mainly of the buildings of a former medical establishment which were in a state of disrepair or even dilapidation, a carabinieri station, a school and a chapel. He lived there principally in the company of other persons subjected to the same measure and of policemen. The permanent population of Asinara resided almost entirely at Cala d'Oliva, which Mr Guzzardi could not visit, and would appear to have made hardly any use of its right to go to Cala Reale. Consequently, there were few opportunities for social contacts available to the applicant other than with his near family, his fellow 'residents' and the supervisory staff. Supervision was carried out strictly and on an almost constant basis. Thus, Mr Guzzardi was not able to leave his dwelling between 10 pm and 7 am without giving prior notification to the authorities in due time. He had to report to the authorities twice a day and inform them of the name and number of his correspondent whenever he wished to use the telephone. He needed the consent of the authorities for each of his trips to Sardinia or the mainland, trips which were rare and, understandably, made under the strict supervision of the carabinieri. He was liable to punishment by 'arrest' if he failed to comply with any of his obligations. Finally, more than 16 months elapsed before his arrival at Cala Reale and his departure for Force (see paras 11, 12, 21, 23—42 and 51 above).

It is admittedly not possible to speak of 'deprivation of liberty' on the strength of any one of these factors taken individually, but cumulatively and in combination they certainly raise an issue of categorisation from the viewpoint of Article 5. In certain respects the treatment complained of resembles detention in an 'open prison' or committal to a disciplinary unit …

The Court considers on balance that the present case is to be regarded as one involving deprivation of liberty."

    The reference to Force is to the fact that Guzzardi was transferred from Asinara to Force, which was a remote country district but did not have the unusual characteristics of Cala Reale on Asinara. He again challenged the order requiring him to live there, but the Commission held in 1977 the application inadmissible. The concluding phrase of paragraph 92, which is the sheet anchor of the appellant's contention that article 5 applies, comes directly from paragraph 59 of Engel v The Netherlands (No 1) (1976) 1 EHRR 647. It is argued that it entitles the court to look at the conditions of life of the person the subject of the control order, and if the restrictions give rise to a sufficiently fundamental alteration of his lifestyle the case may come within article 5.

77.   That this is a questionable assumption may be seen by comparing the dissenting opinion in Guzzardi of Judge Sir Gerald Fitzmaurice. Judge Fitzmaurice came to the conclusion that article 5 did not apply, but it is apparent he regarded the question as one of degree, and in his judgment the restrictions did not amount to deprivation of Guzzardi's liberty. In posing the issue whether Guzzardi's situation amounted to deprivation of liberty or was essentially in the nature of a restriction on freedom of movement and choice of residence, he stated in paragraph 5 of his judgment that this
"must in the long run remain a matter of appreciation and opinion, namely whether the conditions of the applicant's existence on Asinara were sufficiently stringent to amount to a sort of imprisonment, even though a mild one as imprisonments go, or whether on the other hand, there was no more than a banishment accompanied by measures of confinement to house and grounds but, subject to that, without any restriction on movement within an area of at least a half-mile radius, or more according to some accounts. This could be argued about endlessly and either view is reasonably maintainable - for the issue is essentially one of degree."

78.   Judge Fitzmaurice went on to contrast article 5 with article 2(1) of Protocol No 4. In paragraph 6 he drew certain deductions:
"(a)  The existence of this provision [article 2 of Protocol No 4] shows either that those who originally framed the Convention on Human Rights did not contemplate that its Article 5 should go beyond preventing actual deprivation of liberty, or to extend to mere restrictions on freedom of movement or choice of residence; or else that the Governments of the Council of Europe did not see Article 5 as covering measures of 'deprivation of liberty' where the basic character of those measures consisted primarily of restrictions on movement and place of residence, or they would not have considered it necessary to draw up a separate Protocol about that. The resulting picture is that Article 5 of the Convention guaranteed the individual against illegitimate imprisonment, or confinement so close as to amount to the same thing—in sum against deprivation of liberty stricto sensu—but it afforded no guarantee against restrictions (on movement or place of residence) falling short of that. The latter was effected only by the Protocol, so that in those countries (of which Italy is one) that have not ratified it, such restrictions are not prohibited.

(b)  It follows that if Article 5 of the Convention is not to impinge on ground intended to be covered by Article 2 of the Protocol, and is not to do double duty with the latter, it (Art. 5) must be interpreted strictly and regarded as limited to cases of actual imprisonment or to detention close and strict enough to approximate to a virtually complete deprivation of liberty. This was certainly not the situation in regard to the applicant in the present case.

(c)  If Article 5 of the Convention were to be interpreted so widely as to include instances of what was basically restriction on freedom of movement or choice of residence, then not only would Article 2 of the Protocol be rendered otiose, but an indirect means would be afforded of making Governments subject to the obligations of the latter, despite the fact that they had not ratified the Protocol. This could not have been intended, but it is a possibility that can only be avoided by a strict interpretation of Article 5 that confines it to its proper sphere."

    In so defining deprivation of liberty for the purposes of article 5 I do not understand Judge Fitzmaurice to have been taking issue with the majority of the court on the principles to be applied; rather he differed from them on the application to the instant case of those principles. What he stated seems to me to be consistent with a proper interpretation of the majority judgment and to furnish a clear analysis of the ambit of the article.

79.   The key in both judgments to the meaning of deprivation of liberty is, I think, to be found in the majority's comparison of Guzzardi's situation with detention in an open prison or committal to a disciplinary unit and in Judge Fitzmaurice's phrase "illegitimate imprisonment, or confinement so close as to amount to the same thing". It was in this context that the court referred to taking account of "a whole range of criteria such as the type, duration, effects and manner of implementation of the measure in question." In saying this the court was in my view doing no more than recognising that the situations in which a person may be confined or restricted may vary in many ways from the archetypical case of imprisonment in a cell, but still amount to deprivation of liberty. To take the phrase out its context and use it to reach the conclusion that a variety of restrictions which prevent the person from enjoying a normal life is not in my opinion legitimate.
80.   Guzzardi has been applied or distinguished in a number of later cases before the ECtHR or the Commission, but, as I have said, in none of them is there any analysis which would give real assistance in determining the proper ambit of deprivation of liberty. In one line of house arrest cases, represented by NC v Italy (11 January 2001), [2001] ECHR 12, Mancini v Italy (12 December 2001), [2001] ECHR 502, Vachev v Bulgaria (8 October 2004), [2004] ECHR 325, Nikolova (No 2) v Bulgaria (30 December 2004), [2004] ECHR 462, and Pekov v Bulgaria (30 June 2006), [2006] ECHR 299 , it was held that confinement to the subject's house constituted a deprivation of liberty. It did not matter that the supervision of the confinement may have been so lax, as was claimed in Pekov v Bulgaria, that the applicant could in fact leave his house with impunity.
81.   On the other side of the line were a series of cases of compulsory residence imposed by orders made by the Italian Government, in each of which the Commission or the Court rejected the complaints: Ciancimino v Italy (1991) 70 D & R 103, Raimondo v Italy (1994) 18 EHRR 237 and Labita v Italy (6 April 2000), [2000] ECHR 161. In these cases the Strasbourg organs looked at the restrictions, contrasted them with those in Guzzardi, and concluded that they did not amount to deprivation of liberty. In Labita's case the applicant did not allege a breach of article 5(1), and the Grand Chamber concentrated on article 2 of Protocol No 4.
82.   A case on which counsel for the Secretary of State placed some reliance was Trijonis v Lithuania (17 March 2005), [2005] ECHR 875. The applicant was permitted to attend his place of work during the week, but was subject on those days to a curfew between 7 pm and 7 am, while he was obliged to remain in his house during the whole of each week-end. The Court briefly contrasted the case with that of Guzzardi and held that there was not a deprivation of liberty, going on to consider article 2 of Protocol No 4.
83.   It is to my mind notable that the assiduity of counsel has not brought before the House any case in which the ECtHR has held on facts at all comparable with those of the present appeals that there was a deprivation of liberty. If nothing more, this should make your Lordships feel the need to exercise some caution lest they depart from the current of the Strasbourg case-law: see R (Ullah) v Special Adjudicator [2004] 2 AC 323, 350, para 20, per Lord Bingham of Cornhill. Similar caution is required in recourse to the facts of individual Strasbourg cases and their use as factual precedents (R (Gillan) v Commissioner of Police of the Metropolis [2006] 2 AC 307, 342, para 23, per Lord Bingham of Cornhill), a tendency which does appear in the judgments of the courts below in these appeals. In the absence of any subsequent exposition of principle in the case-law, Guzzardi's case remains the fount, and for the reasons I have given I do not think that it supports the conclusions which have been drawn from it.
84.   I accordingly consider that Sullivan J and the Court of Appeal in JJ and others did not approach the meaning of deprivation of liberty in the proper fashion, and I would allow the appeal of the Secretary of State in these cases. I am conscious of the concern which some of your Lordships have felt about the effect of a curfew as long as 18 hours per day, and I would not dismiss that concern lightly. I conclude, however, that on balance even that very long curfew does not take the cases of JJ and others over the line of deprivation of liberty. I am not disposed to enter into discussion of the length of time which would take a case over that line. A great deal depends on the overall factual matrix of any given case. Moreover, I feel that the House ought to focus more on the principles to be followed than in giving detailed directions.
85.   These conclusions make it unnecessary to decide on the question of remedy, but since the issue was fully argued before the House I should perhaps express a very brief view on it. The court is empowered by section 3(12) of the 2005 Act to quash one or more of the obligations imposed by the Secretary of State's order. The detailed nature of the obligations in a typical control order may be seen from the summary of the order made in respect of LL contained in paragraph 33 of Lord Hoffmann's opinion. I incline to the view that the power to quash obligations is intended to cover a case where the court takes the view that the control order is justified but one of the fairly peripheral obligations is not, eg if one of the 5 individuals named in the non-association obligation numbered (7) in LL's case was wrongly included. If, however, the court applying the proper test concludes that the control order constitutes a deprivation of liberty because it is in essence comparable to imprisonment, then I would agree with the conclusion reached by Lord Bingham that it can and should quash the order, leaving the Secretary of State to reconsider the case.

LORD BROWN OF EATON-UNDER-HEYWOOD

My Lords,

86.   Control orders are highly contentious. Many think them essential as a means of providing some protection at least against suspected terrorists, the very minimum which government should do in fulfilment of its undoubted obligation to safeguard public security. Others abhor the whole notion of preventive action against people not even to be charged with a criminal offence and question whether the control order regime, like internment in the past, does not create more terrorists than it disables. That, however, is a debate for the House in its legislative capacity, not for your Lordships in the Appellate Committee. Rather your Lordships have to decide certain very different questions as to the legality of the control order regime, and in particular its compatibility with the European Convention on Human Rights.
87.   Between them the four appeals now before the House raise a number of different issues: see Secretary of State for the Home Department v JJ and others [2007] UKHL 45 , Secretary of State for the Home Department v MB; Secretaryof State for the Home Department v AF [2007] UKHL 46 and Secretary of State for the Home Department v E and another [2007] UKHL 47. Paramount amongst them, however, and lying at the very heart of this (the JJ) appeal, is the contention that control orders which impose, as these six control orders did, eighteen hour curfews necessarily involve a deprivation of liberty contrary to article 5 of the European Convention on Human Rights (the Convention). That contention succeeded before the Court of Appeal (Lord Phillips of Worth Matravers CJ, Sir Anthony Clarke MR and Sir Igor Judge P) [2007] QB 446. The Secretary of State now appeals.
88.   The facts of the appeal I gratefully take from the judgment below. Each of the respondents is a single man. Five are Iraqi nationals who have claimed asylum. They were arrested under the Terrorism Act 2000, released without charge, and then re-detained under immigration powers on notice of intention to deport on national security grounds. There is a dispute as to whether the sixth, LL, (who has since absconded and is now believed to be overseas) is an Iranian or an Iraqi national. He too was detained pending deportation on national security grounds. All deportation proceedings were discontinued on the making of the control orders.
89.   The obligations imposed by the control orders are essentially identical in all six cases. Each respondent is required to remain within his "residence" at all times, save for a period of six hours between 10 am and 4 pm. In the case of GG the specified residence is a one-bedroom flat provided by the local authority in which he lived before his detention. In the case of the other five respondents the specified residences are 1-bedroom flats provided by the National Asylum Support Service. During the curfew period the respondents are confined in their small flats and are not even allowed into the common parts of the buildings in which these flats are situated. Visitors must be authorised by the Home Office, to which name, address, date of birth and photographic identity must be supplied. The residences are subject to spot searches by the police. During the six hours when they are permitted to leave their residences, the respondents are confined to restricted urban areas, the largest of which is 72 square kilometres. These deliberately do not extend, save in the case of GG, to any area in which they lived before. Each area contains a mosque, a hospital, primary healthcare facilities, shops and entertainment and sporting facilities. The respondents are prohibited from meeting anyone by pre-arrangement who has not been given the same Home Office clearance as a visitor to the residence.
90.   The Strasbourg jurisprudence makes plain that the article 5 concept of deprivation of liberty is autonomous and that the court's task in cases like this is to decide whether the restrictions in question amount to a deprivation of liberty within article 5 or merely to a restriction upon liberty of movement within article 2 of Protocol No 4 (a Protocol not in fact ratified by the UK). Deprivation of liberty can only ever be justified if brought within one of paragraphs (a)-(f) of article 5(1)—none of which are available in the present type of case notwithstanding that the judge may well conclude both that there are reasonable grounds for suspecting the person concerned to be or have been involved in terrorist-related activity and that it is necessary to make this restrictive a control order to protect the public from a risk of terrorism. Restriction of movement, on the other hand, can be justified in the public interest under article 2(4) of Protocol No 4.
91.   That, then, is the Court's task: to decide into which category the case falls and, as was made plain by the ECtHR in Guzzardi v Italy (1980) 3 EHRR 333 —still the leading Strasbourg authority on the point—the distinction between the two categories is one of degree or intensity, not of kind:
"The difference between deprivation of and restriction upon liberty is nonetheless merely one of degree or intensity, and not one of nature or substance. Although the process of classification into one or other of these categories sometimes proves to be no easy task in that some borderline cases are a matter of pure opinion, the Court cannot avoid making the selection upon which the applicability or inapplicability of article 5 depends." (Para 93).

92.   The Court of Appeal (agreeing with Sullivan J at first instance) concluded that "the facts of this case fall clearly on the wrong side of the dividing line". Was the Court of Appeal right in that conclusion? That is the critical question.
93.   Before seeking further assistance from Guzzardi, it is worth noting one earlier Strasbourg authority, Engel v The Netherlands (No 1) (1976) 1 EHRR 647. Engel was concerned with disciplinary measures against members of the armed forces and considered four measures in particular: in ascending order of severity, (a) light arrest involving confinement to military buildings or premises or the serviceman's dwelling during off-duty hours, (b) aggravated arrest involving the serviceman spending his off-duty hours in a specially designated place and being unable to visit the canteen, cinema or recreation rooms, (c) strict arrest during which the serviceman was locked in a cell both by day and by night for up to 14 days and so unable to perform his normal duties, and (d) committal to a disciplinary unit unable to leave for upwards of a month. Only (c) and (d) were held to involve the deprivation of liberty. The Court said that article 5 "is contemplating individual liberty in its classic sense, that is to say the physical liberty of the person. Its aim is to ensure that no one should be dispossessed of this liberty in an arbitrary fashion." Because, however, of "the specific demands of military service", the Court held that:
"the bounds that article 5 requires the state not to exceed are not identical for servicemen and civilians. A disciplinary penalty or measure which on analysis would unquestionably be deemed a deprivation of liberty were it to be applied to a civilian may not possess this characteristic when imposed upon a serviceman."

    It appears to follow that the Court contemplated that the imposition of measures akin to (a) or (b) upon a civilian might well constitute a deprivation of liberty although account would always have to be taken of "a whole range of factors such as the nature, duration, effects and manner of execution of the penalty or measure in question."

94.   Guzzardi concerned the confinement of a Mafia suspect for over 16 months within a 2.5 square kilometre area (a corner of Asinara island) reserved for persons in compulsory residence, subject to a nine hour (10 pm to 7 am) curfew and almost permanent supervision and with only limited opportunities for social contact. Concluding "on balance" by a majority of eleven to seven that Mr Guzzardi was deprived of his liberty, the Court said (para 95):
"It is admittedly not possible to speak of 'deprivation of liberty' on the strength of any one of these factors [the constricting circumstances of his confinement] taken individually, but cumulatively and in combination they certainly raise an issue of categorisation from the viewpoint of article 5. In certain respects the treatment complained of resembles detention in an 'open prison' or committal to a disciplinary unit."

    It is important to understand, however, that the Court was not there holding that only treatment resembling detention in an open prison or committal to a disciplinary unit can amount to deprivation of liberty. On the contrary, notwithstanding its recognition that there were clear differences between Mr Guzzardi's treatment and "classic detention in prison or strict arrest imposed on a serviceman" - most notably that for fifteen hours a day he was free to leave and return to his dwelling as he wished, that his wife and son lived with him for fourteen of the sixteen months of his confinement, "the inviolability of his home and of the intimacy of his family life, two rights that the Convention guaranteed solely to free people", and with regard to "his social relations" - the Court nevertheless continued:

"Deprivation of liberty may, however, take numerous other forms. Their variety is being increased by developments in legal standards and in attitudes; and the Convention is to be interpreted in the light of the notions currently prevailing in democratic States."

    The Court also echoed (at para 92) what had been said in Engel as to the focus of article 5 being on "physical liberty" and "the starting point [being the applicant's] concrete situation and account must be taken of a whole range of criteria such as the type, duration, effects and manner of implementation of the measure in question."

95.   By "manner of implementation of the measure in question", the Court indicated (para 94), as indeed was already apparent from para 88, that it was concerned not with the Italian legislation authorising Mr Guzzardi's confinement but rather with his "actual position" during it.
96.   Guzzardi, of course, was decided over a quarter of a century ago. But subsequent cases, whilst establishing certain parameters beyond which it is now clear one way or the other whether article 5 applies, afford little additional assistance. The borderline between deprivation of liberty and restriction upon liberty remains indistinct and around it decisions necessarily remain "a matter of pure opinion."
97.   One parameter is represented by a series of Italian Mafia cases where the applicants were subject to internal exile regimes much like Mr Guzzardi's but in larger areas and communities than Mr Guzzardi on Asinara. One such case, indeed, was that of Mr Guzzardi himself when, after Asinara, he was required to live in the small, remote mainland district of Force. Other such cases are Ciancimino v Italy (1991) 70 DR 103, Raimondo v Italy (1994) 18 EHRR 237 and Labita v Italy (6 April 2000, App No 26772/95), [2000] ECHR 161. All involved curfews of between nine and eleven hours. All (save Labita where no article 5 complaint was even made) were summarily ruled inadmissible either by the Commission or the Court. So too was Trijonis v Lithuania (17 March 2005, at App No 2333/02), [2005] ECHR 875 where for sixteen months the applicant was subject to a 12 hour nightly curfew during weekdays and an entire weekend curfew (60 hours from 7 pm Friday to 7 am Monday). He, however, lived in his own house (which was inviolable) and was free from all restraints outside his home and so was able mid-week, to work or meet people as he wished.
98.   The other end of the spectrum is represented by the "house arrest" cases in which the respective applicants were required (sometimes as a "less restrictive" measure than pre-trial detention) to remain at home at all times save by prior permission of the authorities. These cases include NC v Italy (11 January 2001 App No 24952/94), [2001] ECHR 12, Mancini v Italy (12 December 2001 App. No. 44955/98), [2001] ECHR 502, Vachev v Bulgaria (8 October 2004), [2004] ECHR 325, Nikolova (No 2) v Bulgaria (30 December 2004), [2004] ECHR 462 and Pekov v Bulgaria (30 June 2006), [2006] ECHR 299 All were accepted or held to involve a deprivation of liberty. In Pekov the Government had argued that because in fact the applicant could have left his house with impunity (the monitoring authorities being based elsewhere) he was not deprived of his liberty. Unsurprisingly the argument failed.
99.   Plainly the present cases fall comfortably within the wide spectrum between those parameters. These 18-hour curfews (not to mention the additional constraints placed upon the respondents whether at home or away from it) are substantially more restrictive than those imposed in the Italian Mafia cases, or even those imposed in Trijonis. But they certainly do not amount to around the clock house arrest.
100.   Nor is the direct comparison with the facts of Guzzardi (itself plainly a borderline case) especially helpful. In certain respects the respondents have greater physical liberty than Mr Guzzardi on Asinara: in the non-curfew period they can move around in a larger area with greater facilities and a better opportunity of meeting people (subject, of course, to clearance). But in other respects they have less physical liberty, most notably in their confinement to their small flats for all but six hours a day.
101.   My noble and learned friend Lord Bingham of Cornhill warned in R (Gillan) v Commissioner of Police of the Metropolis [2006] 2 AC 307, 342, of the dangers of attempting to apply Strasbourg judgments as factual precedents:
"The Strasbourg jurisprudence is closely focused on the facts of particular cases, and this makes it perilous to transpose the outcome of one case to another where the facts are different."(para 23)

    The warning is salutary. To my mind no real assistance here can be gained from other such very different cases as X v Austria (1979) 18 DR 154 (forcible submission to a blood test), X v Federal Republic of Germany (1981) 24 DR 158 (a ten year old girl's detention for two hours in a police station for questioning) and Bozano v France (1986) 9 EHRR 297 (enforced travel by car for some twelve hours handcuffed to policemen for several hundred kilometres to the Swiss border).

102.   Ultimately, therefore, these appeals fall to be decided as "a matter of pure opinion" with little further guidance than that deprivation of liberty is concerned with "physical liberty", that it can take "numerous other forms", [other, that is, than "classic detention in prison or strict arrest imposed on a serviceman"], and that it is to be distinguished from mere restriction upon liberty as a question of "degree or intensity", starting with the applicants' "concrete situation" and then by reference to "a whole range of criteria such as the type, duration [and] effects" of that situation.
103.   Plainly there must come a point at which a daily curfew (itself clearly a restriction upon liberty of movement) shades into a regime akin to house arrest, where so little genuine freedom is left that the line is crossed into deprivation of liberty. The 2005 Act itself recognises that control orders could be made that are so onerous as to cross that line and require derogation from article 5 - and it recognises too that physical liberty is so important a freedom that not only must there then be derogation but also a substantially higher threshold for the imposition of such deprivation: proof that the person concerned actually is or has been involved in terrorist-related activity, rather merely than that there are reasonable grounds for suspecting this.
104.   At what point, then, is the line crossed? The question plainly is one for the courts, not for the Secretary of State. She understandably wants to impose in these cases the longest curfews consistent with non-derogation - doubtless to reduce so far as possible the need for surveillance (a scarce and now presumably over-stretched resource) and the suspects' opportunity to engage in terrorist-related activity. She contends for 18 hours. But there is no particular logic in this. Why not 20 hours, or 22? No useful comparison can be made with actual imprisonment. Indeed, conditions of imprisonment vary hugely. Some of those in open prisons daily go out to work unsupervised.
105.   Taking account of all the other conditions and circumstances of these control orders - broadly similar not only in these six cases but in the other cases heard with them - and not least the length of time for which they are imposed, I have reached the clear conclusion that 18 hour curfews are simply too long to be consistent with the retention of physical liberty. In my opinion they breach article 5. I am equally clear, however, that 12 or 14-hour curfews (those at issue in two of the related appeals before the House) are consistent with physical liberty. Indeed, I would go further and, rather than leave the Secretary of State guessing as to the precise point at which control orders will be held vulnerable to article 5 challenges, state that for my part I would regard the acceptable limit to be 16 hours, leaving the suspect with 8 hours (admittedly in various respects controlled) liberty a day. Such a regime, in my opinion, can and should properly be characterised as one which restricts the suspect's liberty of movement rather than actually deprives him of his liberty. That, however, should be regarded as the absolute limit. Permanent home confinement beyond 16 hours a day on a long-term basis necessarily to my mind involves the deprivation of physical liberty. And, although naturally I recognise that this cannot be the touchstone for the distinction, I think that any curfew regime exceeding 16 hours really ought not to be imposed unless the court can be satisfied of the suspect's actual involvement in terrorism, the higher threshold that would apply to the making of a derogating control order.
106.   I would add just this. I have given anxious thought to what Lord Bingham of Cornhill said in R (Ullah) v Special Adjudicator [2004] 2 AC 323, 350 (para 20) about not construing the Convention as conferring greater rights than the Strasbourg jurisprudence itself establishes—something upon which, indeed, I myself commented in R (Al-Skeini) v Secretary of State for Defence [2007] 3 WLR 33, 71-72 (paras 105-106). But whereas the issue in Al-Skeini was as to the reach of article 1 itself—an issue to which the ECtHR in Bankovic v Belgium (2001) 11 BHRC 435 (at paras 64 and 65) had made plain that the "living instrument" approach does not apply—here by contrast the Court recognised in Guzzardi (at para 95 in the passage already quoted in para 94 above) that developing legal standards and attitudes will further increase the variety of forms of deprivation of liberty. I think that nowadays a longer curfew regime than 16 hours a day (with the additional restraints imposed in these cases) would surely be classified in Strasbourg as a deprivation of liberty. It may be, indeed, that 16 hours itself is too long. I would, however, leave it to the Strasbourg Court to decide upon that, were any such argument to be addressed to it. (The government itself, of course, cannot complain to Strasbourg about adverse decisions of your Lordships' House.)
107.   Finally I would say this. The borderline between deprivation of liberty and restriction of liberty of movement cannot vary according to the particular interests sought to be served by the restraints imposed. The siren voices urging that it be shifted to accommodate today's need to combat terrorism (or even that it be drawn with such need in mind) must be firmly resisted. Article 5 represents a fundamental value and is absolute in its terms. Liberty is too precious a right to be discarded except in times of genuine national emergency. None is suggested here.
108.   Since writing the paragraphs set out above I have had the advantage of reading in draft the opinions of each of the other members of the Committee. Despite the explicit reluctance of several of your Lordships to suggest the point at which curfews would, by virtue of their length, involve the deprivation of liberty, I remain unrepentant for doing so. I recognise, of course, that "situations may be many and various" (Baroness Hale of Richmond at para 63), that "the overall factual matrix" is important (Lord Carswell at para 84) and that the decision whether or not a particular non-derogating control order involves a deprivation of liberty is one for the judge, appealable only for error of law. As mentioned, however, the other conditions and circumstances of these six control orders (and, indeed, those under consideration in the related appeals) are all broadly similar and, as Lord Bingham points out in para 11 of his opinion in the E & S appeal [2007] UKHL 47, what principally must be focused on is the extent to which the suspect is "actually confined": "other restrictions (important as they may be in some cases) are ancillary" and "[can] not of themselves effect a deprivation of liberty if the core element of confinement . . . is insufficiently stringent." Just so that there is no mistake about it, my view is that, taking account of the conditions and circumstances in all these various control order cases, provided the "core element of confinement" does not exceed sixteen hours a day, it is "insufficiently stringent" as a matter of law to effect a deprivation of liberty. Beyond sixteen hours, however, liberty is lost.
109.   It follows that I respectfully agree with Lord Bingham and Baroness Hale that in the JJ appeal Sullivan J and the Court of Appeal were right to hold that these six control orders involved a deprivation of liberty having regard to their "type, duration, effects and manner of implementation." That being so, I respectfully agree also (with each of your Lordships save for Lord Hoffmann, and for the reasons given by Lord Bingham) that there was no alternative here but to quash the orders in their entirety rather than strike down part of them only or direct their modification by the Secretary of State.

    

110. Accordingly I too would dismiss this appeal.