4.1 Legal capacity

The first element necessary for valid consent is for the consenting actor to be recognised as a person before the law. People with disabilities, like women and people from other marginalised communities, have often been treated as lacking the legal capacity necessary to provide valid consent. This is evident in present-day laws that preclude individuals under adult guardianship or other substitute decision-making regimes from making certain legally binding decisions without the consent of their guardian, including decisions to consent to or refuse medical treatment, to choose where and with whom to live, or to marry (Fundamental Rights Agency, 2013).Footnote ⁴ Many of these current adult guardianship and substituted decision-making laws now focus on the cognitive functioning of an individual to determine whether she can give valid consent. This functional approach prioritises cognition and rationality, and typically involves an assessment of the individual's ‘mental capacity’ before determining whether she can give valid consent in a particular legal context.

Many people with disabilities (especially individuals with intellectual and psychosocial disabilities as well as acquired brain injuries and dementia) are disproportionately required to undergo functional mental-capacity assessments to demonstrate their understanding of decisions that attract legal consequences, whereas other adults are not routinely required to undergo such assessments. In this way, ‘mental capacity’ is often conflated with, or viewed as a prerequisite for, ‘consent’. The functional assessment of mental capacity therefore operates as a preliminary hurdle that an individual has to pass before her consent or refusal will be considered valid. Where a functional assessment of mental capacity can result in an individual being denied the legal agency to consent or refuse a particular act, it should therefore be understood as a denial of the legal capacity of that individual in respect of the decision at hand.

Our proposal aims to ensure that the functional test of mental capacity is no longer used as a determinant of legal capacity in the context of informed consent. Instead, we propose that all persons over the age of majority should be automatically recognised as persons before the law, and no further requirement to demonstrate ability to consent should be imposed.Footnote ⁵ This does not mean that, in situations in which the person was misled or misunderstood the information about the decision, she will be deemed to have provided informed consent. Rather, it simply means that a discriminatory hurdle will be removed, and questions about whether the person expressed clear agreement to a proposed course of action, with the required information about the action, will be considered as part of the evaluation of informed consent, discussed further below.

While the paradigm is shifting towards a recognition of the legal capacity of people with disabilities (including those with complex and significant disabilities), the ramifications of this shift on existing laws that require free and informed consent from legal subjects have not been fully explored. If we begin with the premise that adults with disabilities are legal persons, capable of providing valid consent, then we can examine the kind of nuanced interpretations of will and preferences that can constitute consent, keeping in mind the environmental, socio-political and other factors that might impact on the validity of that consent.

4.2 Freedom to negotiate: communication and agreement

An important prerequisite for valid consent to exist is the freedom of the individual to negotiate their consent or refusal to a particular action. This idea has primarily emerged from scholarship on consent to sex, and relies on Anderson's (Reference Anderson2004) model of ‘negotiating sex’, which emphasises the communicative nature of consent to sex. One of the key elements of this approach is the requirement that the agreement must be an active one. This approach makes a break with previous rape laws that set out a passive and patriarchal notion of consent in which submission or acquiescence (of a woman) to a (male) partner is sufficient grounds to infer consent. The negotiation model rejects the arguments of Schulhofer (Reference Schulhofer1998) and others who say that consent can be inferred from behaviour – particularly because, as Anderson states:

Anderson (Reference Anderson2004) argues that the duty to negotiate agreement rests with the person who wants to penetrate his or her partner. While this focus on penetration is still problematic in the context of rape and sexual offences law, the notion of requiring the person who wants to act upon another's body to commence a process of active communication and listening is a useful one for the context of medical treatment. If this model is extended beyond rape law into consent to treatment, it requires the person seeking to interfere with another's bodily integrity to commence the negotiation. Communication is key to negotiating such an agreement. Anderson has a preference for verbal communication between partners who have not previously established a form of non-verbal communication. This preference is clearly grounded in the notion that verbal communication is easiest way for a person to understand whether their partner agrees to a particular action. This, however, poses problems from a disability perspective. Nevertheless, as Arstein-Kerslake (Reference Arstein-Kerslake2015) has previously written, the requirement for verbal communication can be re-imagined from a disability perspective to consider the positions of those who use non-verbal methods as their primary or sole form of communication, while maintaining a preference for communicative methods that are objectively verifiable by others wherever possible.

The positive obligation to initiate negotiation and communication to reach an agreement helps to place the final ‘agreement’ in context. In the language of the CRPD, the communication process in this negotiation might be considered a means of exploring the individual's ‘will and preferences’, which the CRPD Committee (2014) maintain must be at the heart of an exercise of legal capacity. A negotiation also implies that a range of options are open for discussion – and that both parties are considered legal equals in the encounter, meaning that refusals will be fully respected.Footnote ⁶ A full consideration of the context in which consent is sought is therefore needed.

In the context of rape law, Palmer (Reference Palmer2013) suggests that a framework of ‘freedom to negotiate’ would not simply consider whether the person ‘agreed’ to sex, but whether they were free to negotiate in the absence of a fear of violence, financial or other dependence and power dynamics. A similar approach could be applied in the context of medical treatment to ensure that ‘mere agreement’ is not taken as evidence of free and informed consent. Agreement to a particular course of medical treatment could be obtained in an environment that is fraught with coercion and this kind of agreement should not be considered valid consent. In light of the power dynamic that exists between doctor and patient and the pressure to choose a particular course of treatment, patients may not have freedom to negotiate consent or refusal. Therefore, in order to respect the right to bodily integrity, the individual must be the ultimate decision-maker on questions of medical treatment. Further, the concept of ‘negotiation’ reflects the centrality of an active communicative process between individuals – without which no negotiation can take place, and no agreement can be reached.

4.3 Information, expression of will and preferences, and voluntariness

Once the individual is recognised as possessing legal capacity, and we can establish that the consent or refusal is occurring in a context in which the individual has ‘freedom to negotiate’ and is free from coercion or the threat of coercion, the next step is to establish that the individual has sufficient information about the action she is being asked to consent to or refuse. In the context of medical treatment, this would require the treating physician to disclose all possible side effects of treatment, which would require far greater disclosure (e.g. about the side effects of proposed medications) in a mental health treatment context than that which currently occurs (Morrison et al., Reference Morrison, Hutton, Shiers and Turkington2012).

Once the relevant information has been provided, in a manner accessible to the person, and the person has been offered support to exercise legal capacity in this context if desired, then the person proposing the relevant course of action must wait until the person has expressed her will and preferences on the matter before any further action can be taken. In keeping with General Comment 1 (UNCRPD Committee, 2000), we consider that, where the person's will and preferences are unknown, every effort to discover them must be made using all possible forms of support and communication. If a person's will and preferences remain unknown after all efforts to discover them have been made, a supporter may have to determine the ‘best interpretation’ of the person's will and preferences in that context to decide, for example, whether the person wishes to consent to, or refuse, medical treatment.

It is particularly important that a supporter, trusted and chosen by the person, undertakes this task, and that it is not undertaken by the health care professional who proposes to treat the person, as this creates a clear conflict of interest. Further, we believe that there are limits on when a best-interpretation approach can be used to determine will and preferences in the context of consent to medical treatment. This approach should only be used where all other options have been exhausted and where a decision about treatment urgently needs to be made. There are certain kinds of medical decisions that in our view should never be subject to best interpretation – especially where the decision is irreversible or has a significant detrimental impact in the long term on the person's life. These decisions include sterilisation and psychiatric treatment. Given the consequences of undergoing these treatments, we believe that they should not be administered without the person's clear expression of their will and preference to consent to such treatment. In the context of decisions to consent to sex, we also do not believe that a best-interpretation approach would be appropriate – in these situations, a clearly communicated agreement to engage in sexual activity must be present for valid consent to exist.

In life-threatening situations in which no communication is possible with the person, such as where a person is found unconscious by paramedics, we believe that medical intervention can be justified where it would be proportionate in the circumstances and is necessary to preserve the person's life, as set out by Flynn and Arstein-Kerslake (Reference Flynn and Arstein-Kerslake2017) in this issue. However, in this situation, we presume that, while the person has not provided consent for the intervention (due to unconsciousness), no desire to refuse medical treatment is indicated (e.g. the person has not previously made a ‘do not resuscitate’ order that the paramedics have access to). If there is any indication that the person is refusing treatment, this must be respected. Once the immediate risk to the person's life is addressed (e.g. where resuscitation revives the person), consent must be sought for any further medical intervention. If, following the provision of support to exercise legal capacity, it remains impossible to establish communication with the person and further medical treatment is offered (e.g. surgery on the person's wounds), then, in accordance with General Comment 1, we suggest that a supporter may be needed to make a best interpretation of the individual's will and preferences that can be used to guide any further treatment decisions.

We acknowledge that a potential difficulty here is determining whether a particular medical emergency is ‘life-threatening’ and whether the treatment offered in that situation is a necessary and proportionate response to the perceived risk to the person's life. It is important to reiterate here that we do not accept that a person perceived by medical practitioners to be in a state of extreme emotional distress should be forcibly sedated, as this does not constitute a proportionate response to any perceived threat to the individual's life. This position is further explored and justified elsewhere in this volume (Flynn and Arstein-Kerslake, Reference Flynn and Arstein-Kerslake2017). For the purpose of this paper, therefore, our position is that, once the individual has expressed her will and preferences or where the person's will and preferences remain unknown, once a supporter has arrived at the best interpretation of that person's will and preferences, the role of the health care provider in the context of medical treatment is simply to either respect the individual's refusal or, if she thinks the individual is expressing a will and preference to undergo treatment, to take steps to verify that this agreement is a voluntary one.

Our focus on information, will and preferences, and voluntariness for consent must be distinguished from the emphasis on rational decision-making from the liberal conception of autonomy described above. These liberal approaches often result in assessment of the person having, or not having, ‘insight’ into the decision to be made – for example, into the need for medical treatment. Medical practitioners’ assumptions about lack of insight can inhibit meaningful dialogue, exaggerating inequalities between people's experiential knowledge and the scientific knowledge base of practitioners (Morant et al., Reference Morant, Kaminskiy and Ramon2015).

Diesfeld and Sjöström (Reference Diesfeld and Sjöström2007) illustrate how the construct of ‘insight’ is used to justify continued detention of people with psychosocial disabilities. Patients’ factual accounts that contradict expert opinion are not treated as countervailing reports, but instead become evidence of symptoms. In the absence of concrete evidence, claims regarding lack of insight are a convenient blanket assertion to address credibility issues. Diesfeld and Sjöström (Reference Diesfeld and Sjöström2007) commented on the circular logic: (1) lack of insight causes non-compliance; (2) non-compliance is evidence of lack of insight. When clinicians claim the person lacks insight (and by implication will not comply with treatment), that assertion acquires medical certainty. Thus it undermines the person's credibility. Failure to comply with medication is taken as an indication of lack of insight, while lack of insight substantiates the claims about non-compliance.

Morgan and Veitch (Reference Morgan and Veitch2004) further critique the notion that the purpose of mental-capacity assessments is to verify the individual's actual understanding of the decision to be made. They argue that:

In support of this argument, Haidt (Reference Haidt2001) and Iyengar and Lepper (Reference Iyengar and Lepper2000) have demonstrated that individuals only conjure up reasons for their decisions when called upon to do so, and these reasons rarely correlate with their actual decision-making process at the time of the original decision, but rather reflect the most persuasive explanation the person can find for their decision. A disability perspective can challenge this form of post-hoc rationalisation where judgments are based on socially intuitive moral reasoning – especially for people with disabilities who communicate differently or express reasons for their decisions that others find hard to understand or empathise with.

However, our approach, which would eliminate mental-capacity assessments as a precursor to determining consent, would also eliminate the use of ‘insight’ as a basis on which to grant or deny the opportunity to consent or refuse treatment. In our view, the law has sufficient tools to determine whether free and informed consent was actually present at a particular moment, based on the discovery of will and preferences, without requiring people with disabilities to first demonstrate their decision-making ability before we consider the validity of their consent, as we will demonstrate in the following section.

4.4 A practical example: the case of Mr B

Here, we will consider how our approach could be applied in practice, using the example of an existing case, Wye Valley NHS Trust v. Mr B.Footnote ⁷ In this case, the man in question, Mr B, was seventy-three years old and had type II diabetes. While he did not belong to any particular religion, he had strong spiritual beliefs, and told the judge that he heard the voice of angels, and that the Virgin Mary spoke to him in a vision and urged him to become a Catholic. He had been subject to forced psychiatric treatment on several occasions. He developed a foot ulcer as a result of diabetes that did not heal despite various interventions. In hospital, he refused medication and antibiotics for his foot and his condition rapidly deteriorated. His medical team explained to him that his foot now needed to be amputated, but he refused to consent to the procedure, although he did allow the wound to be cleaned and dressings changed. Various doctors reported that at different times, Mr B had said that his spiritual beliefs informed his decision to refuse treatment.

The hospital sought a declaration that he lacked mental capacity to consent to an amputation and that an amputation should be authorised against his will, in his best interests. When Jackson J. met with Mr B, the patient told the judge that he believed his leg would get better with proper care and if he was allowed to use it. He also said:

Jackson J. determined that, while Mr B did not have the necessary mental capacity to consent to surgery, he did not think that forcing him to undergo surgery against his will was justified in his ‘best interests’. While this decision represents a good example of an effort to promote autonomy within a discriminatory paradigm of mental-capacity testing, under our proposal, the same outcome would have been reached without the need for a stigmatising and discriminatory labelling of Mr B as lacking the mental capacity necessary to consent.

Under our approach, the first task of medical team would be to make every effort to inform Mr B of his treatment options, likely benefits and side effects, in a manner that was accessible to him and respectful of his preferred form of communication. It appears from the judgment that this was done – although no formal supports were offered to Mr B to exercise his legal capacity in this respect – and it appears that he had no informal support network who could have provided assistance to him, if he desired it, in reaching his decision. Under our proposal, the state would have an obligation to provide such support if Mr B required it to exercise his legal capacity. Mr B made several unequivocal communications of his will and preferences in respect of the proposed surgery, as is evident from his statements to the judge. Therefore, no further inquiry is needed into whether Mr B truly understands the decision he is making – once all efforts to provide information about the decision and support to exercise legal capacity have been made.

The final step in this process is a confirmation that Mr B's decision was a voluntary one – made in the absence of coercion. While Mr B's spiritual beliefs may have influenced his decision, this is true of many treatment decisions that are made, and the reliance on a religious belief in making a medical decision should not be viewed as amounting to coercion that would vitiate his consent. It is clear from the facts that he was under pressure from the medical team to accept treatment, but he vigorously resisted this pressure, and was not subject to undue influence from any other source in reaching his decision. Under our proposal, therefore, Mr B's refusal of treatment would have been respected, and he would not have had the humiliating experience of being brought to court and declared to lack capacity before his wishes would be respected.

This example demonstrates that Article 12 CRPD requires us to go deeper to discover, to the extent possible, the will and preferences of the individual, which can be used as a basis for determining whether that person wishes to consent or refuse. If the person's will and preferences are clear, and the person is clearly consenting or refusing, using their chosen form of communication, then the legal position post-CRPD is relatively clear. All that is required in this scenario is due diligence by those seeking consent. For example, in a treatment context, medical professionals should ensure that the person has the necessary information in a format they can understand, and is given sufficient time and support to reach a decision. This case indicates the role that supported decision-making such as by trauma-informed, intentional peer support (Mead and Hilton, Reference Mead and Hilton2003; Mead, Reference Mead, Stastny and Lehmann2007) can play in establishing the person's will and preference. Once a clear will and preference emerges, the law should simply require respect for the person's decision, even if the decision seems irrational, ill-informed, non-compliant or unwise. However, where it is impossible to determine an individual's will and preferences at the moment at which consent would usually be required – for example, where a person is found unconscious by paramedics – then treatment to preserve the person's life, health or safety can of course be provided; and practitioners in this instance would be protected from liability under the doctrine of necessity (Gooding and Flynn, Reference Gooding and Flynn2015). In this scenario, once the person returns to a position in which she can once again express her will and preferences, consent must be sought for any further treatment.

As authors in this Special Issue have written previously, where a person's will and preferences are conflicting– for example, where a person experiencing anorexia expresses a will to live but a desire not to eat – then intensive support may be required to ascertain what treatment that person wishes to consent to or refuse (Flynn and Arstein-Kerslake, Reference Flynn and Arstein-Kerslake2014). Advance directives that include Ulysses clauses can also provide helpful support in these difficult treatment decisions, if they do not directly or indirectly discriminate in their application against people with psychosocial disabilities (Morrissey, Reference Morrissey2012). At the end of this process, the person may still express a clear refusal of certain forms of treatment, which may even result in death, and, in line with our previous argument, this refusal should be respected. We do not make this argument lightly, and acknowledge the challenge that respecting such refusals of treatment presents, but remain convinced that treatment without consent in these situations is not a solution that adequately respects the human rights of the individual.Footnote ⁹

Our proposal is to integrate questions of information, will and preferences and voluntariness into a determination of the presence or absence of consent. This is an alternative to both the functional assessment of mental capacity undertaken on specific population groups prior to considering consent and status approaches to those considered ‘unable to consent’ on the basis of a label of disability. This may seem to some a mere semantic or symbolic difference from the current legal framework. However, we view it as an important indicator of the paradigm shift (Kayess and French, Reference Kayess and French2008) from viewing people with disabilities as objects to be manipulated to subjects and equal citizens, with equal rights and responsibilities. There is certainly a risk that, within our proposed approach, which is prima facie non-discriminatory, people with disabilities and those with cognitive disabilities in particular may continue to be disproportionately viewed as having will and preferences that are unclear or unknown.

However, the difference with our proposal is that, if the person is not giving free and informed consent, the response is not to remove their legal capacity so that a proxy can consent on their behalf, or force them to undergo treatment against their will. Rather, we argue that, in the context of medical treatment, the person who proposes to administer treatment must provide all of the necessary information to the person about the treatment in a manner that is accessible to the person and respects her chosen form of communication. This may include providing information to the person's appointed supporters, whose role may be legally recognised in a binding support agreement. Once the person has subsequently expressed will and preferences regarding the decision, if the person indicates a refusal of treatment (including where this refusal is communicated via supporters), this must be respected. If the person indicates agreement with the proposed course of action, then, if the treating professional considers in good faith that the person is agreeing voluntarily and has all of the information required to give consent, the agreement should be regarded as valid consent.

5.1 Coercion

The finding of the presence of coercion invalidates most legal agreements. Coercion can be evident in the overt form of practices such as restraint (legal, mechanical or chemical) but, more insidiously, coercion can be latent or unspoken, existing in threats of a range of measures including force, deprivation or encroachments on dignity and autonomy (McSherry and Freckelton, Reference Mcsherry and Freckelton2013). Informal coercion (persuasion, inducements, threats and force) has been widely observed operating in many in-patient mental health units (Lidz et al., Reference Lidz, Mulvey, Hoge, Kirsch, Monahan, Eisenberg, Gardner and Roth1998). Sjöström (Reference Sjöström2006) describes the ‘coercive context’ whereby coercion is invoked to control people and maintain order on hospital wards. A ‘coercive invocation’ may be the overtly locked doors or covert delaying tactics by nurses to distract and deter voluntary patients from leaving a unit. An Australian study on the perspectives of people who have experienced seclusion and restraint, and of their family, friends and supporters, describes the harms and chilling effect of these practices (Brophy et al., Reference Brophy, Roper, Hamilton, Tellez and MCSHERRY2016). They use the term ‘emotional restraint’ to describe the pressure to comply with behavioural expectations within mental health services. ‘Emotional restraint’ silences people and inhibits them from expressing their views to staff for fear of the consequences. Therefore, it is unsurprising that people using mental health services frequently report care as coercive and controlling, without choices either to refuse care, about what is considered care or over who provides this care (Newbigging et al., Reference Newbigging, Ridley, McKeown, Sadd, Machin, Cruse, Dela Haye, Able and Poursanidou2015). Minkowitz (Reference Minkowitz2007) goes further to describe the use of formal and informal coercion in psychiatry as an exercise of violence under the auspices of the state. Where informal coercion occurs within these contexts (often underpinned by formal legal coercion that supports forced treatment), this would, under our proposal, mean that no valid consent is given.

Likewise, coercive practices have been reported in residential care of older people and of people with intellectual disabilities. Chemical and physical restraint are widespread practices to control challenging behaviour in under-resourced and oppressive institutional settings where people with intellectual disabilities and dementia live (Curkovic et al., Reference Curkovic, Dodig-Ćurković, Erić, Kralik and Pivac2016; HSE, 2006). In a study of Swedish care services for people with intellectual disabilities, the authors found high levels of violence. Compulsory medication, along with force and seclusion, were used as means to restrain (physically or chemically) people who were not cooperative with staff (Strand et al., Reference Strand, Benzein and Saveman2004). In such situations, an atmosphere of fear is the antithesis of conditions necessary to foster real consent.

Increasingly, community mental health services in the US and UK are using coercion to ensure that people adhere to treatment (Burns et al., Reference Burns, Yeeles, Molodynski, Nightingale, Vazquez-Montes, Sheehan and Linsell2011; Dawson, Reference Dawson2005; Dunn et al., Reference Dunn, Maughan, Hope, Canvin, Rugkåsa, Sinclair and Burns2012; Zigmond, Reference Zigmond2011). Rendering essential supports such as access to welfare income and accommodation services contingent on adherence to prescribed treatment removes any notion of free, informed consent. The unifying theme in this work, linking back to consent, is that all definitions of consent include the condition that valid consent cannot be said to exist in the presence of coercion.

5.2 Undue influence

Feminist jurisprudence asserts that consent to sex under undue influence is not true consent. In legal terms, undue influence is understood as any act of persuasion that overcomes the free will and judgment of another (Lehman and Phelps, Reference Lehman and Phelps2005). If undue influence is found by a court to have been present at the time that instructions for a will were given (e.g. to an attorney) or when it was executed, it has the effect of rendering a will invalid.

Exhortations, importunings, insinuations, flattery, trickery and deception are all examples of undue influence. Talcott Parsons (Reference Parsons1963) described four types of influence that someone can exert over another: an appeal to shared values, inducement, persuasion and force. An appeal to shared values alone is not generally seen to be likely to constitute undue influence, but inducements (such as payments, early release from confinement, credits for courses or other favourable treatments) to participate in sex or medical treatment may constitute undue influence. In general, the greater the inducement and the participant's need for what is offered them, the less voluntary the choice to participate is regarded. In the case of persuasion, the relationship and position of the persuader vis-à-vis the participant has a significant bearing.

If someone is not beholden, they are more likely to be able to resist persuasion but, when there is an unequal power relationship, as in a child/parent, patient/doctor or student/teacher relationship, then certain kinds of persuasion have been determined to constitute undue influence. This is particularly the case for people dependent on care-givers who may not be in a position to refuse inappropriate demands made on them by their care-givers. For instance, literature on elder abuse is replete with instances of undue influence being brought to bear on isolated and vulnerable older people (following spousal bereavement or other losses) by people in positions of trust (Quinn, Reference Quinn2000). Equally, decision-making by people with intellectual disabilities is frequently shaped by the undue influence their carers wield in mediating people's environments, and the choices made available to people (Jenkinson, Reference Jenkinson1993).

Undue influence (and indeed coercion) are aspects of the interpersonal power dynamics that shape and contribute to occasions of oppression in the realm of consent and decision-making for people with developmental and psychosocial disabilities. Work by Björnsdóttir and Traustadóttir (Reference Björnsdóttir and Traustadóttir2010) illustrates how a complex web of intersecting oppressive structures – gender, class, religion and disability – play a role in shaping and constricting the life choices available to people with intellectual disability in Iceland. In the same vein, many of the same macro social structures limit occasions of respectful negotiation of consent. One of the most pervasive aspects of macro structures that indirectly impacts on free and informed consent are the power imbalances present between parties.

5.3 Power imbalances

Many feminist and other critical scholars move their focus beyond the interactions between individuals to consider the social factors that impinge on and shape individual instances of consent or refusal. For instance, Catherine MacKinnon (Reference Mackinnon1991; Reference Mackinnon1997) argues that, given the societal structures of patriarchy that give men so much power over women, women are not usually in a position to freely consent to sex with men at all. In this context, sex is a feature of men's expectations from women – part of what men feel is due them from women. Indeed, early feminist work on rape asserts that it is not motivated by sexuality so much as violence and hatred of women (Brownmiller, Reference Brownmiller2013).

Susan Stefan (Reference Stefan1992) highlights the power dynamics operating in the determination of competence, which is used to negate the individual's ability to consent. She argues that it is always the dominant professionals who determine the conditions under which the person is assessed and which define the outcome, and that ‘questions of competence arise only as a function of a relationship between two or more people and that this relationship is necessarily a hierarchical one, characterized by dominance and subordination, by power and powerlessness’ (Stefan, Reference Stefan1992, p. 766). Bridget Hamilton and Cath Roper also problematise the ‘power relations between health professionals, who predominantly define what is truth and good practice in mental health, and the people upon whom public psychiatric treatment is visited, commonly without their consent’ (Hamilton and Roper, Reference Hamilton and Roper2006, p. 416).

Lukes and, later, Gaventa present a critical account of power in relation to decision-making fora: work that is applicable to consent to medical treatment (Gaventa, Reference Gaventa2006; Lukes, Reference Lukes1974). Gaventa (Reference Gaventa2006) describes three forms of power: visible, hidden and invisible. Visible power determines the agenda for decision-making, including the ‘who, what and how’ of decision-making. However, the other forms of power are often over-looked when considering how power influences relationships between professionals and lay people. Hidden power is held by the agents who can influence what is or is not allowed onto the agenda, what knowledge is valid for consideration but, more importantly, who is not allowed to the decision-making table. The final form of power is the invisible ‘shaping of the psychological and ideological boundaries of participation’ (VeneKlasen and Miller, 2002, cited in Gaventa, Reference Gaventa2006, p. 29). Not only are issues kept off the table for discussion, but people come to believe that the powerful are acting in their interests, and fail to see how things could be otherwise arranged. Lukes (Reference Lukes1974) asserted that this third face of power is most successful when those subjected to it do not recognise its operation or indeed consider that the status quo is in their best interests.

Masterson and Owen (Reference Masterson and Owen2006) apply Lukes's (Reference Lukes1974) theory of power to the mental health field, and illustrate it using the example of decision-making around medication. The traditional Weberian understanding of power is apparent at the first face of power: here the authority of a psychiatrist is very visible. In addition, though, covert use of power controls the agenda of what can be discussed. If both the psychiatrist and the service-user were to discuss the relative merits of different drugs, with the final decision resting with the service-user, this can be considered an equal exercise of power at the level of covert (hidden) power. What is more common, however, is that the discussion precludes that the person could cope without medication. Another aspect of such hidden power is the omission of potential adverse effects of medications and an informed ‘cost–benefit risk’ discussion. These issues are not on the agenda for discussion, as the psychiatrist sets the agenda and uses covert power to pre-empt the decision-making process.Footnote ¹⁰ The fact that non-medical explanations of psychosocial distress are not considered, plus the legal power to force treatment, is an example of the third form of power: the latent or invisible power operating the hegemony of psychiatry (Brosnan, Reference Brosnan2012).

Health professionals are not the only sources of power imbalance in the context of determining valid consent in law. In all legal contexts in which consent is required, there may be power dynamics at play – including the power dynamics between a patient and their family members and service providers in the context of medical treatment. We cannot expect the law on consent to require an elimination of all power imbalances in these contexts. For example, the person proposing a particular form of medical treatment will almost always have more power in these encounters. Nevertheless, laws on consent can require these actors to take steps to minimise the power imbalance or, at a minimum, be conscious of the power imbalance in determining whether undue influence or coercion might be present to the extent that it negates consent.