A group of eminent experts on the deprivation of liberty safeguards (DoLS) published a very important piece of research this month. Writing in The Psychiatrist
they reported the results of an exercise which asked a group of lawyers, psychiatrists, best interest assessors (BIA’s) and independent mental capacity advocates (IMCA’s) to make judgments about whether situations described in 12 vignettes amounted to a deprivation of liberty or not. Overall agreement between professionals was extremely poor (Kappa score
0.16); in descending order it was highest amongst IMCA’s, then psychiatrists, then BIA’s, and lowest amongst the lawyers. Lawyers were the most likely to find deprivation of liberty to be occurring
The results of this study are really concerning. The proper functioning of the DoLS relies upon professionals being able to identify deprivation of liberty where it is occurring in order to apply the safeguards, but this research suggests that whether or not the safeguards are applied will depend upon the highly variable judgment of individual professionals. Furthermore, the study supports the comment by the Mental Health Alliance
last year that
‘…deprivation of liberty’ is very ill defined and therefore subject to a very wide range of local interpretations. This was obvious even prior to 1 April, when the forecasts of likely cases from supervisory bodies of similar size, and based on the same guidance, varied from single figures to many hundreds, and it is apparent now in the frequent reports of disputes between managing authorities, CQC inspectors, IMCAs and assessors as to whether a particular situation amounts to deprivation.’ (page 13
I suspect, in fact, the overall levels of disagreement may be higher even than this study suggests, given that it focuses exclusively on hospital settings. Judgments about whether a situation amounts to a deprivation of liberty may be even more variable in community settings (care homes, supported living, family homes). It would be a very welcome exercise if an equivalent study were carried out looking at community based settings and included also the judgment of registered managers, care inspectors and social workers who are not BIA’s.
Part of the problem, of course, is that the judgments themselves that professionals must rely upon apply highly variable reasoning as to why a person is – or is not – deprived of their liberty. It was hoped by many that the Court of Appeal judgment in P and Q 
would offer some more determinative guidelines, but as I wrote in June
more recent judgments are difficult to reconcile with the reasoning given in P and Q. I hear on the grapevine that many major judgments are to be appealed in any case (including P and Q in the Supreme Court, Re RK
is said to be being appealed by the official solicitor, and Cheshire West and Chester Council are currently appealing against Re P
). The cost of this uncertainty is high for all. It is high for all
of those involved in court proceedings because they must first deal with the contentious question of whether Article 5 is even engaged – before they even get onto issues of capacity and best interests. It is high for all of those who work with the DoLS, because their time must be spent considering and resolving disputes over whether they should even be engaged or not. The human rights cost of this uncertainty will be high for all those people who really need safeguards (more on which below) but for whom they are not applied because of uncertainty.
And in the background to many of these cases about the scope of Article 5 are difficult questions about resources, and political questions about the proper role of the state in private life. Why else would local authorities take up court time arguing that if deprivation of liberty were found to be occurring in the present case ‘the vast, if unquantifiable, number of necessary reviews of such a character would surely be beyond the present capacity of the Official Solicitor’s department and in particular of the Court of Protection’ (P & Q, ), and ‘the implications will be formidable’ (Re RK, ). The judges are keen to emphasise that resource considerations cannot impact upon the ‘objective test’ of whether Article 5 is engaged or not; however it is hard to know how far their reasoning can be insulated against the knowledge that the system as it stands simply cannot cope with the strain of a wider definition of deprivation of liberty. The impact assessment
for the DoLS predicted they would cost £13.9m in their first year, £12.1m in their second year and eventually arrive at £4.3m ‘steady state’ cost. Research suggests
that the cost of a single assessment under the DoLS was £1277, not £600 as the impact assessment estimated; but the overall number of assessments has been far lower than anticipated. I am unaware of any research on the cost of the DoLS appeals to the justice system as a whole (including the courts, Official Solicitor, legal costs to supervisory bodies and the Legal Services Commission), but I would suspect that the cost of an individual appeal is far in excess of the amount predicted as cases are turning out to be lengthy, complicated and frequently appealed.
All of which brings me to the question – what were the DoLS supposed to do? The meaning of ‘deprivation of liberty’ employed will very much depend upon the perceived threat that we are meant to be safeguarding people from. Whilst I am sure that many individuals involved in the process of consultation and drafting that preceded the DoLS may have had a vision of what they were intended to do, it is very unclear from the policy documents produced at the time. The consultation documents are depressingly circular on this matter, emphasising only that people who are ‘deprived of their liberty’ need ‘safeguards’, but offering no clear guidance on what ‘liberty’ itself (and deprivation of) might look like. My sense through talking to different people involved in the DoLS has been that it meant different things to different people, and I would like to consider some possible accounts of what the DoLS may be been intended to achieve. I confess to once having held quite a strong personal view of what DoLS could achieve – extra protection for people who were subject to the ‘complete and effective control’ of others (any other) to ensure greater accountability and contestability of any restrictions placed upon them. Today, I am less sure than ever that DoLS could achieve this, and I am uncertain what their future could be. Perhaps much of what I hoped DoLS would achieve is better achieved through wider access to advocacy services and more robust and aspirational regulation of care.
Guarding against ‘states of domination’ in care settings
In Republican political philosophy
(nothing to do with the Republican Party in the US) liberty is defined as ‘non-domination’. This means that although – inevitably – everyone will always be subject to the power of other individuals or groups (the state, corporations, employers, family, etc) in some respects, we can still enjoy a state of liberty so long as we are not subject to the arbitrary
power of another. So, for instance, an employer’s power over employees is limited by employment law and contractual arrangements. The power of a government may be limited by constitutional arrangements or international law. A judge’s power is limited insofar as the common law and statute directs him to a certain set of outcomes. In essence, the task is to reduce discretion
in how power is exercised, and in this sense Republican philosophy has many parallels with the ideas enumerated by Tom Bingham in The Rule of Law
(as an aside, I would suggest that the DoLS and even the MCA fail the ‘rule of law’ test given the wide variations in interpretation and lack of ease with which applications may be contested). We can reduce the discretion in how power is exercised through a variety of means, including increasing transparency
so that others can see and understand who and why power is being exercised and contestability
so that abuses of power can be brought to account and corrected.
In care settings, carers of all kinds (be they family-carers or paid care workers) may exercise considerable power over those they care for. The DoLS could be seen as a means to ensure that where that power reaches a certain threshold of interferences with liberty it is held to account to ensure that it is not exercised inappropriately. This would see the DoLS as a means of ensuring that restrictions on liberty were kept to a minimum, and were not used to abusive ends. They would, on this analysis, have tremendous safeguarding potential. BIA’s, IMCA’s and ultimately the court would be able to scrutinise the exercise of power and – if necessary – contest it, to ensure it was in accordance with the Mental Capacity Act 2005 and human rights values.
There is some support for the notion Article 5 is concerned with reducing arbitrariness in the exercise of power in HL v UK (2004)
The Court observes that, as a result of the lack of procedural regulation and limits, the hospital’s health care professionals assumed full control of the liberty and treatment of a vulnerable incapacitated individual solely on the basis of their own clinical assessments completed as and when they considered fit: as Lord Steyn remarked, this left “effective and unqualified control” in their hands. While the Court does not question the good faith of those professionals or that they acted in what they considered to be the applicant’s best interests, the very purpose of procedural safeguards is to protect individuals against any “misjudgments and professional lapses” (Lord Steyn, paragraph 49 above).
The point is that the safeguards are there to scrutinise decision making in those who exercise power, to ensure it is exercised appropriately. No doubt this sense of being scrutinised is not always desirable to those who have full faith in their own judgment; but it is a principle of Republican liberty that even if a person exercising power never abuses it, the mere possibility that they could diminishes liberty because subjects remain vulnerable to abuse nevertheless.
But if the DoLS were meant to achieve this in hospitals, it is less clear that they were elsewhere. The definition of deprivation of liberty given in HL v UK (2004) – being subject to ‘complete and effective control’ and being ‘under continuous supervision and control and… not free to leave’ – has not been widely taken up by the courts in relation to other care settings. In P and Q  the Court of Appeal rejected the analysis that P (MEG) was deprived of her liberty because she was subject to the ‘complete and effective control’ of those who cared for her. Instead it pointed to the ‘normality’ of the setting (a group home) and the absence of P’s objections; no matter that HL himself had not been objecting, nor that ‘normality’ itself may – for many – turn on how restrictive a setting is, not it’s legal status or aesthetic appearance. It is possible that the Supreme Court (if the case ever gets there) may overturn this ruling and supply a definition more in accordance with Republican ideals of limiting the exercise of arbitrary power by care professionals regardless of setting, but it seems highly unlikely to me that they will ever extend this to include family carers.
There is no a priori
reason why care by one’s family should be more ‘free’ that institutional settings – meant in the sense of the free exercise of will and unrestricted movements – but I imagine the courts would find it highly impracticable and politically toxic to find deprivation of liberty to be occurring there. There is a sense in some of the judgments that they recognise this tension – in Neary v Hillingdon 
Jackson J says Steven Neary was subject to deprivation of liberty ‘to some degree or other necessarily… wherever he lives
’ . There are cases where it seems to me that the court could potentially have labelled care in the family home an abusive deprivation of liberty on this analysis (A Primary Care Trust v P
; Re MP
) but they did not discuss this; no doubt this would open up a real can of worms about the Article 5(5) right to sue for compensation for unlawful detention amongst other things. The courts have indicated that in the context of a ‘loving family home’ even quite serious restrictions on liberty will not engage Article 5, and yet they have not ruled out the possibility altogether (A Local Authority v A 
Safeguarding against the intrusion of the state into home and family life
All of which would seem to point towards a view that ‘deprivation of liberty’ is defined by reference to interferences with liberty by the state
. Indeed, ‘imputability to the state’ is part of the tripartite analysis of deprivation of liberty that is cited in almost every judgment on Article 5 in the Court of Protection (see paragraph 77 in JE v DE (2006)
for its first formulation). This would help explain why E (G v E), HL (HL v UK), Q (P and Q) and Steven Neary were all deprived of their liberty in institutions but not deprived of their liberty by their families (or foster families). It would also tie into the definition proffered by the Joint Committee on Human Rights
We consider that if it is known that a person will be taken from their home to a place where they will be prevented from leaving, and complete and effective control will be exercised over their movements, that person is deprived of liberty from the point of removal from their home
The idea that the deprivation of liberty safeguards are there to safeguard against an intrusive state will no doubt appeal to many, particularly those of a conservative or libertarian bent. And it does help to make sense of many of the disputes that have come before the Court of Protection and been published. The DoLS have in many ways become an elaborate (and costly) dispute resolution service for struggles between public authorities and P’s family over where P should live. But surely, this is more the territory of Article 8, not Article 5? I do wonder how far the DoLS have been ‘captured’ by these types of disputes because it is often very difficult and costly to get them into court in other ways. The DoLS offer the benefits of detailed reports which consider the views of family, of advocacy services and of legal aid and permission-free routes of appeal to court (in theory, at least). Given that disputes may break out in respect of other matters, however, which cannot be characterised as ‘detention’ (for instance, treatment – and withdrawal of treatment – decisions, contact with family members, best interests decisions about different types of accommodation), it seems rather unfortunate that there are not more accessible (and perhaps more lightweight) procedures for resolving these issues than the sledgehammer of the DoLS.
In any case, this interpretation of the purpose of the DoLS is fraught with danger. Many people – particularly those who grew up in long-stay hospitals and are still confined either in private hospitals or care homes – would not obviously receive protection from this interpretation. Furthermore, many people who are subject to highly restrictive care may be deprived of their liberty in private care homes, under privately commissioned care, which has had very little to do with the state – or at least to do with local authorities and NHS bodies. The ruling in Storck v Germany (2005)
found deprivation of liberty to have occurred in a private hospital, and found a number of routes through which it was imputable to the state including the positive obligations of states to prevent unlawful deprivation of liberty by private parties. We should not forget that sometimes care that amounts to detention – as in the case of Storck – may be commissioned by families themselves (although my research suggests
that these cases may not be being picked up on by care providers for the purposes of DoLS).
If we remember our history, it was not the public asylums that concerned the Reformists of the 18th
Century, but also the private madhouses where families would pay a keeper to lock away their relations, ‘peddling discreet silences’
has put it. The 1763 Select Committee on Madhouses, chaired by Thomas Townsend, reported that the insane were kept in appalling conditions in Madhouses, and that they were being used to confine people who were not lunatics. The 1774 Madhouses Act required ‘an order in writing, under the hand and seal of some physician, surgeon or apothecary (*d) that such person is proper to be received into such house or place as a lunatic’ in order for a person to be admitted. These are the first procedural safeguards against arbitrary detention, and the Act also introduced the licensing and inspection of Madhouses (albeit somewhat ineffectually). The reforms were fiercely resisted by physicians and clergy, who resented the scrutiny and bureaucracy, and feared it would hamper the profitability of their operations. Similar forces may also be at work today in the operation of the DoLS across a variety of settings. Importantly though, the 18th century reforms brought private sites of care within the ambit of state control, and perhaps it could be through the route of inspection and regulation of care settings that we might distinguish private institutional care from the family home. But even some contractual care arrangements may lie outside the CQC’s scope of registration (e.g. employment of personal assistants or nursing staff directly by family members).
It seems, in any case, as if the courts do wish to reserve the meaning of ‘deprivation of liberty’ for the family home for some cases, as Munby LJ enumerated a procedure for guarding against unlawful deprivation of liberty in the family home where local authorities found it to be occurring
(A Local Authority v A 
). He offered less clarity on when it could be said to be occurring; and so we are back full circle to our original question – what is ‘deprivation of liberty’ and what are safeguards against it for? In some cases, far from guarding against intrusion into home and family life by the state, they may facilitate it.
Ensuring ‘P’ is not confined against his wishes
One other way of defining ‘deprivation of liberty’ would be by reference to ‘P’ being confined in a place against his wishes. This interpretation is lent weight by the ‘objections’ criterion outlined by Wilson J in P and Q . As I
, and others
, have argued elsewhere there is a real difficulty here around interpreting communications of people’s wishes – particularly where people lack verbal ability. Once again, it seems to me to place too much power and discretion in the hands of caregivers to decide whether ‘challenging behaviour’ or ‘wandering’ or pleas to go back to a home that has perhaps long since been sold, are ‘genuine’ objections or merely symptoms of mental disorder that they can discount. Clearer guidance would be welcome. There is also the difficulty that medication may mask objections – acknowledged by Wilson J in P and Q, but then strangely ignored in relation to Q herself being prescribed Risperidone. Again, the courts seem to have left it open to caregivers to argue that medication that has the effect
of suppressing behaviours that could be signs of unhappiness, or even clearly expressed objections, is prescribed with the intention
of managing ‘symptoms’ (anxiety, aggression, distress etc) and should not be taken into account. In some cases, a person’s tractability can be directly controlled by medication, but the medication in turn is controlled by the caregivers, which seems to me a rather alarming form of control in its own right.
In any case, I think for purely legal reasons the courts should be cautious of eliding deprivation of liberty with objections as it risks contravening the fundamental message of the Bournewood ruling itself: that HL was detained, despite the fact he made no protest and never attempted to escape.
Safeguarding against ‘institutionalisation’
Another way of attempting to distinguish care in the family home from care settings where Article 5 has been found to be engaged is through distinguishing ‘institutional’ from ‘non-institutional’ settings. This appears roughly congruent with the criterion of ‘normality’ given by Wilson J in P and Q :
If the person is living with her parents or other members of his natural family in their home, she is living – in that respect – the most normal life possible. Typically – but sadly not always – there will be no deprivation of liberty in such circumstances: A v A Local Authority, cited above, at , per Munby J. Not much less normal for this purpose is the life of a child in the home of foster parents or of an adult, such as Mr HL, in the home of carers; such Mr Gordon is constrained to accept. But, even when the person lives in an institution rather than in a family home, there is a wide spectrum between the small children’s home or nursing home, on the one hand, and a hospital designed for compulsory detentions like Bournewood; and it is in my view necessary to place each case along it. 
It also seems to me to be roughly the basis of the decision of the Labour government not to include supported living in the safeguards
‘because it considered it less likely that severe restrictions would be placed on people in supported living arrangements, who would tend to lead more independent lives.’ Likewise it explains an extensive focus on how person centred care could reduce the risk of deprivation of liberty in the draft deprivation of liberty safeguards code of practice – passages which, I note, did not make it into the final DoLS code of practice – despite considerable support for them in consultation responses I have seen.
Where one stands on this analysis may depend to a considerable degree what one regards as an institution. Famously the sociologist Erving Goffman defined a ‘total institution’ as a place which is segregated from the rest of the community, where a ‘batch’ of people live, work (if at all), play (if at all), and eat and sleep together in accordance with institutionally prescribed regimes. There is a strong division between those in authority (staff) and those who are subject to authority. Even on outings outside the institution, its inhabitants are still regarded by the staff – and the wider world – as being subject to its overall control. The distinction between ‘total institutions’ like convents and army barracks and those institutions we do want the DoLS to pick out may turn on consent to admission and freedom to leave. It is pretty straightforward to argue that hospitals would be captured by the definition, and probably most large care homes (perhaps more than the Ministry of Justice and care providers might like, given the running costs of the safeguards). But the definition might also pick out other sites that are less attractive for the engagement of Article 5, like boarding schools or children’s homes – although these are matters of dispute in the courts at present (see Re RK, and also C v A Local Authority 
The ‘total institution’ analogy with deprivation of liberty may break down where people regularly leave the institution, for instance to attend school, college or a day centre, as in the case of G v E. Legally speaking, a ‘supported living’ service is domiciliary care in a person’s own home, but such arrangements can take on a very institutional character such that they are in practice hard to distinguish from other more explicitly institutional care arrangements (see this post
, for more discussion). This may be the case even when the ‘batch’ of cohabitants is very small, or a person even lives alone. Researcher Noga Avni has argued
that the family home may take on many characteristics of Goffman’s ‘total institutions’ for battered women, but the reasons she gives are closely connected with the degree of control a partner exercises over the woman, and how ‘free’ she is to leave. It seems to me that this route to defining institutions might take us straight back to debates around how far ‘deprivation of liberty’ is connected with values like autonomy, privacy and being subject to the control of others. Goffman, for his part, explicitly states that ‘total institutions’ are distinct from the family home, but I would suggest that this position is maintained from an experience and understanding of home that is characterised by being an able-bodied and able-minded adult American male, and may not be an experience shared across all spaces called ‘home’. In the high court ruling on P and Q (MIG and MEG 
) Parker J asserts that the primary intention was ‘to provide them each with a home
’ , but it is doubtful this is the kind of home Goffman had in mind when distinguishing home from institutions.
In any case, it is entirely possible for care to be ‘person centred’ and yet still be controlling. Indeed, this is arguably the desirable outcome if the MCA and DoLS are applied properly; all restrictions on liberty would be kept to a minimum and tailored to the individual’s own ‘best interests’ and not the needs of the institution. But restrictions on liberty they remain, nonetheless. The crucial point is that the institution – with some limited oversight from the commissioners of care – retain ultimate control over these restrictions and a person’s autonomy. We also know that there can be a large reality gap between the ‘person centred’ aspirations of a service and the reality (see this post
on a Castlebeck care home, for more discussion). We may not be able to assess the size of that gap without a mechanism – like the DoLS or good regulation – that scrutinizes in detail the restrictiveness of care regimes. And without a definition of either Article 5 or Article 8 that captures excessively institutional regime and renders them unlawful, it is hard to see how either regulator or any set of safeguards could police these issues.
We must be careful, when we consider what the deprivation of liberty safeguards are really there for, that we are not ‘beguiled by the brochure’. Care that seems good in the care plan and inspection reports often is not when you scratch beneath the surface. Family care is not always ideal. And more importantly, circumstances can change very quickly and care can deteriorate. This is problematic in all care settings, of course, but where people are subject to high levels of control and little concourse with the wider world – who will know, and who will challenge? Where care receivers have few opportunities to engage with the wider community, where there are few channels of communication not controlled by caregivers, where there are no mechanisms for care-receivers to challenge the authority of those who care for them, it seems there is ample opportunity for darker practices to flourish.
Were the DoLS intended to shine a light on these practices? I am not sure that they were; and certainly that is not what they are achieving in practice on any systematic basis. The only great achievement of the DoLS, it seems to me, is to make it extremely difficult for the UK to be challenged again at the European Court of Human Rights in a similar fashion to HL v UK (2004), because you would have to exhaust domestic remedies first. And the DoLS do, eventually, provide a remedy. But they provide a remedy in an arbitrary and unsystematic fashion – applied haphazardly according to the discretion of professionals, and interpreted in what seems from the outside to be an unsystematic way by the courts. Given the high cost to the state, and the paucity of protection for care recipients, I would be very surprised if this system were still here in ten years time. But next time we sit down to consult the real question we need to ask ourselves is this: what are we trying to protect people from?