In the matter of D (A Child)

The common law has long protected the liberty of the subject, through the machinery of habeas corpus and the tort of false imprisonment. Likewise, article 5 of the European Convention on Human Rights begins: “Everyone has the right to liberty and security of person. No one shall be deprived of his liberty save in the following cases and in accordance with a procedure prescribed by law”. In Storck v Germany(2005) 43 EHRR 6, paras 74 and 89, confirmed by the Grand Chamber in Stanev v Bulgaria (2012) 55 EHRR 22, paras 117 and 120, and adopted by this court in Surrey County Council v P; Cheshire West and Chester Council v P [2014] UKSC 19; [2014] AC 896 (commonly known as Cheshire West), para 37, the European Court of Human Rights held that there were three components in a deprivation of liberty for the purpose of article 5: (a) the objective component of confinement in a particular restricted place for a not negligible length of time; (b) the subjective component of lack of valid consent; and (c) the attribution of responsibility to the State.

At the same time, the common law and equity have long recognised the authority of parents over their minor children, now encapsulated in the concept of “parental responsibility” in the Children Act 1989. Likewise, article 8 of the European Convention on Human Rights begins “Everyone has the right to respect for his private and family life, his home and his correspondence”; and, as this court recognised in Christian Institute v Lord Advocate [2016] UKSC 51; 2017 SC (UKSC) 29, paras 71 to 74, the responsibility of parents to bring up their children as they see fit, within limits, is an essential part of respect for family life in a western democracy.

This case is about the interplay between the liberty of the subject and the responsibilities of parents, between the rights and values protected by article 5 and the rights and values protected by article 8, and between the relationship of parent and child at common law and the Convention rights. The principal issue can be simply stated: is it within the scope of parental responsibility to consent to living arrangements for a 16 or 17-year-old child which would otherwise amount to a deprivation of liberty within the meaning of article 5? But this principal issue cannot sensibly be addressed without also considering further issues. What difference, if any, does it make that the child lacks the mental capacity to make the decision for himself? What difference, if any, does it make that the holder of parental responsibility is a public authority rather than an individual? Furthermore, although the concentration in this case is upon 16 and 17-year-old children, similar issues would arise in a case concerning a child under 16.

A further issue was raised by the court after the hearing: do the restrictions on placing children in accommodation provided for the purpose of restricting liberty, arising from section 25 of the Children Act 1989, apply to the sort of living arrangements in question here? We are grateful to the parties for their written submissions on this complicated issue. It is addressed by Lady Black at paras 91 to 115 of her judgment, with which I agree.

The child in question, D, was born on 23 April 1999, and so is now aged 20 and an adult. Nevertheless, the importance of the issues is such that this court gave the Official Solicitor, who acts for him as his litigation friend, permission to appeal from the decision of the Court of Appeal.

D was diagnosed with attention deficit hyperactivity disorder at the age of four, Asperger’s syndrome at seven, and Tourette’s syndrome at eight. He also has a mild learning disability. His parents struggled for many years to look after him in the family home, despite the many difficulties presented by his challenging behaviour. Eventually, in October 2013 when he was 14, he was informally admitted to Hospital B for multi-disciplinary assessment and treatment. Hospital B provided mental health services for children between the ages of 12 and 18. He lived in a unit in the hospital grounds and attended a school which was integral to the unit. The external door to the unit was locked and D was checked on by staff every half hour. If he left the site, he was accompanied by staff on a one to one basis. His visits home were supervised at all times.

In 2014, the Hospital Trust issued an application under the inherent jurisdiction of the Family Division of the High Court relating to children, seeking a declaration that it was lawful for the Trust to deprive D of his liberty and that this was in his best interests. In March 2015, Keehan J held: first, that the conditions under which D lived amounted to depriving him of his liberty (by which he meant confinement under limb (a) of Storck v Germany, para 1 above); the fact that he enjoyed living in the unit made no difference; second, that it was “within the zone of parental responsibility” for his parents to agree to what would otherwise be a deprivation of liberty; it was a proper exercise of parental responsibility to keep an autistic 15-year-old boy who had erratic, challenging and potentially harmful behaviours under constant supervision and control; but third, once he reached 16 he would come under the jurisdiction of the Court of Protection and the different regime there, largely contained in the Mental Capacity Act 2005: In re D (A Child) (Deprivation of Liberty)[2015] EWHC 922 (Fam); [2016] 1 FLR 142.

By then, the clinical team had agreed that D should be discharged from Hospital B to a residential placement. Birmingham City Council took the lead in making the arrangements for D to move to Placement B. His parents agreed to the move. On 23 April 2015, his 16th birthday, proceedings were issued in the Court of Protection. Birmingham sought a declaration that D would not be deprived of his liberty at Placement B because his parents could consent to it. On 20 May 2015, Keehan J made an order for D’s transfer from Hospital B to Placement B. This took place on 2 June 2015. D was accommodated there under section 20 of the Children Act 1989 and thus became a “looked after child” within the meaning of section 22(1) of that Act.

It has always been common ground between the parties to this case that the arrangements under which D lived at Placement B would have amounted to a deprivation of liberty were it not for his parents’ consent to them. Placement B was a large house set in its own grounds, with 12 residential units in the grounds, each with its own fenced garden. D lived with three other young people in House A. The external doors were locked. If he wanted to go out into the garden, he had to ask for the door to be unlocked. He was not allowed to leave the premises except for a planned activity, such as attending his school, which was also on the site, swimming and leisure activities. He received one to one support during waking hours and staff were in constant attendance overnight.

The application was heard by Keehan J in the Court of Protection in November 2015. In January 2016, he held: first, that the parents could no longer consent to what would otherwise be a deprivation of liberty now that D had reached 16; his principal reasons for doing so were that Parliament had, on numerous occasions, distinguished the legal status of those who had reached the age of 16 from that of those who had not; and that the Mental Capacity Act 2005 applied to people who had reached the age of 16. He also held that this deprivation of liberty was attributable to the state, a matter which is no longer in dispute: Birmingham City Council v D (by his litigation friend, the Official Solicitor) [2016] EWCOP 8; [2016] PTSR 1129.

Birmingham City Council appealed to the Court of Appeal. Before the hearing, D was transferred to Placement C, where the arrangements were not materially different from those in Placement B. Once again, his parents agreed to his being accommodated under section 20 of the 1989 Act, to the arrangements themselves and to the restrictions on D’s liberty which they entailed. On 23 November 2016, Keehan J authorised the placement and the deprivation of liberty. There has never been any doubt that both placements were in D’s best interests but that D himself did not have the capacity to consent to them.

The appeal was heard in February 2017, but judgment was not given until 31 October 2017. In the meantime, D had reached the age of 18 and parental responsibility for him ceased. However, by virtue of his age, it was now possible for a deprivation of liberty in a hospital or care home to be authorised under the deprivation of liberty safeguards in Schedule A1 to the Mental Capacity Act 2005 (which applied only to those aged 18 or over), as well as by the Court of Protection.

The Court of Appeal agreed with Keehan J that D’s accommodation in Placement B and Placement C was attributable to the State. However, the appeal was allowed on the ground that Keehan J had been wrong to hold that a parent could not consent to what would otherwise be a deprivation of the liberty of a 16 or 17-year-old child who lacked the capacity to decide for himself. Sir James Munby P, in a typically erudite judgment, traced the development of the common law responsibilities of parents in great detail. He concluded, first, that the approach of Keehan J did not give effect to the fundamental principle, established by the majority of the House of Lords in Gillick v West Norfolk and Wisbech Area Health Authority [1986] AC 112, that the exercise of parental responsibility comes to an end, not on the child’s attaining a fixed age, but on his attaining “Gillick capacity”; and second, that none of the statutory provisions upon which Keehan J had relied had a bearing on the matter in hand (para 125).

The Official Solicitor now appeals to this court on D’s behalf. The Official Solicitor’s primary case is that, whatever may once have been the position at common law, no person can consent to the confinement of a child who has reached the age of 16 and lacks the capacity to decide for himself. If such a child is confined, and that confinement is attributable to the state, he is deprived of his liberty within the meaning of article 5 and there must be safeguards to ensure that the deprivation is lawful. The Mental Capacity Act 2005 provides a complete decision-making framework for the care and treatment of people aged 16 and above who lack the capacity to decide for themselves. His alternative case is that, even if such consent is within the scope of parental responsibility, the person giving it should apply the principles and procedure for deciding whether the arrangements are in the child’s best interests set out in section 4 of the Mental Capacity Act 2005.

Birmingham City Council’s case is that it is within the scope of a parent’s lawful exercise of parental responsibility to authorise the confinement of a 16 or 17-year-old child who is not “Gillick competent” to consent. The common law as to the scope of parental responsibility in this respect has not been eroded by the Mental Capacity Act 2005 or by any other legislation. If it affects the exercise of parental responsibility at all, it does so only by substituting the concept of lack of “capacity” within the meaning of the Act for the concept of lack of “competence” within the meaning of Gillick (to the extent that these two may differ - an issue which does not arise on this appeal).

The Equality and Human Rights Commission have been given permission to intervene in this court, as they did in the Court of Appeal. Their case is that, while parental responsibility can in principle extend to the age of 18, whether it applies in particular circumstances has to be judged in the light, not only of the common law, but also of statute, the European Convention on Human Rights and other international instruments. Parents should not be able to consent to the confinement of a 16 or 17-year-old child, thereby removing the protections given by article 5 of the European Convention. Further, to remove those protections from a child who lacks Gillick competence because of a disability, while according them to a competent child, is unjustified discrimination on the ground of his or her disability.

The Secretaries of State for Education and for Justice did not intervene in the Court of Appeal but have been given permission to intervene jointly in this court. The Secretary of State for Education has policy responsibility for children and young people and depriving them of liberty; the Secretary of State for Justice has overall policy responsibility for the Mental Capacity Act 2005 and in relation to parental responsibility generally. Their case is, first, that a child will only be “confined” if the restrictions on his liberty go beyond those which would be imposed upon a child of the same age and relative maturity who is free from disability; and second, that even if a child is “confined”, a person with parental responsibility may provide a valid consent to that confinement if the child is not Gillick competent to make the decision for himself; however, a person with parental responsibility must be acting in the best interests of the child for there to be a proper exercise of that responsibility.

The Secretaries of State agree with both the Official Solicitor and Birmingham City Council that a local authority with parental responsibility by virtue of a care order or interim care order, or with any other statutory responsibilities for a child, cannot supply a valid consent to the confinement of a child (as Keehan J held in In re AB (A Child) (Deprivation of Liberty: Consent) [2015] EWHC 3125 (Fam), [2016] 1 WLR 1160; see also In re K (A Child) (Secure Accommodation Order: Right to Liberty) [2001] Fam 377, para 35 below). Only a natural person with parental responsibility can do so. Quite what the basis is for distinguishing between the content of parental responsibility according to the person who holds it is not explained: the Children Act 1989, which both defines and governs the allocation of parental responsibility, makes no such distinction.

This case turns on the inter-relationship between the concept of parental responsibility, as defined by the Children Act 1989, the common law and other relevant statutory provisions, and the obligation of the State to protect the human rights of children under the European Convention on Human Rights. The one cannot supply an answer without reference to the other. It makes sense, therefore, to begin with parental responsibility.

Parental responsibility is defined in the Children Act 1989 as “all the rights, duties, powers, responsibilities and authority which by law a parent of a child has in relation to the child and his property” (section 3(1)). “By law” obviously refers to the common law, but also includes those statutory provisions which give rights etc to parents, such as the Marriage Act 1949, which gives them the right to withhold consent to the marriage of a 16 or 17-year-old child (section 3).

As Lady Black explains in more detail in paras 55 to 68 of her judgment, the common law and equity recognised the parental power of a father over his legitimate minor children (the mother did not acquire equal rights and authority with the father until the Guardianship Act 1973): see the valuable account of the history by P H Pettitt, “Parental Control and Guardianship” in R H Graveson and FR Crane, A Century of Family Law (1957, Sweet & Maxwell). The high water-mark of this was the well-known case of In re Agar-Ellis (No 2) (1883) 24 Ch D 317, at 326, where the Master of the Rolls declared: “… the law of England … is, that the father has the control over the person, education and conduct of his children until they are 21 years of age. That is the law.” However, as Sachs LJ explained in Hewer v Bryant [1970] 1 QB 357, at 372, in passages quoted in full at para 55 of Lady Black’s judgment, a distinction was drawn between custody in the narrow sense of the right to physical possession of the child and custody in the wider sense of the right to control every aspect of the child’s life, including his religion, education and property. The common law courts would enforce the former by way of the writ of habeas corpus. But they would refuse to do this against the wishes of the child, once he or she had reached the “age of discretion”. In the 19th century, this was regarded as fixed at 14 for boys and 16 for girls (the latter by reference to the Abduction Acts of 1557 and 1828). The Court of Chancery would enforce all parental powers and authority, but by way of the Crown’s parens patriae jurisdiction rather than by way of enforcing the parent’s rights, potentially up to the age of majority. Parental rights were never absolute and became increasingly subject to the overriding consideration of the child’s own welfare. This was put on a statutory footing by the Guardianship of Infants Act 1925, which famously declared that “Where in any proceedings before any court … the custody or upbringing of an infant, or the administration of any property belonging to or held in trust for an infant, or the application of the income thereof, is in question, the court, in deciding that question, shall regard the welfare of the infant as the first and paramount consideration …” (section 1). That remains the guiding principle in section 1(1) of the Children Act 1989, which provides that “the child’s welfare shall be the court’s paramount consideration”.

Section 1, of course, deals with the position if a case about a child’s upbringing gets to court. But what about the powers and authority of holders of parental responsibility before a case gets to court? Two 20th century cases show how, whatever may have been the earlier position, the common law is capable of moving with the times. In Hewer v Bryant [1970] 1 QB 357, the issue was the meaning of “in the custody of a parent” in the Limitation Act 1954. The High Court had held that a child remained, by law, in the custody of his father until the age of majority, applying In re Agar-Ellis. The Court of Appeal held that a 15-year-old living away from home and working as an agricultural trainee was not in the custody of a parent for this purpose. “Custody” in the Limitation Act meant the actual exercise of powers of control. But both Lord Denning MR and Sachs LJ recognised that the parent’s legal powers of physical control diminished as the child got older. Sachs LJ expressly referred to the parent’s “ability to restrict the liberty of the person”, which lasted until the age of discretion, and distinguished between the parental power and the court’s power, which lasted until the age of majority (p 372). Lord Denning MR put it this way:

That dictum was approved by the majority of the House of Lords in Gillick v West Norfolk and Wisbech Area Health Authority [1986] AC 112, at pp 172, 186, 195. The issue was whether it was lawful to give contraceptive advice and treatment to a child under 16 without her parent’s consent if she herself was capable of giving that consent. The Court of Appeal had held that this would be infringing the inalienable and legally enforceable rights of parents relating to the custody and upbringing of their children which, except in emergencies, could only be overridden by a court. A girl under 16 was incapable either of consenting to treatment or prohibiting a doctor from seeking the consent of her parents. The House of Lords, by a majority, disagreed. The earlier “age of discretion” cases had established the principle that children could achieve the capacity to make their own decisions before the age of majority. It was no longer, if it ever had been, correct to fix that at any particular age, rather than by reference to the capacity of the child in question: it had already been established that a child below the age of 16 could consent to sexual intercourse so that it was not rape (R v Howard [1966] 1 WLR 13) or to being taken away so that it was not kidnapping (R v D [1984] AC 778). Parental rights and authority existed for the sake of the child, to enable the parent to discharge his responsibilities towards the child, and not for the sake of the parent. Lord Scarman put it thus (p 185):

As Lady Black explains in paras 69 to 72 of her judgment, the Gillick case is not directly relevant to the issue before us now. It had to do with medical treatment and not with deprivation of liberty. It was concerned with whether a child might acquire the capacity, and the right, to make such decisions for herself before she reached the common law age of discretion, not with whether parental authority endured beyond that age if the child lacked the capacity to decide for herself. And as Lady Black has shown, it is, to say the least, highly arguable that such authority did not extend to depriving such a child of her liberty once she had reached the age of discretion.

Some support for that conclusion is supplied by the earliest legislation dealing specifically with people with mental disabilities, the Mental Deficiency Act 1913. Section 6(3) provided both for detention in an institution and admission to guardianship. Section 10(2) provided that an order that a person be placed under guardianship conferred upon the person named as guardian “such powers as would have been exercisable if he had been the father of the defective and the defective had been under the age of 14”. That provision remained in force until repealed by the Mental Health Act 1959, under which the powers of a guardian were defined, by section 34(1), in materially identical terms. (The powers of a guardian are now more narrowly defined in the Mental Health Act 1983.) It is highly likely that it was contemplated that a guardian might have to accommodate the person under his guardianship in conditions which deprived him of his liberty. The fact that the guardian was therefore given the powers of the father of a child under 14, rather than the powers of the father of a child of any age, suggests that it was not then thought that a father had the right to deprive a child of any age of his liberty if the child lacked the capacity to make his own decisions. Many of the people falling within the definition of “mental defective” would lack that capacity.

Statute has also intervened to make specific provision for children who have reached the age of 16. In the Court of Protection, Keehan J referred to the following (para 64) although not in the same terms or the same order:

The age of 16 is significant for various other purposes, such as leaving school, joining the armed forces or getting married (albeit that parental consent is usually required). So Keehan J was correct to suggest that the law accords children who have reached 16 a status which is in some respects different from that of children under that age. However, Sir James Munby P was also correct to hold that these provisions do not supply the answer to the issues in this case. Items (iv) and (v) are concerned with the limits on making court orders. Items (i) and (ii) relate only to children who have the capacity to make a decision for themselves, and it is quite possible that item (vi) is also so limited. Furthermore, as Gillick holds, a child may acquire the capacity to make certain decisions for himself before the age of 16. We are concerned with the extent of parental responsibility for a child who lacks the capacity to make the decision for himself.

It may well be that, as a general rule, parental responsibility extends to making decisions on behalf of a child of any age who lacks the capacity to make them for himself. This would always be subject to the courts’ powers of intervention, whether at the behest of another parent or individual in private law proceedings under Part 2 of the Children Act 1989, or at the behest of a local authority in public law proceedings under Part 4 on the ground that the child is suffering or likely to suffer significant harm as a result of the parents’ decisions. The question, however, is whether there are any limits to that general rule, and in particular whether it is within the scope of parental responsibility to make arrangements which have the effect of depriving a child of his liberty. In view of the conclusion which I have reached as to the effect of article 5 of the European Convention on Human Rights, and the interaction between parental responsibility and the child’s rights under article 5, it is strictly unnecessary to reach a concluded view on that question. But I acknowledge the force of the conclusion reached by Lady Black at para 90 of her judgment. As she says, it reinforces the conclusion which I have reached for other reasons.

Article 1 of the European Convention on Human Rights requires the High Contracting Parties to secure to “everyone within their jurisdiction” the rights and freedoms set out in the Convention. There can be no doubt that “everyone” includes minor children, or indeed that the Convention rights may require adaptation to cater for their special needs as children: see, for example, the case of Thompson and Venables dealing with the fair trial rights of children accused of serious crime: T v United Kingdom(1999) 30 EHRR 121.

We are here concerned with article 5, which, as already stated, accords to “everyone … the right to liberty and security of person. No one shall be deprived of his liberty save in the [listed] cases and in accordance with a procedure prescribed by law”. That this applies to children is made clear by article 5(1)(d) which permits “the detention of a minor by lawful order for the purpose of educational supervision …”. That it applies to people who lack the capacity to make decisions for themselves is made clear by article 5(1)(e) which permits “the lawful detention of … persons of unsound mind”. Article 5(1) contains within it the requirement that decisions made under it are not arbitrary and accord with the Convention concept of legality: see, for example, HL v United Kingdom (2005) 40 EHRR 32. Article 5 also contains various specific procedural safeguards, including article 5(4), which requires that “everyone who is deprived of his liberty by arrest or detention shall be entitled to take proceedings by which the lawfulness of his detention shall be decided speedily by a court and his release ordered if the detention is not lawful”.

Prima facie, therefore, article 5 protects children who lack the capacity to make decisions for themselves from being arbitrarily deprived of their liberty. All parties to this case agree that this means that a local authority which has parental responsibility for a child cannot deprive the child of his liberty without the authority of a court. But, say Birmingham City Council and the Secretaries of State, the position is different if the parents - or other individuals with parental responsibility such as special guardians - agree to it. Why should that be?

The facts of RK v BCC, YB and AK [2011] EWCA Civ 1305; [2012] COPLR 146, were remarkably similar to the facts of this case. A young woman aged 17 suffered from autism, attention deficit hyperactivity disorder and severe learning difficulties, as well as epilepsy. She had been looked after at home for nearly 16 years but was then accommodated by the local authority under section 20 of the Children Act 1989 in a private care home. In proceedings brought by her mother in the Court of Protection, the Official Solicitor raised concerns that her living arrangements might amount to a deprivation of liberty. Mostyn J held: first, that the provision of accommodation under section 20 could never amount to a deprivation of liberty because the parents must have agreed to it; and second, that in any event the restrictions authorised by her parents did not amount to a deprivation of liberty. In relation to the first, the Court of Appeal upheld the consensus reached at the Bar (para 14):

That decision was, of course, before the Supreme Court’s decision in Cheshire West, which clarified the objective elements of a deprivation of liberty (limb (a) of Storck v Germany, para 1 above). The “acid test” is that a person is under continuous supervision and control and not free to leave. The fact that such restrictions may be necessary in order to prevent a person from harming himself or others makes no difference. Nor does the fact that the person’s living arrangements are as close to a normal home life as they could possibly be. It seems likely, therefore, that the conditions under which RK was living would now be regarded as depriving her of her liberty. But it is not clear why that should make any difference to the validity of the consensus reached at the Bar and endorsed by the court as to the scope of parental responsibility.

The basis for that consensus was said to be the case of Nielsen v Denmark (1988) 11 EHRR 175 in the European Court of Human Rights, together with the Court of Appeal decision in In re K (A Child) (Secure Accommodation Order: Right to Liberty) [2001] Fam 377. It is worth looking at the case of Nielsen v Denmark in some detail. It concerned a 12-year-old child of unmarried parents. The mother alone had custodial rights over him. Nevertheless, at the age of eight he had refused to return to his mother after a holiday with his father and then disappeared to live in hiding with his father for more than three years. He re-appeared with his father during proceedings in which his father was attempting to obtain a change in custody. His father was arrested and his mother requested that the boy be admitted to the State Hospital’s child psychiatric ward as it was clear that he did not want to stay with her. He was eventually discharged after six months and went to live with another family. However, five months after that, the Danish Supreme Court awarded custody to the father.

The boy complained to the European Court of Human Rights that his rights under article 5(1) and 5(4) of the European Convention had been breached. The European Commission found, by 11 votes to one, that there had been a breach of article 5(1) and by ten votes to two that there had been a breach of article 5(4). However, the Court found, by nine votes to seven, that there had been no breach. It is not easy to identify how the majority’s decision fits into the tri-partite scheme of later decisions such as Storck v Germany and Stanev v Bulgaria. The majority held that article 5 was “not applicable in so far as it is concerned with deprivation of liberty by the state” (para 64), a conclusion which was strongly disputed by the dissenters who pointed out that the boy was detained in a State Hospital, with the agreement of the responsible psychiatrist, and was returned to hospital by the police when he disappeared on the day that he was due to be discharged into his mother’s care. Nevertheless, the majority went on to consider whether article 5 was applicable “in regard to such restrictions on the applicant’s liberty as resulted from the exercise of the mother’s parental rights” (para 64). This can only be explained on the basis that, again as the minority observed, article 5 imposes a positive obligation upon the state to protect individuals from being deprived of their liberty by private persons.

The majority went on to consider the boy’s actual situation. They found that he was in need of medical treatment for his “nervous condition”. The treatment did not involve medication, but consisted of regular talks and “environmental therapy”. The restrictions on his freedom of movement and contact with the outside world were “not much different from restrictions which might be imposed on a child in an ordinary hospital”. In general “conditions in the ward were said to be ‘as similar as possible to a real home’” (para 70). The restrictions to which the boy was subject were “no more than the normal requirements for the care of a child of 12 years of age receiving treatment in hospital”. They were not in principle different from those in many hospital wards where children with physical disorders were treated. The boy was still at an age at which it would be normal for a decision to be made by the parent even against the wishes of the child (para 72). Accordingly, “the hospitalisation of the applicant did not amount to a deprivation of liberty within the meaning of article 5, but was a responsible exercise by his mother of her custodial rights in the interest of the child” (para 73).

The minority considered that the placing of a 12-year-old boy who was not mentally ill for several months in a psychiatric ward was a deprivation of liberty for which the state was accountable.

Whether one agrees with the majority or the minority assessment of the facts of the case, the majority judgment clearly turned on the comparative normality of the restrictions imposed upon the freedom of a 12-year-old boy. In In re K (A Child) (Secure Accommodation Order: Right to Liberty), the Court of Appeal was faced with an argument that the regime for authorising the placement of a child in secure accommodation, under section 25 of the Children Act 1989, was incompatible with the right to liberty in article 5. The first issue was whether such a placement was indeed a deprivation of liberty, even though it was agreed to by the local authority, which had parental responsibility under an interim care order, as well as by the child’s parents. Dame Elizabeth Butler-Sloss P and Judge LJ held that it clearly was. Butler-Sloss P recognised the force of the principles in Nielsen (and Family T v Austria, 64 DR 176, which followed it). Nevertheless, “There is a point, however, at which one has to stand back and say: is this within ordinary acceptable parental restrictions upon the movements of a child or does it require justification?” (para 28) Judge LJ was to the same effect: “In short, although normal parental control over the movements of a child may be exercised by the local authority over a child in its care, the implementation of a secure accommodation order does not represent normal parental control” (para 102). (They went on to hold that section 25 was not incompatible with article 5 as it fell within article 5(1)(d).)

That, as it seems to me, is the crux of the matter. Do the restrictions fall within normal parental control for a child of this age or do they not? If they do, they will not fall within the scope of article 5; but if they go beyond the normal parental control, article 5 will apply (subject to the question of whether parental consent negates limb (b) of the Storck criteria, see para 42 below). Quite clearly, the degree of supervision and control to which D was subject while in Placement B and Placement C was not normal for a child of 16 or 17 years old. It would have amounted to a deprivation of liberty in the case of a child of that age who did not lack capacity. The question then arises what difference, if any, does D’s mental disability make?

The answer to that question lies in the illuminating discussion by Lord Kerr in Cheshire West:

Indeed, the principal point of Cheshire West was that the living arrangements of the mentally disabled people concerned had to be compared with those of people who did not have the disabilities which they had. They were entitled to the same human rights, including the right to liberty, as any other human being. The fact that the arrangements might be made in their best interests, for the most benign of motives, did not mean that they were not deprived of their liberty. They were entitled to the protection of article 5, precisely so that it could be independently ascertained whether the arrangements were indeed in their best interests.

It follows that a mentally disabled child who is subject to a level of control beyond that which is normal for a child of his age has been confined within the meaning of article 5. Limb (a) of the three Storck criteria for a deprivation of liberty (see para 1 above) has been met. There was, however, an argument that the consent of D’s parents supplied a substitute for the consent of the person confined, so that limb (b) was not met. It suited counsel in Cheshire West (as recorded in the last sentence of para 41) to argue that Nielsen should be regarded as a case of substituted consent, because no person has the right to give such consent on behalf of a mentally incapacitated adult. But, as also pointed out in Cheshire West, it is striking that the European Court of Human Rights has consistently held that limb (b) can be satisfied despite the consent of a person with the legal right to make decisions on behalf of the person concerned: see Stanev v Bulgaria (2012) 55 EHRR 696, DDv Lithuania [2012] MHLR 209, Kedzior v Poland [2013] MHLR 115, Mihailovs v Latvia, unreported, and now Stankov v Bulgaria [2015] 42 ECtHR 276. In Stanev, the court did observe, in passing, that “there are situations where the wishes of a person with impaired mental facilities may be validly replaced by those of another person acting in the context of a protective measure and that it is sometimes difficult to ascertain the true wishes or preferences of the person concerned” (para 130). However, as Keehan J observed in the Court of Protection (para 118) that is very far from adopting a general principle of substituted consent. The consent of a legal guardian may have been sufficient to make the confinement lawful in the domestic law of the country concerned, but that did not prevent its being a deprivation of liberty, or guarantee that it fulfilled the Convention requirement of legality. In the cases where limb (b) has been held to be satisfied, it is because the evidence showed that the person concerned was willing to stay where he or she was and was capable of expressing that view. Parental consent, therefore, cannot substitute for the subjective element in limb (b) of Storck.

As already mentioned, limb (c) of Storck is no longer disputed and rightly so. Not only was the State actively involved in making and funding the arrangements, it had assumed statutory responsibilities - albeit not parental responsibility - towards D by accommodating him under section 20 of the Children Act 1989, thereby making him a “looked after child”. Even without all this, it is clear that the first sentence of article 5 imposes a positive obligation on the State to protect a person from interferences with liberty carried out by private persons, at least if it knew or ought to have known of this: see, for example Storck, para 89.

In conclusion, therefore, the accommodation of D in Placement B and Placement C did amount to a deprivation of liberty within the meaning of article 5 and the fact that his parents agreed to them did not rob the arrangements of this quality. The procedural requirements of article 5 applied. (As it happens, both placements were authorised by a High Court Judge sitting in the Court of Protection and it is common ground that they were in D’s best interests. His rights under article 5 have not, in fact, been violated.)

This conclusion is consistent with the whole thrust of Convention jurisprudence on article 5, which was examined in great detail in Cheshire West. But it is reinforced by the consideration that it is also consistent with the principle of non-discrimination in article 2.1 of the United Nations Convention on the Rights of the Child, which requires that the rights set out in the Convention be accorded without discrimination on the ground of, inter alia, disability, read together with article 37(b), which requires that no child shall be deprived of his liberty unlawfully or arbitrarily, and article 37(d), which requires the right to challenge its legality. It is also consistent with article 7.1 of the United Nations Convention on the Rights of Persons with Disabilities, which requires all necessary measures to ensure the full enjoyment by children with disabilities of all human rights and fundamental freedoms on an equal basis with other children.

But what is the relationship between holding that the placement did deprive D of his liberty within the meaning of article 5 and the view that it might otherwise have been within the scope of parental responsibility? Parental responsibility is about the relationship between parent and child and between parents and third parties: it is essentially a private law relationship, although a public authority may also hold parental responsibility. As Irwin LJ correctly pointed out (para 157) human rights, on the other hand, are about the relationship between individuals (or other private persons) and the state. It is, however, now agreed that any deprivation of liberty in Placement B or Placement C was attributable to the state. So is there any scope for the operation of parental responsibility to authorise what would otherwise be a deprivation of liberty?

There are two contexts in which a parent might attempt to use parental responsibility in this way. One is where the parent is the detainer or uses some other private person to detain the child. However, in both Nielsen and Storck it was recognised that the state has a positive obligation to protect individuals from being deprived of their liberty by private persons, which would be engaged in such circumstances.

The other context is that a parent might seek to authorise the state to do the detaining. But it would be a startling proposition that it lies within the scope of parental responsibility for a parent to license the state to violate the most fundamental human rights of a child: a parent could not, for example, authorise the state to inflict what would otherwise be torture or inhuman or degrading treatment or punishment upon his child. Likewise, section 25 of the Children Act 1989 recognises that a parent cannot authorise the State to deprive a child of his liberty by placing him in secure accommodation. While this proposition may not hold good for all the Convention rights, in particular the qualified rights which may be restricted in certain circumstances, it must hold good for the most fundamental rights - to life, to be free from torture or ill-treatment, and to liberty. In any event, the state could not do that which it is under a positive obligation to prevent others from doing.

In conclusion, therefore, it was not within the scope of parental responsibility for D’s parents to consent to a placement which deprived him of his liberty. Although there is no doubt that they, and indeed everyone else involved, had D’s best interests at heart, we cannot ignore the possibility, nay even the probability, that this will not always be the case. That is why there are safeguards required by article 5. Without such safeguards, there is no way of ensuring that those with parental responsibility exercise it in the best interests of the child, as the Secretaries of State acknowledge that they must. In this case, D enjoyed the safeguard of the proceedings in the Court of Protection. In future, the deprivation of liberty safeguards contained in the Mental Capacity Act 2005 (as amended by the Mental Capacity (Amendment) Act 2019) will apply to children of 16 and 17. I would therefore allow this appeal and invite the parties’ submissions on how best to incorporate this conclusion in a declaration.

Logically, this conclusion would also apply to a younger child whose liberty was restricted to an extent which was not normal for a child of his age, but that question does not arise in this case. The common law may draw a sharp distinction, in relation to the deprivation of liberty, between those who have reached the age of 16 and those who have not, but the extent to which that affects the analysis under the Human Rights Act is not clear to me and we have heard no argument upon it. I therefore prefer to express no view upon the question. Nor would I express any view on the extent of parental responsibility in relation to other matters, such as serious and irreversible medical treatment, which do not entail a deprivation of liberty. Some reference to this was made in the course of argument, but it does not arise in this case, which is solely concerned with depriving 16 and 17-year-olds of their liberty. It follows that I agree with what Lady Black says about those last two points in para 90 of her judgment.

The purpose of this judgment is two-fold. It addresses the question of whether the restrictions, in section 25 of the Children Act 1989, on placing children in accommodation provided for the purpose of restricting liberty apply to the sort of living arrangements in question here. It also provides an opportunity to explain a little further why I agree with Lady Hale’s conclusion that the consent of D’s parents to his confinement cannot operate as a substitute for D’s own consent. Lady Hale bases this conclusion, essentially, on there being no room for substituted consent in cases such as the present, for reasons she sets out commencing at para 42 of her judgment; I agree with her on this point and I do not seek to detract from what she says there. My comments are directed at the prior issue of whether it is actually within the scope of parental responsibility to consent to living arrangements for a 16 or 17-year old child which would otherwise amount to a deprivation of liberty within the meaning of article 5.

Like Sir James Munby P (at para 50 of the Court of Appeal judgment), I consider that, in order to answer this question, it is necessary to look to the domestic law, set in its proper historical context. Before us, the parties have not dwelt on the legal history in relation to parental responsibility. This is no doubt because the Official Solicitor accepts that the Court of Appeal was entitled to hold that, immediately following the decision in Gillick v West Norfolk & Wisbech Area Health Authority [1986] AC 112 (Gillick), parental responsibility was, in principle, exercisable to authorise the confinement of a 16 or 17-year old child who, for whatever reason, lacked capacity. The Official Solicitor’s argument is that that position has changed since Gillick with the passage of the Mental Capacity Act 2005 and the trends in international rights norms.

Notwithstanding the Official Solicitor’s approach, I have gone back to look in detail at the old authorities, many of them signposted in the President’s judgment. The President observed that the domestic law is far from straightforward, an observation which I have no difficulty in endorsing. At para 62 below, I have summarised an explanation given by Bowen LJ in In re Agar-Ellis (No 2)(1883) 24 Ch D 317 of the terminology used in this area of the law, which might help a little in understanding the earlier authorities.

Lady Hale sets out, in para 21, the themes which are to be found in the old cases. I do not disagree with what she says, but merely seek to add a little more detail to the picture and to explain that, as will become apparent (particularly from paras 88 to 90), I have reached a firmer conclusion than she has on this aspect of the case.

It might be useful to set the scene by citing the passage from Sachs LJ’s judgment in Hewer v Bryant [1970] 1 QB 357, at 372-373, which, after his own study of the earlier authorities, the President found (para 65) to be an accurate analysis of the position:

It is, of course, custody in what Sachs LJ called “its limited meaning” that is material in the present appeal. In this sense, it is concerned with, as he put it, “an ability to restrict the liberty of the person concerned”, otherwise described as a “power of physical control over an infant”, and even “physical possession”. There are a number of earlier cases, notably Rex v Greenhill (1836) 4 Ad & E 624, R v Maria Clarke (In the Matter of Alicia Race) (1857) 7 E & B 186, and R v Howes (1860) 3 El & El 332, which deal with the common law position in relation to this aspect of custody. The context in each case is a habeas corpus application by a parent. What is important about the decisions for present purposes is that they establish, as the common law position, that (i) up to the age of discretion, the parent’s right to restrict the child’s liberty was absolute (subject to some very limited exceptions), (ii) once the child reached the age of discretion, that right disappeared, and (iii) reaching the age of discretion was a matter of attaining the requisite chronological age, and not a matter of mental capacity.

It is necessary to begin with Rex v Greenhill, although it is only in the subsequent cases that its import becomes clear. It concerned children who were all under six years of age and were with their mother. Their father obtained an order for them to be delivered up to him and their mother applied for that to be set aside. She was unsuccessful, a father being entitled to custody in the absence of any sufficient reason to separate the children from him. The four judges each gave separate short judgments, from which the following are extracts:

Rex v Greenhill was relied upon in R v Maria Clarke (In the matter of Alicia Race) (1857) where Lord Campbell CJ interpreted it as having:

The issue in the Maria Clarke case was whether the ten year old girl’s widowed mother, as her guardian for nurture, had a legal right to custody against the wishes of the girl, however intelligent she was, or whether the court was bound to examine the child to ascertain whether she had the mental capacity to make a choice. There was no argument but that children under seven were delivered to the guardian without any such examination; the argument was about those between the ages of seven and 14 (when guardianship for nurture ended). The court held that the guardian was absolutely entitled to the custody of the child until the age of 14, irrespective of the child’s capacity. Lord Campbell said:

In R v Howes (1860) 3 El & El 332, a father sought the delivery to him of a girl of 15 who was unwilling to return to him, and had been brought into court in obedience to a writ of habeas corpus and interviewed privately by the judges before they heard argument. Cockburn CJ, giving the judgment of the court, said:

In re Agar-Ellis(No 2) (1883) 24 Ch D 317 concerned a girl of 16 who wanted to spend time with her mother, contrary to her father’s directions. She and her mother argued that, given her age, her father had no right to the control or custody of her. Their petition failed at first instance and they appealed. This was not a habeas corpus application because, said Sir William Brett MR, “the child is not away from her father - the child is under the control of her father”. It was, as Lindley LJ described it the following year in Thomasset v Thomasset (see below), “an attempt … to remove a girl over 16 from the care of her father”. Therefore, although the habeas corpus cases were considered, they were found to be inapplicable. The court was concerned, rather, with an exercise of what Cotton LJ described as “the jurisdiction which the Court of Chancery has always exercised, delegated probably from the Crown as parens patriae”. The law was declared to be that the father had, as the Master of the Rolls put it, “control over the person, education, and conduct of his children until they are 21 years of age”. Exercising its Chancery jurisdiction, the court proceeded on the basis that it could interfere with the discretion of the father, but would not do so except in very extreme cases, of which this was not one. It is to be noted that this might have been seen by some as an overly narrow application of the Chancery approach, see the judgment of Kay LJ in R v Gyngall [1893] 2 QB 232. Sachs LJ commented, in the passage I have quoted above from Hewer v Bryant, on the use during the second half of the 19th century of the parens patriae jurisdiction “unquestionably … in aid of the wishes of the father”, citing Agar-Ellis (No 2) as an example of this, but pointing out that any powers exercised by way of physical control in the later years of infancy were not in fact the father’s personal powers but the more extensive ones of the Crown. And in Gillick, Lord Scarman referred to In re Agar-Ellis (No 2) and an earlier decision concerning the same family (In re Agar-Ellis(1878) 10 Ch D 49), as “the horrendous Agar-Ellis decisions” (p 183E of Gillick).

Bowen LJ’s judgment in In re Agar-Ellis (No 2) includes an interesting explanation of how confusion had been caused by earlier law books making distinctions which were no longer adhered to. He explained that the strict common law gave the father guardianship of his children during the age of nurture and until the age of discretion (14 for a boy, and 16 for a girl), and thereafter, apart from in the case of the heir apparent (in relation to whom he was “guardian by nature” until 21), he had no guardianship. But he said that, “for a great number of years”, more especially in the courts of equity, the term “guardian by nature” had not been confined to heirs apparent, so that there was “a natural paternal jurisdiction between the age of discretion and the age of 21, which the law will recognise”.

The distinction between the common law jurisdiction (concerned to declare rights between the parties) and the broader jurisdiction of the Courts of Chancery is made very clear in R v Gyngall (supra). The father of the child in that case (a girl of about 15) was dead and it was the mother who was the guardian, it seems by operation of the Guardianship of Infants Act 1886. The decision of the first instance court not to return the girl to her mother, despite there being no misconduct on the part of the mother derogating from her right to custody, was interpreted as an exercise of the Chancery jurisdiction, taking into account the welfare of the child, rather than an exercise of the common law habeas corpus jurisdiction.

It is worth looking at the case of Thomasset v Thomasset, the following year, because the Court of Appeal there set out its understanding of the position that the law had now reached. Dealing first with the approach of the Courts of Common Law, Lindley LJ said (at p 298) that the father had a legal right to custody of his child until the child attained 21, but the child would not be forced to remain with the father after he or she had attained the age of discretion. He quoted the passage from Coleridge J in Rex v Greenhill which is set out above, and continued:

Although the Court of Chancery would be in the same position as the Courts of Common Law when dealing with a writ of habeas corpus, Lindley LJ emphasised that it also had its much more extensive parens patriae jurisdiction. This had been exercised in aid of fathers and guardians of children who had attained the age of discretion, and also to control the rights of fathers and guardians in order to secure the welfare of infants, and it was available to the Divorce Court since the Judicature Acts. However, whilst the Divorce Court could make orders respecting the custody, maintenance, and education of infants during the whole period of infancy, that is up to 21, both members of the court expressed caution on the subject. Lindley LJ said (p 303):

Lopes LJ said (p 306):

It is not easy for us, accustomed to a legal system which has changed very considerably since Victorian times, to reach a perfect understanding of these cases, and I do not pretend to have done so. Like Sir James Munby P, I would settle for Sachs LJ’s summary, quoted at para 55 above. How does this historical perspective, as summarised by Sachs LJ, inform the present issue?

It can be seen from the old cases that the mere fact that a parent had guardianship of a child did not give him a free hand to make decisions about that child’s life. The extent of his ability to dictate depended upon the circumstances. At common law, the ability to restrict the liberty of the child lasted only until the age of discretion, and the age of discretion was fixed chronologically, and not by considering the attributes of the particular child. Equity had wider powers to govern a child’s behaviour, but this was essentially by stepping in as the parent, and making decisions for the welfare of the child, rather than by enforcing the parent’s rights. Terminology has changed, of course, and what was referred to as guardianship in those days has been translated into today’s parental responsibility, but I do not see this older chapter in the evolution of the scope of parental authority as irrelevant in the search for “the rights, duties, powers, responsibilities and authority which by law a parent of a child has in relation to the child and his property” (section 3(1) of the Children Act 1989). Equally, it is not the end of the investigation, because the common law can evolve. I turn therefore to the important case of Gillick decided in 1985. The President took the view (see in particular paras 85 and 125 of his judgment) that the fixed “age of discretion”, encountered in the old cases, has been replaced by a test of, as he termed it, “Gillick capacity” to determine when the exercise of parental responsibility comes to an end.

I will look in more detail at Gillick below, but it may help to preface this examination by explaining that I do not share the President’s confidence that the Gillick test extends to the aspect of parental responsibility with which the present case is concerned. The issue in Gillick was different in at least two important ways. First, it was not about restricting the physical liberty of the child, but concerned decision making in the sphere of medical treatment. Secondly, the question was whether the parent could lose his or her exclusive decision-making powers before the child reached the age of 16, if the child was capable of making his or her own decision, not whether the parent was entitled to continue to make decisions after the child reached 16, if the child was not capable.

As to the first of these differences, the considerations in relation to decisions about physical liberty are not the same as those involved in other spheres where parental responsibility may operate. In particular, article 5 of the European Convention on Human Rights was not material in the Gillick decision, but is of central importance to the present case, as can be seen from Lady Hale’s judgment. Moreover, it is not only in the Convention, and the cases decided by the European Court of Human Rights, that one can find special attention being given to liberty, whether of a child or a vulnerable adult. It is evident in the older habeas corpus authorities to which I have referred. And it has been firmly engrained in domestic law in certain statutes, notably in the Children Act 1989 in the secure accommodation provisions, with which I deal later in this judgment, and in the Mental Capacity Act 2005 which, by section 4A, marks out “deprivation of liberty” for special treatment, with more attendant safeguards than other acts performed in relation to a person who lacks capacity.

Turning to the second of the differences, it has a number of components, one of which is the pivotal age of 16. Although not determinative of the question before us, I think it is far from irrelevant that, as Lady Hale puts it at para 27 (after listing various statutory provisions in her para 26), the law accords to children who have reached the age of 16 years a status which is in some respects different from children below that age. Of the provisions listed, I would single out section 2(5) of the Mental Capacity Act 2005 (para 26(iii) of Lady Hale’s judgment). I cannot accept the Official Solicitor’s case that the 2005 Act constitutes a complete decision-making framework for the care and treatment of those aged 16 and above who lack capacity, not least because there is an obvious overlap between the reach of the Children Act 1989 and that of the 2005 Act, and I can find nothing in the 2005 Act that could be said to indicate a general rule to the effect that, where it applies, it does so to the exclusion of other common law and statutory provisions. However, it does seems to me that the deliberate choice of the legislature to include children of 16 to 18 years within the scope of the 2005 Act, and now (by virtue of the recent amendment to the Act, see para 49 of Lady Hale’s judgment) to extend a regime of administrative deprivation of liberty safeguards to them, indicates an appreciation of the different needs of this particular age group.

At risk of underlining the obvious, another important element in this second difference is that Gillick was about contracting the boundary of parental responsibility and empowering the child at an earlier age, whereas the present case is about extending the boundary of parental responsibility beyond the age at which, in relation to the particular matter in issue, it would have been taken, at common law, to have ended. As can be seen from SachsLJ’s analysis, the common law position at the time of Hewer v Bryant, in 1969, was that the power of physical control over an infant ended at the age of discretion. For present purposes I would take that age as 16. The common law, even in 1969, might have balked at continuing to treat a boy as reaching the age of discretion at 14, but a girl as having to wait until 16, and, if an issue about this had come up, would no doubt have evolved to iron out the difference. It is pointless to worry, in the present context, about whether it would have equalised up or down, because we are dealing here with the position of a child between the ages of 16 and 18. If the age of discretion had been raised above 16, thus extending the parent’s power of physical control into the age group with which we are concerned, that would have been of considerable significance to the debate, but I have seen nothing to suggest that that was done, and it would be extremely surprising if it had been.

So to the detail of Gillick. Lady Hale deals with the case in para 23 of her judgment. I propose to refer to some further passages, in order to provide some substance for the views that I have just expressed about it. The Family Law Reform Act 1969 had provided that the consent of a minor who has attained the age of 16 years to medical treatment was as effective as if the child were of full age, and rendered it unnecessary to obtain consent from the parent or guardian. It was the position of a child of under 16 that was in question, and in particular whether such a child could, herself, provide the necessary consent for contraceptive treatment. It will be recalled that Mrs Gillick was arguing (see p 168D of the report) that the custody that parents have of a minor under the age of 16 “necessarily involves the right to veto contraceptive advice or treatment being given to the girl”, and that she failed in this argument.

Giving the first speech, Lord Fraser of Tullybelton, one of the three members of the House in the majority, spoke in terms of “parental rights to control a child”, which he held (p 170D) existed for the benefit of the child, not the parent, and were justified only in so far as they enabled the parent to perform his duties towards the child, and towards other children in the family. Understandably, given the issue in the case, his speech is directed at diminishing control on the part of a parent as a child ages, rather than at the opposite problem of the child who needs parental input for a longer than usual period, although I accept that some of what he said is in general terms and could be applied to either situation. He spoke (p 171E) of wise parents relaxing their control gradually as the child develops, and of the degree of parental control actually exercised over a particular child varying according to his understanding and intelligence. He looked at R v Howes, In re Agar-Ellis, and Hewer v Bryant (see above). As to R v Howes, where the court declined to consider a child having discretion to consent to leaving the father before she reached 16, he said that the view that the child’s intellectual ability is irrelevant cannot now be accepted. He endorsed the criticism that had been heaped on Agar-Ellis, dismissed the concept deployed in that case of absolute paternal authority continuing until a child attains majority as so out of line with present day views that it should no longer be treated as having any authority, and shared Lord Denning’s view that the legal right of a parent to custody of a child was “a dwindling right”, as the child approached majority. He said (p 173D):

This abandonment of the rule of the parents’ absolute authority led to the conclusion that a doctor could prescribe contraceptive treatment to a girl of under 16 without the consent of her parents, provided that, amongst other things, she would understand his advice.

Lord Scarman agreed with Lord Fraser but added his own speech. Having considered the earlier case law, he enunciated what he had found to be the principle of law (hereafter in my discussion of his speech “the principle”), saying (pp 183-184):

Later in his speech, Lord Scarman formulated the principle in slightly different ways. At p 185E, the following formulation can be found:

Understandably, in his consideration of “the common law’s understanding of the nature of parental right”, Lord Scarman’s focus was upon the particular type of parental right/duty that was in issue there, as the following passage from his speech makes clear (p 184F):

Although varying ages of discretion had been fixed by statute and case law for various purposes, Lord Scarman found it clear that (p 185F):

In modern times, statute had intervened in respect of a child’s capacity to consent to medical treatment from the age of 16 onwards, but neither statute nor case law had ruled on the extent and duration of parental right in respect of children under the age of 16. So, Lord Scarman said, it was “open … to the House to formulate a rule” (p 185H).

He was influenced, in so doing, by the fact that the law relating to parent and child is concerned with the problems of growth and maturity of the human personality. This disposed him against the fixed age limit of 16 (below which a girl could not give valid consent) that had commended itself to the Court of Appeal. He observed (p 186B):

He found the earlier cases no guide to the application of the principle in the conditions of today. He dealt specifically with the habeas corpus cases in these terms (p 187B):

After citing from R v D[1984] AC 778 (an appeal relating to the conviction of a father on indictment of kidnapping his five year old daughter) on the subject of parental right and a child’s capacity to give or withhold a valid consent, Lord Scarman concluded (p 188H) that:

The third member of the majority, Lord Bridge, agreed with what Lord Fraser and Lord Scarman had said, without adding further reasoning of his own.

What the President drew from the speeches of Lord Fraser and Lord Scarman was, as he set out at paras 83 to 85, that the attainment of “Gillick capacity” is “child-specific”, to be decided as a matter of fact in relation to each particular child. He said:

Therefore, he said (para 85), after Gillick, the position of domestic law in relation to the aspect of custody described in Hewer v Bryant as, inter alia, “the ability to restrict the liberty of the person” was that:

So, the President concluded, Keehan J was wrong to decide that a parent of an incapacitous 16-year-old may not consent to confinement which would otherwise amount to a deprivation of liberty (para 115 of Keehan J’s judgment) because (para 125 of the President’s judgment), none of the statutory provisions upon which he relied assisted on the matter and:

As I have explained (see above at para 69 et seq), I do not share the President’s confidence that the Gillick test extends to the aspect of parental responsibility with which the present case is concerned, or that the Gillick decision can, without more, be treated as regulating the situation where the objective is not to contract the boundaries of parental responsibility, but to extend them. In my view, as I said above, it is of real significance that in Gillick, the House of Lords were dealing with a materially different issue. The respondent recognises that the focus of Gillick was specific to the issue of consent to medical treatment of children under 16, but invites this court to conclude that the test laid down there applies beyond that scope and up to the age of majority. I accept that certain things that were said in Gillick were capable of being interpreted as applying to a situation such as the present, but it would not, in my view, be appropriate to interpret them in that way, so as to draw into the Gillick net a situation which is diametrically opposed to that with which the House was concerned (not the tempering of parental responsibility in relation to the under 16 age group, but its expansion in relation to those aged 16 and 17 so as to give it a role which would not otherwise be afforded by the common law). My unwillingness to adopt this interpretation is reinforced by what I perceive to be the distinct, and rather special, features of the field of deprivation of liberty with which we are here concerned. It follows that the rights of a parent in relation to restricting the liberty of a child remain, at common law, as described in Hewer v Bryant. The inescapable result of that is, I think, that it is not within the scope of parental responsibility for parents to give authority for their 16 year old child to be confined in a way which would, absent consent, amount to a deprivation of liberty. In so saying, I do not intend in any way to water down the important changes brought about by Gillick or to alter the way in which it has been applied in many spheres of family law. I have only been concerned to consider its application in the very specific context of confinement of children of the ages of 16 to 18.

The position in relation to the confinement of children who are under 16 might be different for a variety of reasons. It could be argued, for example, that the Gillick decision is more readily applicable to under 16s than to over 16s, given that this was the age group with which the House was concerned. It would then be arguable that the position in relation to that group was as the President set out at para 85 of his judgment (quoted above) ie that the parental ability to restrict a child’s liberty continues to be as described by Sachs LJ in Hewer v Bryant, but with a Gillick test rather than the previous fixed ages. But the effect of this, applied to a child who lacked capacity, would not be to leave a gap in the parent’s powers to cater for the particular needs of a child with disability. On the contrary, the child not having attained Gillick capacity, there would be nothing to bring to an end the parent’s common law power to confine the child as required in the child’s interests. To put it in the terms used in this appeal, it would remain within the ambit or zone of the parent’s parental responsibility. However, there would, no doubt, be other arguments to be aired on the point, and I have not formed even a preliminary view about it.

In summary, therefore, I would hold that as a matter of common law, parental responsibility for a child of 16 or 17 years of age does not extend to authorising the confinement of a child in circumstances which would otherwise amount to a deprivation of liberty. For me, this reinforces the conclusion to which Lady Hale has come by the route she sets out in paras 42 to 49 of her judgment. She concludes, in para 50, by saying that logically her conclusion would also apply for a younger child, but I would prefer to leave this separate question entirely open, to be decided in a case where it arises. I should also stress, before moving on to the discrete issue in relation to section 25 of the Children Act 1989 and its potential application to living arrangements such as D’s, that I have been looking specifically at the common law power of a parent in relation to a child’s liberty. I have not intended to cast doubt on any existing understanding about the operation of parental responsibility in different spheres of a child’s life. And nothing that I have said is intended to cast any doubt on the powers of the courts, recognised in the early cases to which I have referred, and still available today in both the parens patriae jurisdiction and under statute, notably the Children Act 1989, to make orders in the best interests of children up to the age of majority, with due regard to their wishes and those of their parents, but not dictated by them.

Where it applies, section 25 of the Children Act 1989 regulates the circumstances in which children can be placed and kept “in accommodation … provided for the purpose of restricting liberty” and dictates that, save for very short periods, the court’s authorisation of the arrangements is required. If the section applies to living arrangements like D’s, making court authorisation obligatory, the debate as to whether it falls within the scope of parental responsibility to authorise a child’s confinement would be of far less practical significance. In order to set that debate in its proper context, the scope of section 25 was therefore explored. In the light of this exploration, it appears likely that a significant number of children living in confined circumstances will be outside the ambit of the section, although clearly each case will depend upon its own facts. Accordingly, the parental responsibility issue has a real practical importance. The reasons for this provisional conclusion are set out below. They deal with the law as it applies to accommodation in England; there is a separate statutory and regulatory regime where the accommodation is in Wales, albeit in similar terms.

Section 25 provides as follows, omitting provisions concerned solely with Scotland:

Where applicable, the section operates to prevent a child being placed or kept in secure accommodation unless one of the two conditions set out in section 25(1)(a) and (b) is satisfied. The initial placement need not involve the court, but regulations made under section 25(2) provide that a child may not be kept in secure accommodation without court authority for more than 72 hours in any period of 28 consecutive days (regulation 10, Children (Secure Accommodation) Regulations 1991 (SI 1991/1505), hereafter “the 1991 Regulations”). There are limits on the period that can be authorised by the court, being three months in the first instance, and a further period of up to six months thereafter (regulations 11 and 12).

There is a misconception that section 25 applies only to children who are being looked after by a local authority. These are the children to whom section 25(1) refers, but section 25(7) gives the Secretary of State power to provide, by regulations, that the section shall or shall not apply to other descriptions of children, and he did so in the 1991 Regulations. Various categories of children are excluded from the operation of the section including, by regulation 5(1), a child who is detained under the provisions of the Mental Health Act 1983, and, by regulation 5(2)(a), a child who is being accommodated under section 20(5) of the Children Act 1989 (which relates to certain accommodation in a community home of people who are over 16 but under 21 years of age). In contrast, regulation 7 widens the reach of section 25, extending it to children other than those looked after by a local authority. It provides:

With regulation 7 casting the section 25 net beyond looked after children, the possibility that a child is in secure accommodation cannot be dismissed simply on the basis that the child is not being looked after by the local authority. Furthermore, the inclusion within section 25 of children in hospitals and care homes demonstrates that the traditional view that secure accommodation has a punitive quality (see for example In re AB (A Child) (Deprivation of Liberty: Consent) [2015] EWHC 3125 (Fam); [2016] 1 WLR 1160, para 31) will not always be valid, so that cannot be used as a reliable hallmark of secure accommodation either.

Deprived of obvious insignia such as these, how is it to be determined whether a particular child’s circumstances are covered by section 25? As established by section 25(1), the concern of the section is a child who is “placed, and if placed, … kept in accommodation … provided for the purpose of restricting liberty (‘secure accommodation’)”. This definition, which is mirrored in regulation 2(1) of the 1991 Regulations, is the only definition of “secure accommodation”, so the hallmark by which such accommodation has to be identified is that it is “accommodation … provided for the purpose of restricting liberty”.

The Secretaries of State argue that identification is simplified in the case of children’s homes because, they say, “Parliament has provided a mechanism for determining which children’s homes have the nature of being secure accommodation”. The mechanism suggested derives from regulation 3 of the 1991 Regulations. This provides that “[a]ccommodation in a children’s home shall not be used as secure accommodation unless … it has been approved by the Secretary of State for that use”. It seems that the Secretaries of State argue that where the accommodation in question is in a children’s home, it will count as “secure accommodation” only if it has been approved by the Secretary of State for that use. The logical corollary of that would appear to be that no matter what the living arrangements of a child in an unapproved children’s home, he or she is not placed/kept “in accommodation … provided for the purpose of restricting liberty” and therefore not within section 25.

This argument might owe something to the regime in relation to local authority homes which was discussed in R v Secretary of State for the Home Department, Ex p A[2000] 2 AC 276 (see later). It is not necessary to determine, in the present case, whether it is correct in the different context of section 25, but it should be acknowledged that it does give rise to some questions. Whilst it can readily be accepted that the intention is that only properly authorised children’s homes are to be used as accommodation for the purpose of restricting liberty, it does not necessarily follow that, in practice, a child could not find him or herself placed or kept in a children’s home which, but for the fact that it does not have the Secretary of State’s approval, has every appearance of being secure accommodation. If the argument advanced by the Secretaries of State is right, such children might be doubly prejudiced ie placed in an unapproved children’s home and outside the protective regime of section 25. Given the shortage of approved secure children’s homes, highlighted by the Court of Appeal in In re T (A Child) [2018] EWCA Civ 2136, this is a risk which cannot be ignored. In In re T, the appellant was 15 years old and subject to a full care order. The local authority proposed that she be detained in a unit which was not an approved children’s home, and sought authority from the High Court for the restriction of the child’s liberty, relying upon the inherent jurisdiction. It is evident from the judgment of the President of the Family Division (with whom the other members of the court agreed) that such applications are not uncommon. At para 5, he said that there are “many applications being made to place children in secure accommodation outside the statutory scheme laid down by Parliament”, expressing concern about the situation (see also paras 88 and 89). No question seems to have been raised as to whether it is proper for the High Court’s inherent jurisdiction to be used to authorise the restriction of a child’s liberty in an unapproved children’s home. This might, perhaps, have been because the child’s accommodation was not in fact in a children’s home as defined for the 1991 Regulations (see regulation 2) and therefore not covered by the prohibition in regulation 3, but given the limited details available about the child’s circumstances, it is impossible to know.

In any event, even if the approach commended by the Secretaries of State is correct, it would not serve to identify “secure accommodation” in all its various settings, but only in so far as children’s homes are concerned, and it would leave unanswered questions in relation to many other children. Accordingly, there being no reliable and universally applicable shortcuts to identifying secure accommodation, it is necessary to look more closely at the wording of section 25(1) in order to determine what circumstances fall within it.

The parties rightly stress the need to interpret the section with an eye to the whole scheme in which it takes its place. Local authorities have far-reaching welfare obligations towards children. Notably, under section 20 of the Children Act 1989, they have a duty to provide accommodation for children in need, and they must also address the accommodation and other needs of children in relation to whom care orders have been made. The children who require help will present with all sorts of different problems, and there will be those whose care needs cannot be met unless their liberty is restricted in some way. But by no means all of these children will fall within the criteria set out in section 25(1)(a) and (b), which are the gateway to the authorisation of secure accommodation. It seems unlikely that the legislation was intended to operate in such a way as to prevent a local authority from providing such a child with the care that he or she needs, but an unduly wide interpretation of “secure accommodation” would potentially have this effect. It is possible to imagine a child who has no history, so far, of absconding, and who is not likely actually to injure himself or anyone else, so does not satisfy section 25(1)(a) or (b), but who, for other good reasons to do with his own welfare, needs to be kept in confined circumstances. If section 25 applies whenever a child’s liberty is restricted, local authorities will not be able to meet the welfare needs of children such as this. And, of course, it is similarly possible to envisage children in hospitals and care homes who need a degree of confinement, but do not satisfy either of the limbs of section 25(1). Putting it another way, the criteria set for the placing or keeping of a child in secure accommodation might be taken to reveal something of the problems which it was anticipated that children in secure accommodation would present. This, in turn, could be taken as a pointer towards the characteristics that one could expect to find in secure accommodation being used to meet those problems.

It is also worth noting, when considering how section 25 fits into the statutory scheme, that a court determining an application under the section does not have the child’s welfare as its paramount consideration, as would normally be the case when the court determines any question with respect to the upbringing of a child (section 1 of the Children Act 1989). If any of the relevant criteria for keeping a child in secure accommodation are satisfied, the court is obliged to make the order authorising the child to be kept in secure accommodation (section 25(4)). It would be surprising if section 25 were intended to be interpreted in such a way as to extend this displacement of the court’s welfare role beyond a relatively circumscribed group of children whose circumstances make this unavoidable. Underlining this, it is worth noting that where the position of a child of 16 or 17 is being considered in the Court of Protection under the Mental Capacity Act 2005, welfare is the touchstone, as deprivation of liberty will only be endorsed where it is in the best interests of the child.

So, the challenge is to interpret section 25 in such a way as to provide the protection intended by the legislature, without getting in the way of meeting the varied needs of the children for whom hospitals, care homes, and local authorities (in the exercise of their social services and education functions) have responsibility. We are grateful to the parties for the valuable detailed written submissions they have all made to assist with this process; as, for the most part, they traverse similar ground, it is unnecessary to attribute submissions in what follows. It is unnecessary also to address all the arguments advanced, given that we are not making a definitive decision as to the operation of section 25.

It is submitted that the focus should be on the accommodation and the purpose for which it is provided, rather than upon the regime within the accommodation. This would be consistent with section 25(1)(a) and (b) which, in setting the criteria for the use of secure accommodation, stipulate that the child may not be placed/kept in secure accommodation unless it appears that he is likely to abscond from “any other description of accommodation” or to injure himself or others if he is kept in “any other description of accommodation”. This contrast of “secure accommodation” with “any other description of accommodation” can be read as supporting the notion that “secure accommodation” is a “description of accommodation”, rather than a description of a regime of care.

This is an interpretation which also gains support from R v Secretary of State for the Home Department, Ex p A[2000] 2 AC 276. The 15 year old offender in that case had been remanded to a local authority children’s home which was not approved by the Secretary of State for the purpose of restricting liberty, but he was subject to a curfew and other conditions whilst there. The issue was whether he should be given credit, in serving his sentence of detention in a young offender institution, for his period in the local authority accommodation. That depended on whether it was covered by section 67(1A)(c) of the Criminal Justice Act 1967 which entitled an offender to have his sentence reduced by:

The legislative scheme with which the House was concerned was, of course, different from the provisions which concern us. Broadly speaking, by virtue of (inter alia) section 23 of the Children and Young Persons Act 1969, a court remanding a child or young person who had committed, or was alleged to have committed, a criminal offence could release him on bail or remand him to local authority accommodation, in either case with or without conditions. In the case of certain offenders who had reached 15 years of age, the court could “require [the designated local] authority to comply with a security requirement, that is to say, a requirement that the person in question be placed and kept in secure accommodation” (section 23(4)). Section 23(12) defined “secure accommodation” as “accommodation which is provided in a community home, a voluntary home or a registered children’s home for the purpose of restricting liberty, and is approved for that purpose by the Secretary of State”. It will be noted that this definition differs from that in section 25 of the Children Act, in that it makes approval by the Secretary of State an integral part of the definition of “secure accommodation”, whereas the section 25 definition makes no reference to such approval which is, instead, the subject of regulation 3 of the 1991 Regulations.

The local authority home in which the offender was accommodated on remand was not approved by the Secretary of State, so did not qualify as “secure accommodation” as such. But it was argued that section 67(1A)(c) was satisfied anyway, by virtue of the restrictions placed upon him whilst he was living there. A useful review of the history of the provisions as to secure accommodation in the civil and the criminal spheres can be found in the speech of Lord Clyde (with whom all the other members of the House agreed), commencing at p 285, although inevitably the law has moved on again since the decision. Then, at p 287, dealing with the construction of the phrase “and in accommodation provided for the purpose of restricting liberty” at the end of section 67(1A)(c), Lord Clyde said:

Similarly, Lord Hope (with whom the members of the Appellate Committee other than Lord Clyde agreed) said, at p 282, that:

Both Lord Clyde and Lord Hope were persuaded not only by the wording of the provision but also by practical considerations that this construction was correct. By focusing on whether the offender was in what Lord Clyde called “qualifying accommodation” (p 289E), the institution detaining the offender would be able to apply the appropriate credit against the sentence without having to form judgments about the precise conditions under which the individual offender had been held there (see Lord Hope at p 283 and Lord Clyde at p 289). Lord Clyde was clearly equating qualifying accommodation for section 67(1A)(c) purposes with secure accommodation as defined in section 23(12), as he envisaged that all that was necessary to ascertain whether the offender had been in qualifying accommodation was to see whether it had been approved by the Secretary of State as secure accommodation.

It would not be right to regard R v Secretary of State for the Home Department, Ex p A as determinative of the ambit of section 25 of the Children Act 1989. Although the phrase considered by the House of Lords there also features in section 25, the context is obviously different. There, by training the lens on the accommodation itself, the House was able to ensure that there was a simple means of identifying relevant periods on remand, merely by looking to see whether or not the particular local authority accommodation had the Secretary of State’s approval. Focusing on the accommodation itself does not, however, provide such a simple answer to the problem of what is secure accommodation within section 25. Section 25 extends well beyond local authority homes, and undoubtedly encompasses secure accommodation which does not have to be approved by the Secretary of State. Furthermore, the purpose of the provisions considered by the House of Lords was very different from the purpose of section 25. They were concerned with a scheme which conferred power on a court remanding a child to local authority accommodation to dictate that the child should be kept in secure accommodation as narrowly defined by section 23(4) of the Children and Young Persons Act, and confined credit for time spent in local authority accommodation to that type of accommodation. In contrast, what section 25 has to say about secure accommodation is of much wider application. It does not set out to dictate where a local authority must place/keep a particular child, but to regulate, in both local authority and non-local authority settings, the circumstances in which a child can be placed/kept in secure accommodation as defined in the section. Nevertheless, given that the House of Lords were concerned with the same phrase as features in section 25, their interpretation must carry weight.

Coming closer to home, we are invited to endorse the approach that Wall J took to the phrase “accommodation … provided for the purpose of restricting liberty” in In re C (Detention: Medical Treatment)[1997] 2 FLR 180 at p 193. The case concerned a 16-year old girl suffering from anorexia nervosa. The local authority made an application for an order under the inherent jurisdiction authorising her detention in a clinic for medical treatment. Wall J was faced with the question whether the court’s powers under the inherent jurisdiction were ousted by the scheme laid down by Parliament in section 25, and in addressing that issue, he needed to determine whether the clinic was, in fact, “secure accommodation” within section 25. Having reviewed three earlier authorities (R v Northampton Juvenile Court, Ex p London Borough of Hammersmith and Fulham [1985] FLR 193, South Glamorgan County Council v W and B[1993] 1 FLR 574, and A Metropolitan Borough Council v DB[1997] 1 FLR 767) he said (p 193):

Having said that, he went on to examine the regime operated by the particular clinic, before finding that it was not secure accommodation. In his view, the “purpose of placement of a child in the clinic is to achieve treatment: the accommodation provides a structure for that treatment”. The fact that there was a degree of restriction on the patient’s liberty was an incident of the treatment programme, and the fact that steps could be taken to prevent the child from leaving the premises did not, of itself, render the clinic secure accommodation.

Section 25 has played no direct role in the proceedings in the present case, and the bulk of the argument about it has occurred in writing after the conclusion of the hearing in this court. Nothing that we say about it will conclusively resolve the difficult questions that arise as to its scope and operation, and that is as it should be, because it would be undesirable that final views should be formed, without there having been an opportunity for oral argument. Furthermore, it would be better that such issues as there are about the scope of section 25 should be resolved in a case where the relevant facts have been found, so that the section can be interpreted with reference to a real factual situation. Because the issue was not under consideration at all before the appeal arrived in this court, factual findings have not been made in relation to all the matters relevant to the application of section 25 in D’s case. As, by virtue of his age, D is now no longer within the scope of the Children Act, there would be absolutely no point in remitting the case for evidence to be heard, particularly as none of the parties contends that this is a section 25 case.

The exercise in which we have engaged has, however, been sufficient to persuade us that section 25 is not intended to be widely interpreted, so as to catch all children whose care needs are being met in accommodation where there is a degree of restriction of their liberty, even amounting to a deprivation of liberty.There is much force in the argument that it is upon the accommodation itself that the spotlight should be turned, when determining whether particular accommodation is secure accommodation, rather than upon the attributes of the care of the child in question. This fits with the language used in section 25(1), when read as a whole. It is also consistent with the objective of ensuring that the section is not so widely drawn as to prejudice the local authority’s ability to offer children the care that they need, and it ought to make it more straightforward to apply than would be the case if the issue were dependent upon the features of a child’s individual care regime, so that the child might be found to be in secure accommodation in all manner of settings.

A restrained construction of the section is also justified by the fact that, far from being concerned with the routine sort of problems that might require a child’s freedom to be curtailed, the section has a “last resort” quality about it. It is concerned with accommodation which has the features necessary to safeguard a child with a history of absconding who is likely to abscond from any other description of accommodation or to prevent injury where the child in question would be likely to injure himself or others if kept in any other description of accommodation.

Of course, training the spotlight on the accommodation itself does not provide a complete answer to the question as to what falls within the definition of secure accommodation. Some secure accommodation will be readily recognisable from the fact that it is approved as such by the Secretary of State, but that is by no means a universal hallmark, as that approval is not needed for all types of secure accommodation. Moreover, given that it is contemplated that secure accommodation might be provided in places such as hospitals, it seems likely that there will not infrequently be more than one purpose of the child being in the accommodation, and there is much to commend Wall J’s approach to such a situation, that is to count within the definition of secure accommodation “designed for or having as its primary purpose” the restriction of liberty. Equally, the section will have to be interpreted in such a way as to allow for situations where only a part of the premises is made over to restricting liberty.

I agree with the judgment of Lady Hale on the effect of article 5 of the European Convention on Human Rights (the Convention). She has held that parental consent to a child’s living circumstances is not effective to prevent a child, who has mental disabilities and cannot give any relevant consent to those circumstances, from being deprived of their liberty for the purposes of article 5 if their living circumstances mean that they are not free and the restrictions on them go beyond those which are normal for a child of their age.

In this case, the child, D, is over 16 years of age. I agree with Lady Hale (para 50) that it is unnecessary in this case to express any view on the question whether there would be a deprivation of liberty for the purposes of article 5 if a child who has not yet attained that age has their liberty restricted to an extent that is not normal for a child of their age. Likewise I express no view on the question of parental consent for medical treatment or other matters outside article 5.

The key case on article 5 in this context is Nielsen v Denmark (1988) 11 EHRR 175, which Lady Hale analyses at paras 34 to 38 above. As Lady Hale explains, it is the normality of the parent’s control over the child, as compared with arrangements for children of a similar age, that is the key to understanding this difficult decision of the European Court of Human Rights (the Strasbourg court). In the present case, the position can simply be compared with the position of other children in the UK. It might in future be necessary to have regard to the practice in other contracting states to the Convention, but that does not arise in this case.

I have one qualification. Article 5 is not a qualified right and there is no scope for holding that the denial of a person’s liberty engages article 5 but does not amount to a violation because it serves a legitimate aim and is proportionate and necessary in a democratic society. Exceptionally there are situations where the Strasbourg court finds that in effect those tests were met but it can only do so by holding that there is no deprivation of liberty for article 5 purposes. Thus, in Austin v United Kingdom 35 EHRR 14, the complainants were demonstrators who had been “kettled” by the police, that is, kept against their will within a police cordon. The Strasbourg court held that there was no violation because the need for the police to maintain order in this situation meant the denial of liberty was not a deprivation of liberty for article 5 purposes. So, too, in Nielsen, the Strasbourg court had held that there was no deprivation of liberty for article 5 purposes.

It follows that there will be cases where a person loses their liberty but the acid test in Cheshire West, as Lady Hale describes it, does not apply. That conclusion is shown by observing that D’s case is about living arrangements. It is not about a child, or anyone else, needing life-saving emergency medical treatment. For the reasons which the Court of Appeal (McFarlane LJ, Sir Ross Cranston and myself) gave in R (Ferreira) v Inner South London Senior Coroner[2018] QB 487, the situation where a person is taken into (in that case) an intensive care unit for the purpose of life-saving treatment and is unable to give their consent to their consequent loss of liberty, does not result in a deprivation of liberty for article 5 purposes so long as the loss of liberty is due to the need to provide care for them on an urgent basis because of their serious medical condition, is necessary and unavoidable, and results from circumstances beyond the state’s control (para 89).

I pass on to section 25 of the Children Act 1989 and to the judgment of Lady Black. So far as section 25 is concerned, this was unfortunately dealt with only on written submissions. I have read the judgment of Lady Black, in which Lady Hale concurs, with admiration. I have read it as laying down a marker for the future. I have read it conscious of the depth of experience which Lady Black and Lady Hale bring to bear in the field of family law, and particularly the circumstances in those children’s cases which may be affected by a ruling on section 25. It is evident that there is a very serious issue here, but I do not express any final view until a case arises which raises this very question. I am far from disagreeing with them, but I would like to reach a final view against the facts of an actual case. I express no view on the other issues as to the common law in Lady Black’s judgment for the same reason.

It follows that I would allow this appeal.

As Lady Hale says, this case is about the limits of parental responsibility in the case of a young person who has reached the age of 16, but does not have the mental capacity to make decisions for himself. This arises in the context of article 5 of the European Convention on Human Rights by which:

As she also explains (para 1), and as is common ground, the application of article 5 is to be tested by reference to three components: (a) the objective component of confinement in a particular restricted place for a not negligible length of time; (b) the subjective component of lack of valid consent; and (c) the attribution of responsibility to the State. This is the effect of Storck v Germany(2005) 43 EHRR 6 (“Storck”), followed by this court in Surrey County Council v P; Cheshire West and Chester Council v P[2014] UKSC 19; [2014] AC 896 (“Cheshire West”). It is further common ground that on the facts of this case, components (a) and (c) are satisfied. The area of debate is about component (b): whether on the facts of this case the exercise of parental responsibility could make up for the lack of consent by D himself.

That it could do so while he was under the age of 16 was not in dispute in the courts below. That was supported by reference to the decision of the Strasbourg court in Nielsen v Denmark (1988) 11 EHRR 175 (see Lady Hale para 34). It is worth stating at the outset the reasons for this view, as stated by Keehan J, and adopted by Sir James Munby P giving the leading judgment in the Court of Appeal (para 108):

The good sense of that appraisal has not, as I understand it, been challenged by any of the parties to this court. Nor is it suggested that, when D became 16, anything changed in practical terms, whether in respect of his own needs and best interests, or of his relationship with his parents or the public agencies involved. However, Keehan J was persuaded by the Official Solicitor that there was a fundamental change in the legal position so that the approval of the court was now required. The Court of Appeal disagreed.

That view is challenged in this court by the Official Solicitor, with the support of the Equality and Human Rights Commissioner as first intervener. The Court of Appeal’s approach is defended by the Council as the statutory authority responsible for safeguarding D’s interests. They are supported by the Secretaries of State for Education and Justice. They have intervened as having policy responsibility respectively for the Mental Capacity Act 2005 (MCA 2005) and the Children Act 1989, and for the Court Service. They are concerned that the outcome of the appeal could have significant implications for a large number of 16 and 17-year olds, who are being held in care across a variety of settings, ranging from foster care placements to residential holiday schemes for disabled children.

They also point out that the appeal takes place against the background of the Law Commission’s review of the law of Mental Capacity and Deprivation of Liberty, and in particular the deprivation of liberty safeguards (“DoLS”) (Law Com No 372). As the Commission explained in its Consultation Paper (CP No 222, paras 2.39-40), that review was prompted by the massive and unanticipated increase in cases requiring to be dealt with under the DoLS arrangements (from 11,300 in 2013-4 to 113,300 in 2014-5), following the decision of this court in the Cheshire West case. Their review has been followed more recently by the consideration by Parliament of the Mental Capacity (Amendment) Bill (now the Mental Capacity (Amendment) Act 2019). That provides for the replacement of DoLS by a new scheme of safeguards (the “Liberty Protection Safeguards”) for those who lack capacity under the MCA and who are deprived of their liberty, which will extend to 16 and 17-year olds. The background to the 2019 Act was described in the Explanatory Notes:

The Law Commission noted that its remit had been limited to children of 16 or over (para 7.20). It also noted the “complicating factor” that in the Nielsen case the Strasbourg court had recognised the right of parents in certain cases to consent to what would otherwise be a deprivation of liberty for their children; but it also noted that Keehan J (in the present case, decided since the consultation paper) had limited that approach to children under 16 (para 7.22). When what became the 2019 Act was presented to Parliament it was limited to those over 18, but it was later extended to those over 16. That followed an amendment proposed in the House of Lords by (inter alios) Baroness Thornton. It is of interest that she referred to evidence of the Royal College of Psychiatrists which

It seems therefore that the fixing of the age threshold in the new Act at 16 was directly related to the then understanding of the scope of parental responsibility as reflected in the judgment of Keehan J in the present case.

There is no dispute about the importance of the principle of parental responsibility in the common law. As Sir James Munby P said in In re H-B (Contact)[2015] EWCA Civ 389; [2015] 2 FCR 581, para 72:

Not surprisingly a corresponding principle is recognised under the European Convention on Human Rights. As the Strasbourg court said in Nielsen v Denmark (1988) 11 EHRR 175,“family life” in the Contracting States -

The common law principle is given specific statutory recognition in section 3 of the Children Act 1989, which defines parental responsibility as encompassing -

The judge’s conclusion that the legal position changed when D became 16 turned principally on his view of the change in the statutory framework applicable to such children. He said:

The President disagreed. As to the correct approach to article 5, and in particular the effect of Nielsen, he extracted the following propositions from the judgments in Cheshire West of Lady Hale and Lord Neuberger:

This led him to a discussion of the scope of parental responsibility in the context of Storck component (b), which in his view was governed by domestic law:

There followed a comprehensive review of the authorities culminating in the leading modern authority in Gillickv West Norfolk and Wisbech Area Health Authority [1986] AC 112. From an extended discussion of the speeches in Gillick (paras 74ff) he derived three propositions: (in very brief summary) first, that parental rights exist for the benefit of the child not the parent; secondly, that parental rights are to be exercised in the best interests of the children, and subject to the control of the court by reference to “general community standards”; thirdly, “the rejection of the rule that the age of discretion was fixed and the substitution of what it has now become customary to refer to as the acquiring of ‘Gillick capacity’” (para 79).

He concluded his discussion of Gillick with this summary:

He thought that the judge’s approach was inconsistent with the Gillick principle, and unsupported by the statute:

In relation to the 2005 Act itself he made two points

In this court the Official Solicitor submits that the President was wrong to treat Nielsen as relevant only to limb (b) of the Storck test, having regard to the later authorities reviewed in the judgment of Keehan J. To do justice to the argument, I cannot do better than quote from the written submissions advanced by Mr Setright QC and his team. He suggests that argument about the scope of limb (a) (confinement) is “to some extent, a sterile one” since -

In the remainder of his case he puts the main emphasis on developments in the law since Gillick, in particular the 2005 Act. Indeed he accepts that the President’s approach was “undoubtedly correct as a statement of the operation of ‘Gillick competence’ at common law in 1985 …”, but argues that it has been overtaken by developments in the law, in particular, the passage of the 2005 Act, and “also the trends in international human rights norms” (Case para 65).

He argues that these changes justify a change in the approach of the courts:

This argument is resisted by the City Council.

Without disrespect I can deal relatively shortly with the central argument in the Official Solicitor’s case, because I agree essentially with the reasons given by the President for rejecting the corresponding part of the judge’s reasoning. Like the President I see nothing in the 2005 Act which detracts from the common law principle or from section 3 of the 1989 Act. There is a presumption that Parliament does not change the common law by implication. Certainly in respect of a concept as basic and sensitive as parental responsibility one would expect clear words to indicate the nature of the change and its practical consequences. Not only is there nothing in the 2005 Act itself to indicate such a change, but, as the Secretary of State has shown (without challenge), there is nothing in the background to the Act to indicate such an intention. On the contrary it was made clear by the Law Commission and in Ministerial statements to Parliament, that there would be an overlap between the proposed regime and the 1989 Act. That position has been reinforced by the lack of anything in the 2019 Act to undermine the common law position, as reflected in the 1989 Act. In the absence of any specific legislative change I do not see how unincorporated international instruments can add anything to the argument.

I am also satisfied that this is not an area in which it would be appropriate for this court to accept the invitation to develop the law to fill a supposed gap left by the legislation, or otherwise “to reflect contemporary understandings of the rights of children”, as the Official Solicitor invites us to do. There is no parallel with the Montgomery case (Montgomery v Lanarkshire Health Board [2015] 1432) where a seven-justice court had been convened specifically to consider whether to depart from the controversial and much-criticised reasoning of the majority in Sidaway v Board of Governors of the Bethlem Royal Hospital and the Maudsley Hospital[1985] AC 871. Not only does the Secretary of State, who is responsible for legislative policy in this area, resist such a development, but the treatment of mentally incapacitated 16 or 17-year olds has been subject of Parliamentary scrutiny in connection with legislation passing through Parliament during the course of this appeal. In addition the experience of the Cheshire West decision should lead us to extreme caution in this difficult and sensitive area of the law.

On the question whether Nielsen was a case about Storck limb (a) or (b), I accept that, as the case was decided before the identification of the Storck components, it is a little artificial to attempt to fit the reasoning of the majority directly into that scheme. It is enough to say that, on the authorities as they stood before him, I see no error in Sir James Munby P’s approach. He referred to two passages in Cheshire West to support his view. The first (para 26) was from Lady Hale:

There was some discussion before us whether the second passage in the quotation from Lady Hale was an expression of her own opinion, or simply a recitation of counsel’s submission. Either way I can see nothing in the remainder of her judgment to indicate disagreement with that proposition, which also seems to me consistent with the first passage, and with the passage quoted from Lord Neuberger’s judgment.

It also seems to me the more natural interpretation. If Storck component (a) is directed to the objective quality of “confinement”, it is difficult to see how that quality is affected by whether or not it has been sanctioned by the parent. It is true that the attributes of “confinement” may vary in relation to children of different ages, as explained by Lord Kerr in Cheshire West paras 77-79 (a passage quoted by the Lady Hale: para 38). However, I am not persuaded that the clarity of the concept would be improved by further “nuancing” as the Official Solicitor suggests. In this case, as I have said, it is not in dispute that component (a) is satisfied.

I need to deal separately with Lady Hale’s judgment in the present appeal. She takes a rather different approach from that advocated by any of the parties before us, and perhaps for that reason finds it unnecessary to address in any detail the reasoning of the Court of Appeal. She deals relatively briefly with the majority judgment in Nielsen, which she treats as turning “on the comparative normality of the restrictions imposed on the freedom of a 12 year-old boy” (para 38). Later in the judgment (para 42), she discounts suggestions in Cheshire West that Nielsen was a case of “substituted consent”, because it had “suited counsel” so to argue. Instead she relies on later Strasbourg authorities as showing that limb (b) (that is, lack of valid consent) can be satisfied “despite the consent of a person with the legal right to make decisions on behalf of the person concerned”, the only exceptions being where “the evidence showed that the person concerned was willing to stay where he or she was and was capable of expressing that view”. She concludes accordingly that “parental consent cannot substitute for the subjective element of limb (b) of Storck”.

Later in her judgment (para 48) she reinforces that view by equating deprivation of liberty with other “fundamental human rights” such as the right to life or freedom from torture. She argues that it would be a “startling proposition” that it lies within the scope of parental responsibility to authorise violation of such rights. I say at once, with respect, that I am not persuaded that such comparisons are fair or helpful. D’s parents were not authorising the state to commit torture or anything comparable to it. They were doing what they could, and what any conscientious parent would do, to advance his best interests by authorising the treatment on which all the authorities were agreed. That this involved a degree of confinement was an incidental but necessary part of that treatment, and no more than that. On the President’s view, with which I agree, they were not “authorising a violation of his rights”, but rather exercising their parental responsibility in a way which ensured that there was no such violation.

More importantly, I do not accept that the majority reasoning in Nielsen, nor indeed what was said about it in Cheshire West, can simply be brushed aside. It was not just about the relative “normality” of the confinement (although some might share the minority’s doubts about that description of the forcible confinement of a 12-year old child for five months in a locked psychiatric ward). As has been seen (para 147 above), in Cheshire West Lady Hale herself described it as turning, not on the “normality” of the arrangements, but on “the proper limits of parental authority in relation to a child”. More specifically Lord Neuberger said that it was decided “on the basis that his mother was exercising her article 8 rights responsibly, in good faith and on the basis of medical advice …”.

Sir James Eadie QC, on behalf of the Secretaries of State, has helpfully analysed the majority judgment in Nielsen in terms which I would in substance endorse. He accepts that, not surprisingly in a case decided before Storck, there may be some overlap between the categories. But, in agreement with the President, he sees it as primarily about limb (b). He points to the emphasis given by the majority “at the outset” to family life under the Convention, encompassing the “broad range of parental rights and responsibilities in regard to the care and custody of minor children …” (para 61). He notes the following points from the judgment (paras 68-72):

The mother’s decision to have the applicant hospitalized “was a lawful exercise of parental powers under Danish law” and was also well-founded. The Danish courts found that the hospitalization decision fell within the mother’s competence as holder of parental rights.

The mother had taken her decision on the basis of medical advice from her family doctor and a professor, and had as her objective the protection of the applicant’s health. This was a proper purpose for the exercise of parental rights.

The mother’s decision was approved by the relevant social services authorities.

There was no suggestion that the treatment given at the hospital and the conditions under which it was administered were inappropriate in the circumstances. The applicant was in need of medical treatment for his condition and the treatment administered to him was curative.

There was no evidence of bad faith on the part of the mother. Hospitalization was decided upon by her in accordance with expert medical advice. It must be possible for a child to be admitted to hospital at the request of the holder of parental rights, a case which was not covered by para 1 of article 5.

That the court based its reasoning principally on the exercise of parental responsibility seems to me put beyond doubt by its concluding comment:

The Secretaries of State further submits (without contradiction) that the Strasbourg court has not departed from Nielsen in the three decades since the judgment was delivered. That submission is confirmed by the exhaustive review by Keehan J of the Strasbourg authorities relied on by the Official Solicitor before him (paras 44-61). The case has been consistently explained by the court itself as a case about the “responsible exercise by the applicant’s mother of her custodial rights”: see Koniarski v United Kingdom (Application No 33670/96) 30 EHRR CD 139 and DG v Ireland (Application No 39474/98) (2002) 35 EHRR 33. (Notably, the fact that both those cases related to 17-year olds was not cited as a ground of distinction). In HL v United Kingdom(2004) 40 EHRR 761, it was cited with approval, but distinguished on the basis that no one had legal authority to act on the adult HL’s behalf in the same way as Jon Nielsen’s mother (para 93). It was also cited with approval by the Grand Chamber in Stanev v Bulgaria(2012) 55 EHRR 696, where it was explained as “the exercise of exclusive custodial rights over a child who was not capable of expressing a valid opinion” (para 122). Keehan J (para 42) noted some doubts about the case expressed by Lord Walker in Austin v Comr of Police [2009] 1 AC 564, para 42. But no such doubts appear to have found their way into the Strasbourg jurisprudence. It was and remains the leading Grand Chamber decision on the scope of such “parental rights and responsibilities” in the context of article 5.

It is unnecessary in my view to decide whether the case is to be regarded as a case of “substituted consent” so as to bring it directly with Storck limb (b) (as was hinted at in Stanev v Bulgaria [2012] 55 EHRR 22, para 130), or whether it is simply an exception to the Storck categorisation, justifiable in its own terms by reference to the scope of family life under article 8. For the present purposes, it provides amply sufficient support in Strasbourg case law for the President’s reliance on equivalent domestic law principles to determine the present case.

Lady Black, while agreeing with Lady Hale, has introduced a new line of reasoning based on a review of the common law authorities preceding Gillick. This leads her to reject the President’s view of the relevance of that case to decisions relating to detention. Instead she reads those authorities as showing that in this context “reaching the age of discretion was a matter of attaining the requisite chronological age, and not a matter of mental capacity” (para 56, 68). As I understand her judgment (paras 71-72), she would regard 16 as the appropriate age in the modern law, taking account inter alia of the recognition by the legislature in successive Acts of that age as a “pivotal” turning point, most recently in the 2019 Act.

This line of reasoning was not subject to detailed argument at the hearing. For the moment I remain unconvinced that the earlier cases can be relied on to limit the scope of the judgments in Gillick in the way she proposes, or that the President’s conclusions are undermined. However, I acknowledge that this approach, if correct, may have advantages for the certainty and coherence of the law, particularly if taken with another important point which emerges from her review of the earlier cases. That is the willingness of the courts since the 19th century to take guidance from the legislature as to where to draw the lines in relation to the limits of parental responsibilities (see para 60, citing Cockburn CJ in R v Howes (1860) 3 El & El 332). In the present case there is the added consideration that, as noted above (para 130), the exclusion of those under 16 from the new legislative scheme appears at least in part to be a reflection of the legislature’s understanding of the law following Keehan J’s judgment, which to that extent may be seen as having the implicit endorsement of Parliament.

I note with some concern that Lady Hale (para 50) has raised a question as to the logic of the differential treatment of those under 16, at least in the context of article 5 taken on its own. That does not reflect any issue between the parties. Keehan J’s application of parental responsibility to those under 16 has not been questioned by any of the parties in the Court of Appeal or in this court. Nor does Lady Hale, as I understand it, suggest that there is anything in the Strasbourg law as it stands which invalidates that aspect of Keehan J’s judgment. For the time-being his reasoning remains the law, and as such appears to fit well with the new legislative scheme.

I have nothing to add to what Lady Black says in respect of section 25 of the Children Act 1989, with which I agree.

For the reasons stated earlier in this judgment, in substantial agreement with the reasoning of the Court of Appeal, I would have dismissed the appeal.