Introduction
The Mental Capacity Act 2005 (MCA) represents the principal legislative framework governing decision-making for adults who lack capacity in England and Wales. Its stated philosophical commitment is to autonomy: the Act enshrines a presumption of capacity, requires that all practicable steps be taken to support decision-making, and mandates respect for the individual’s past and present wishes in any best-interests determination. Yet the structural logic of the MCA contains an internal tension. Once a person is found to lack capacity under section 2, the decisional authority shifts from the individual to a substitute decision-maker applying the best-interests checklist in section 4. The question is whether the safeguards built around this process — the statutory principles, the participation requirements, the role of the Independent Mental Capacity Advocate (IMCA), the involvement of the Court of Protection, and the Deprivation of Liberty Safeguards (DoLS), now under reform — genuinely protect the autonomy of the person concerned, or whether they operate primarily to legitimise decisions made by others in ways that may not faithfully reflect or prioritise the individual’s own values, preferences and wishes.
This essay argues that while the MCA’s safeguards contain important autonomy-protecting elements, they do not properly protect autonomy in best-interests decisions for three principal reasons. First, the binary capacity threshold operates as a cliff-edge that extinguishes decisional authority rather than supporting its graduated exercise. Second, the best-interests test, despite its subjective elements, permits and often encourages an objective welfare-based override of the person’s wishes without adequate justificatory rigour. Third, the procedural safeguards — IMCAs, DoLS, and the Court of Protection — suffer from practical deficiencies and institutional dynamics that weaken their autonomy-protective function. The overall effect is a framework that professes fidelity to autonomy but structurally prioritises protection.
The Autonomy Principle and Its Philosophical Foundation in the MCA
The MCA’s commitment to autonomy is expressed most directly in its statutory principles. Section 1(2) establishes a presumption of capacity. Section 1(3) requires that a person must not be treated as unable to make a decision unless all practicable steps to help them do so have been taken without success. Section 1(4) provides that a person is not to be treated as lacking capacity merely because they make an unwise decision. These principles draw on a liberal autonomy framework: competent adults are entitled to make decisions about their own lives, including decisions that others consider foolish or harmful (Herring, 2020). As Butler-Sloss LJ stated in Re MB (Medical Treatment) [1997] 2 FLR 426, the right of the competent patient to refuse treatment exists regardless of the rationality of the refusal.
However, the MCA’s autonomy commitment is conditional. It operates only up to the point at which capacity is found to be absent. The Act does not adopt a supported decision-making model, in which the individual retains decisional authority even where their cognitive abilities are compromised. Instead, it follows a substituted judgment model tempered by participatory elements. This philosophical structure means that autonomy, as protected by the MCA, is gatekeeper-autonomy: it depends entirely on whether the individual passes the capacity threshold. Once they do not, the framework pivots from autonomy to welfare, mediated through the best-interests checklist. The critical question, therefore, is whether the safeguards surrounding this pivot are adequate.
The Binary Capacity Threshold as an Autonomy Cliff-Edge
The MCA’s capacity test, set out in sections 2 and 3, requires a two-stage assessment. Under section 2(1), a person lacks capacity if at the material time they are unable to make a decision for themselves because of an impairment of, or a disturbance in the functioning of, the mind or brain. Under section 3(1), a person is unable to make a decision if they cannot understand, retain, use or weigh the relevant information, or communicate their decision. This test is decision-specific and time-specific, which is an important safeguard against blanket declarations of incapacity.
Nevertheless, the test operates as a binary threshold: a person either has capacity or does not. There is no intermediate category in which a person with reduced but not absent decision-making ability retains partial decisional authority or receives enhanced support to exercise residual autonomy. The consequence is that individuals at the margin — those who can express clear preferences and understand the broad nature of a decision but who struggle with the weighing requirement — lose all decisional authority once the threshold is crossed.
The difficulty with the weighing component is particularly significant. In A Local Authority v A [2010] EWHC 1549 (Fam), Bodey J found that a woman who wished to live with her husband, whom the local authority considered abusive, lacked capacity because she was unable to weigh the relevant information. The court accepted that she could understand the information and retain it, but that her emotional attachment prevented genuine weighing. Kong (2017) has argued persuasively that the weighing criterion risks conflating emotional reasoning with cognitive inability, thereby pathologising choices that are value-driven rather than cognitively impaired. If weighing is interpreted so as to require that the person give appropriate weight to risk, it collapses into a requirement of rational deliberation, which contradicts section 1(4)’s protection of unwise decisions.
The Supreme Court in PC v City of York Council [2013] EWCA Civ 478 (Court of Appeal, subsequently considered in related proceedings) emphasised that capacity must not be conflated with wisdom, and that the question is whether the person can use or weigh the information, not whether they do so wisely. Yet the practical application of this distinction remains contested. Assessors — often social workers or clinicians rather than lawyers — exercise significant discretion in determining what it means to weigh information adequately. As Series (2015) has noted, the assessor’s own risk aversion or welfare orientation can subtly influence the outcome of the assessment, particularly in cases involving vulnerable adults whose choices are perceived as dangerous.
This binary threshold therefore functions as an autonomy cliff-edge. Below the threshold, the individual’s voice is reduced to one factor among many in a best-interests determination, rather than being the primary or controlling consideration. A more autonomy-protective framework would incorporate a graduated approach, in which the degree of support and the weight given to the person’s expressed wishes would increase as capacity diminishes, rather than decisional authority being entirely removed at a single point. The UN Committee on the Rights of Persons with Disabilities (CRPD Committee) has gone further, arguing in General Comment No. 1 (2014) that the CRPD requires the abolition of substituted decision-making entirely in favour of supported decision-making. While this position is not reflected in English law, it highlights the extent to which the MCA’s binary model is in tension with a robust conception of autonomy.
The Best-Interests Test: Autonomy as One Factor Among Many
Section 4 of the MCA sets out the best-interests checklist. It requires the decision-maker to consider, so far as reasonably ascertainable, the person’s past and present wishes and feelings, the beliefs and values that would be likely to influence the decision, and other factors the person would be likely to consider (section 4(6)). The decision-maker must also permit and encourage the person’s participation (section 4(4)) and consult those close to the person (section 4(7)).
These provisions contain genuine autonomy-protective elements. The requirement to consider the person’s wishes and feelings, and the broader requirement to consider their values, introduces a subjective dimension into the determination. The aspiration, at least in principle, is that the best-interests decision should approximate what the person would have decided for themselves.
However, the statutory framework stops short of making the person’s wishes determinative or even presumptively controlling. Section 4(6) requires only that the decision-maker “consider” these matters. There is no statutory hierarchy among the best-interests factors. The person’s wishes may be outweighed by welfare considerations, and the Act provides no structured framework for determining when such an override is justified. This contrasts with the position under Article 12 of the CRPD, which, on the CRPD Committee’s interpretation, requires that the person’s will and preferences be respected rather than overridden by objective welfare assessments (CRPD Committee, General Comment No. 1, 2014).
The leading authority on the weight to be given to wishes and feelings is Aintree University Hospitals NHS Foundation Trust v James [2013] UKSC 67, in which Lady Hale stated that the purpose of the best-interests test is to consider matters from the patient’s point of view, and that the focus must be on whether the proposed treatment is in the patient’s best interests, not whether it is in the interests of others. This formulation is autonomy-oriented in its framing, but it does not resolve the central difficulty: when the person’s expressed wishes conflict with what clinicians or courts consider to be objectively in their welfare, the Act permits the latter to prevail.
In Wye Valley NHS Trust v B [2015] EWCOP 60, Peter Jackson J gave particularly strong weight to the wishes and feelings of a man with a gangrenous foot who refused amputation, finding that his best interests were served by respecting his refusal even though it would likely lead to his death. Peter Jackson J stated that the wishes and feelings of the person concerned should be given the most prominent consideration, and that the court must be very cautious about overriding the expressed wishes of a person with capacity issues. This decision represents a high-water mark of autonomy protection within the best-interests framework. Yet subsequent decisions have not consistently followed this approach. In N v ACCG [2017] UKSC 22, the Supreme Court, while emphasising the importance of the person’s wishes, did not establish a general presumption in their favour. The weight to be given to wishes remains discretionary and fact-specific.
The absence of a structured justificatory framework for overriding the person’s wishes is the most significant autonomy deficit in the best-interests test. Donnelly (2016) has argued that the MCA should adopt a rebuttable presumption that the person’s expressed wishes should be followed, with the burden of proof on those seeking to override them to demonstrate that there is a compelling reason relating to serious harm. Such a model would preserve the ability to protect individuals from exploitation or immediate danger whilst ensuring that the default position is one of respect for the person’s own values and preferences. The current law contains no such presumption, and as a result, autonomy competes on equal terms with welfare considerations in a checklist that offers no principled basis for resolving the conflict.
The Functional Limits of Procedural Safeguards
Independent Mental Capacity Advocates
The IMCA service, established under sections 35–41 of the MCA, provides an important safeguard for unbefriended individuals — those who lack family or friends to consult. IMCAs must be appointed where decisions concern serious medical treatment or changes in accommodation. The IMCA’s role is to represent and support the person, ascertain their wishes and feelings, and present those to the decision-maker.
In principle, the IMCA service strengthens autonomy by ensuring that even isolated individuals have a voice in the process. In practice, however, the service is limited in several respects. First, the statutory duty to appoint an IMCA is triggered only for defined categories of decision; many routine but significant best-interests decisions — such as day-to-day care arrangements, contact restrictions, and financial management — fall outside the mandatory scope. The Mental Capacity Act 2005 (Independent Mental Capacity Advocates) (General) Regulations 2006 (SI 2006/1832) extended the IMCA role to care reviews and adult protection cases on a discretionary basis, but take-up has been uneven (House of Lords Select Committee, 2014). Second, IMCAs do not have decision-making authority; they can challenge decisions, but the ultimate determination remains with the original decision-maker or the court. Third, the quality and resourcing of IMCA services vary considerably across England and Wales, with the House of Lords Select Committee on the MCA (2014) concluding that the Act was not being properly implemented and that the IMCA service was under-resourced and inconsistently applied.
Deprivation of Liberty Safeguards and Their Reform
The Deprivation of Liberty Safeguards (DoLS), introduced by the Mental Health Act 2007 and inserted as Schedule A1 to the MCA, were designed to provide Article 5 ECHR-compliant authorisation for deprivations of liberty of persons lacking capacity in hospitals and care homes, following the European Court of Human Rights’ decision in HL v United Kingdom (2004) 40 EHRR 32. The DoLS require a series of assessments, including a best-interests assessment, before a deprivation of liberty can be authorised by a supervisory body.
The DoLS have been widely criticised as bureaucratically complex, poorly understood by practitioners, and ineffective as an autonomy safeguard. The House of Lords Select Committee (2014) described the DoLS as not fit for purpose. The Law Commission (2017) conducted a thorough review and recommended their replacement with the Liberty Protection Safeguards (LPS), enacted as Schedule AA1 to the MCA by the Mental Capacity (Amendment) Act 2019. The LPS were intended to simplify the authorisation process, extend its scope to domestic settings and community arrangements, and introduce a more streamlined assessment procedure.
However, the LPS have not been brought into force. The Government announced in April 2023 that it did not intend to implement the LPS in the current Parliament, effectively leaving the widely criticised DoLS regime in operation (Department of Health and Social Care, 2023). This creates a significant gap in the safeguard framework. The DoLS remain cumbersome, under-resourced, and poorly understood. Local authorities have consistently reported backlogs in processing DoLS applications, meaning that many deprivations of liberty occur without lawful authorisation. This is not merely an administrative failing; it is a fundamental breach of the autonomy of the individuals concerned, who are deprived of their liberty without the procedural protections required by Article 5 ECHR.
Even where the DoLS operate as intended, they provide limited autonomy protection. The best-interests assessment within the DoLS framework is carried out by a best-interests assessor, often a social worker or healthcare professional, who applies the section 4 checklist. As discussed above, this checklist does not give presumptive weight to the person’s wishes. The Relevant Person’s Representative (RPR), appointed under the DoLS to support the individual, has in practice often been a family member or friend whose understanding of their role is limited. The overall effect is that the DoLS authorise deprivations of liberty rather than critically scrutinising them from an autonomy perspective.
The Court of Protection
The Court of Protection, established under Part 2 of the MCA, has jurisdiction over decisions concerning the property, affairs, healthcare, and personal welfare of persons who lack capacity. In principle, the Court provides a judicial safeguard: contested best-interests decisions can be brought before a judge who will independently determine the person’s best interests.
In practice, access to the Court of Protection is limited by cost, complexity, and delay. The vast majority of best-interests decisions are made informally by carers, family members, clinicians, and social workers, without judicial involvement. The Court typically becomes involved only where there is a dispute between professionals and family, or where the decision concerns particularly serious matters such as life-sustaining treatment, organ donation, or disputed residence. The ordinary, everyday best-interests decisions that cumulatively shape a person’s life — what they eat, what they wear, whom they see, where they go — are rarely subject to any formal oversight, let alone judicial determination.
Moreover, the Court of Protection’s own approach to best-interests decisions has not always been consistently autonomy-protective. While decisions such as Wye Valley demonstrate a robust commitment to the person’s wishes, other decisions have adopted a more paternalistic approach. In Re M (Statutory Will) [2009] EWHC 2525 (Fam), Munby J (as he then was) acknowledged the importance of wishes and feelings but ultimately determined best interests on a broader welfare basis. The variable weight given to the person’s wishes across different decisions reflects the inherent subjectivity of the best-interests test and the absence of a principled hierarchy among its constituent factors.
The CRPD Challenge: Substituted Judgment versus Supported Decision-Making
The most fundamental challenge to the MCA’s approach to autonomy comes from the United Nations Convention on the Rights of Persons with Disabilities, which the UK ratified in 2009. Article 12 provides that persons with disabilities enjoy legal capacity on an equal basis with others, and that States Parties shall take appropriate measures to provide access to the support that they may require in exercising their legal capacity. The CRPD Committee, in General Comment No. 1 (2014), interpreted Article 12 as requiring the replacement of all substituted decision-making regimes — including best-interests frameworks — with supported decision-making models in which the person’s will and preferences are always the primary consideration.
This interpretation, if accepted, would render the MCA fundamentally non-compliant with international human rights law. The UK Government has not accepted the CRPD Committee’s interpretation, maintaining that the MCA is compatible with Article 12 because it incorporates participatory elements and gives weight to the person’s wishes (UK Government, Initial Report to the CRPD Committee, 2011). Nevertheless, the CRPD Committee’s critique has been influential in academic scholarship. Series (2015) has argued that the MCA’s best-interests framework, despite its subjective elements, remains structurally a substituted decision-making regime because ultimate decisional authority rests with someone other than the person concerned. Clough (2018) has similarly contended that the MCA’s safeguards provide a procedural legitimation of paternalistic decisions rather than a genuine protection of autonomy.
The difficulty with the CRPD Committee’s position, however, is that it does not adequately address the situation of individuals who cannot express any will or preferences — for example, persons in a vegetative state or with the most severe cognitive impairments. In such cases, some form of substituted decision-making appears unavoidable. The question is whether the MCA strikes the right balance, and the stronger view is that it does not. A framework that treated the person’s expressed wishes as presumptively determinative, subject only to a structured override in cases of serious and demonstrable harm, would better protect autonomy while retaining a residual protective function. This is broadly the approach advocated by Donnelly (2016) and is consistent with the spirit, if not the letter, of the CRPD Committee’s interpretation of Article 12.
The Implementation Gap: Safeguards in Theory and in Practice
Even if the MCA’s statutory safeguards were optimally designed, their autonomy-protective function would depend on faithful implementation. The evidence suggests that implementation has been seriously deficient. The House of Lords Select Committee on the Mental Capacity Act (2014), in its landmark report Mental Capacity Act 2005: Post-legislative Scrutiny, found that the Act was not being properly implemented across health and social care. The Committee identified widespread ignorance of the Act among professionals, a tendency to conflate capacity with compliance, and a failure to apply the statutory principles — particularly the presumption of capacity and the requirement to support decision-making.
Subsequent research has confirmed these findings. Series, Fennell and Doughty (2017) found that best-interests decisions in practice were frequently made informally, without the structured application of the section 4 checklist, and that the person’s wishes and feelings were often given token consideration rather than meaningful weight. Williams et al. (2019) reported that care home staff often made decisions about residents’ daily lives without recognising that the MCA applied, and that the concept of best interests was frequently conflated with clinical or institutional convenience.
This implementation gap is not merely a failure of training or resources, although both are significant factors. It reflects a deeper structural problem: the MCA places the duty to assess capacity and determine best interests on the very professionals who are also responsible for delivering care and managing risk. This creates an institutional conflict of interest. The care provider who must decide whether a person has capacity to refuse a particular intervention is also the provider who bears the regulatory, reputational and legal consequences if that person comes to harm. Risk-averse cultures in health and social care therefore tend to err on the side of finding incapacity and determining that the protective course of action is in the person’s best interests (Series, 2015). The safeguards cannot protect autonomy if the institutional incentives systematically favour protection.
Advance Decisions and Lasting Powers of Attorney: Autonomy Through Anticipation
The MCA provides two mechanisms through which individuals can exercise prospective autonomy: advance decisions to refuse treatment (sections 24–26) and lasting powers of attorney (LPA) (sections 9–14). Advance decisions allow a capacitous person to refuse specified medical treatment in advance of losing capacity. Lasting powers of attorney allow a person to appoint a trusted individual to make decisions on their behalf, either in relation to property and affairs or personal welfare, in the event of incapacity.
These instruments are, in principle, powerful autonomy-protective tools. An advance decision to refuse treatment, if valid and applicable, is legally binding on healthcare professionals (section 26(1)). A personal welfare LPA allows the person to choose their own substitute decision-maker, rather than leaving the determination to professionals or the court.
In practice, however, these mechanisms are underutilised. The take-up of advance decisions remains low, partly because many people do not anticipate losing capacity and partly because the requirements for a valid advance decision refusing life-sustaining treatment — that it must be in writing, signed, and verified by a statement that it applies even if life is at risk (section 25(5)–(6)) — create formality barriers. The take-up of LPAs has increased significantly in recent years, but registrations are disproportionately concentrated among wealthier, more legally aware populations (Office of the Public Guardian, Annual Report, 2022–2023). Those who are most likely to need the protection of an LPA — individuals with learning disabilities, older people from lower socio-economic backgrounds, individuals without family support — are least likely to have one in place.
Moreover, even where an LPA exists, it does not guarantee that autonomy will be protected. The attorney is required to act in the person’s best interests (section 9(4)(a)), and the same difficulties with the best-interests test discussed above apply to attorneys as to other decision-makers. There is limited oversight of attorneys’ decisions, and abuse of the LPA is a recognised problem, as the Office of the Public Guardian’s investigation caseload demonstrates. The Court of Protection can revoke an LPA where the attorney is acting improperly, but such cases depend on the abuse being identified and reported, which is often difficult where the person lacks capacity and has no independent advocate.
Reform Possibilities: Towards a More Autonomy-Protective Framework
Several reform options might better protect autonomy within the best-interests framework. First, the MCA could be amended to introduce a rebuttable presumption that the person’s ascertainable wishes and feelings should be followed, with the burden on those seeking to override them to demonstrate that doing so is necessary to prevent serious harm. This would preserve the protective function of the best-interests test while shifting the default position towards autonomy. Donnelly (2016) has developed this proposal in detail, arguing that it is compatible with the MCA’s existing structure and would require relatively modest legislative amendment.
Second, the capacity test could be reformed to introduce a more graduated approach. Rather than a binary determination of capacity or incapacity, the law could recognise a spectrum of decision-making ability and calibrate the degree of support and the level of intervention accordingly. This would better reflect the clinical reality that capacity exists on a continuum and would reduce the autonomy cliff-edge identified above. However, a graduated model would raise significant practical difficulties concerning the legal authority of the decision-maker at each point on the spectrum, and would require careful design to avoid creating new forms of uncertainty.
Third, the procedural safeguards could be strengthened by ensuring that all significant best-interests decisions — not just those concerning serious medical treatment or accommodation — are subject to independent oversight, whether through an expanded IMCA service, a dedicated advocacy role, or a more accessible form of judicial or quasi-judicial review. The resourcing implications would be substantial, but the current position, in which most best-interests decisions are made informally without any independent scrutiny, is inadequate from an autonomy perspective.
Fourth, the implementation of the Liberty Protection Safeguards, or a revised equivalent, should be treated as a priority. The indefinite continuation of the DoLS regime, which has been found not to be fit for purpose, is a significant failure of the safeguard framework. Any replacement scheme should incorporate a stronger autonomy orientation, including a requirement that the person’s wishes and feelings be given presumptive weight in the authorisation of any deprivation of liberty.
Conclusion
The MCA’s safeguards do not properly protect autonomy in best-interests decisions. The Act’s stated commitment to autonomy is genuine but structurally limited. The binary capacity threshold extinguishes decisional authority rather than supporting its graduated exercise. The best-interests test permits the person’s wishes to be overridden without a structured justificatory framework that gives those wishes presumptive weight. The procedural safeguards — IMCAs, the DoLS, and the Court of Protection — are under-resourced, inconsistently applied, and institutionally biased towards protection over autonomy. The implementation evidence confirms that in practice, the MCA operates more as a framework for legitimising protective decisions than as a robust mechanism for safeguarding autonomy.
The most important reason for this conclusion is structural rather than merely practical. Even if the safeguards were perfectly implemented, the Act’s architecture — a binary capacity threshold followed by a best-interests determination in which the person’s wishes are one factor among many — does not give autonomy the primacy that a genuine commitment to self-determination requires. A framework that introduced a rebuttable presumption in favour of the person’s wishes, combined with a graduated approach to capacity and enhanced independent oversight, would more faithfully protect autonomy while retaining the protective function that is necessary for individuals who cannot express any preferences at all. The MCA represents a significant advance over the common law position that preceded it, but its safeguards fall short of the standard required to vindicate the autonomy of those whom it purports to protect.
References
- Mental Capacity Act 2005.
- Mental Health Act 2007.
- Mental Capacity (Amendment) Act 2019.
- Mental Capacity Act 2005 (Independent Mental Capacity Advocates) (General) Regulations 2006, SI 2006/1832.
- United Nations Convention on the Rights of Persons with Disabilities (adopted 13 December 2006, entered into force 3 May 2008).
- UN Committee on the Rights of Persons with Disabilities (2014) General Comment No. 1: Article 12 – Equal Recognition before the Law, CRPD/C/GC/1.
- HL v United Kingdom (2004) 40 EHRR 32.
- Re MB (Medical Treatment) [1997] 2 FLR 426.
- A Local Authority v A [2010] EWHC 1549 (Fam).
- Aintree University Hospitals NHS Foundation Trust v James [2013] UKSC 67.
- PC v City of York Council [2013] EWCA Civ 478.
- Wye Valley NHS Trust v B [2015] EWCOP 60.
- N v ACCG [2017] UKSC 22.
- Re M (Statutory Will) [2009] EWHC 2525 (Fam).
- Clough, B. (2018) ‘New Legal Landscapes: (Re)Constructing the Boundaries of Mental Capacity Law’, Medical Law Review, 26(2), pp. 246–275.
- Donnelly, M. (2016) ‘Best Interests in the Mental Capacity Act: Time to Say Goodbye?’, Medical Law Review, 24(3), pp. 318–332.
- Herring, J. (2020) Medical Law and Ethics. 8th edn. Oxford: Oxford University Press.
- House of Lords Select Committee on the Mental Capacity Act 2005 (2014) Mental Capacity Act 2005: Post-legislative Scrutiny. HL Paper 139. London: The Stationery Office.
- Kong, C. (2017) Mental Capacity in Relationship: Decision-Making, Dialogue, and Autonomy. Cambridge: Cambridge University Press.
- Law Commission (2017) Mental Capacity and Deprivation of Liberty. Law Com No. 372. London: The Stationery Office.
- Series, L. (2015) ‘Relationships, Autonomy and Legal Capacity: Mental Capacity and Support Paradigms’, International Journal of Law and Psychiatry, 40, pp. 80–91.
- Series, L., Fennell, P. and Doughty, J. (2017) The Participation of P in Welfare Cases in the Court of Protection. Cardiff: Cardiff University.
- Department of Health and Social Care (2023) Statement on Liberty Protection Safeguards Implementation. London: DHSC.
- Office of the Public Guardian (2023) Annual Report and Accounts 2022–2023. London: HMSO.
