The Separate Grounds of Competence 

ii.a demanding common grounds views

Common Grounds views can be divided into those that are more demanding and those that are more liberal.

Most writers on competence to consent in the medical field adopt a relatively demanding view. Take, for example, the “four abilities” model that is widely accepted in the United States and operationalized for capacity assessments through instruments like the MacArthur Competence Assessment Tool-Treatment (MacCAT-T).⁸ Following this model, someone is competent to consent if and only if they have: (i) the capacity to communicate a choice; (ii) the capacity to understand relevant information; (iii) the capacity to appreciate their situation and the consequences for them of different choices; and (iv) the capacity to reason about their options.

In the context of medical care and clinical research, the four abilities model implies that many cognitively impaired individuals cannot decide for themselves. To illustrate, consider Medhi, a man with moderate dementia. Medhi has problems holding new information in his short-term memory, he sometimes fails to recognize familiar people, and he has trouble following through on decisions if they involve multiple steps. Medhi has recently been diagnosed with leukaemia, and among the options now available to him is a phase 1 chemotherapy trial, which would provide him with standard of care treatment but add on an experimental drug that is being evaluated for toxicity. Medhi takes great pride in helping others, but he cannot grasp several important aspects of the clinical trial. For example, despite repeated explanations, he does not understand that the primary purpose of the trial is to explore what this experimental drug does to the human body; he is convinced that it will cure his illness. He is also unable to recall the risks and the potential benefits of the options of trying further standard care versus enrolling in this trial. The four abilities model would find Medhi incapable of consenting to a risky Phase 1 clinical trial as he does not have the capacity to understand or appreciate the reasons for or against participating in the study.

It seems quite reasonable that Medhi should not be allowed to decide for himself about trial participation. But in other contexts this model appears overly demanding. Suppose that Medhi wants to donate blood. He has been a blood donor for many years, and he understands that the procedure involves inserting a needle into his arm and taking out blood to give to other patients. His local blood donation center now collects plasma, so Medhi is invited to donate plasma instead, if he prefers. His nurse explains what plasma is, what it is used for, how the procedures differ, and how blood donations still involve giving plasma but not as much as this new procedure. The explanation is clear, but Medhi cannot retain the difference between blood and plasma donations for long enough to compare the relative value of doing one versus the other. He just wants to give blood, but he is no longer capable of engaging in the kind of complex comparative assessment of alternatives that would be required to meet the appreciation or reasoning conditions. The four abilities model would therefore imply that Medhi is not capable of giving or refusing consent to blood donation; someone else would need to authorize the procedure on his behalf. This is counterintuitive.

This verdict reflects a general problem with demanding Common Grounds views: they often fail to meet the first desideratum of rendering the intuitively correct judgement in cases in which few would disagree about the right answer. This can be seen even more starkly if we consider some cases outside of medicine, such as how we treat decisions made by children. Consider the acts of purchasing a toy or agreeing to a request for a hug. Both involve exercising rights and so require competence. A baby cannot do either while—we think—a typical eight-year-old can do both. Yet that eight-year-old does not have the capacities needed to make most decisions according to the four abilities model.⁹

ii.b liberal common grounds views

In response to the counterintuitive implications of demanding Common Grounds views, one might be tempted by a more liberal approach to competence.

One such approach is suggested by Agnieszka Jaworska.¹⁰ She writes:

The capacity for autonomy [which Jaworska regards as the capacity underlying the right to make one’s own decisions] is best understood not as the ability to lead one’s life by one’s lights when one is left to one’s own devices, not as a full capacity to make a decision from the beginning to end, but as the capacity to value.¹¹

Jaworska thinks that the cognitive requirements for someone to have the capacity to value are relatively minimal, which means that someone can have normative standards for themselves despite substantial cognitive impairments. She identifies three marks of valuing:

the person thinks she is correct in wanting what she wants; achieving what she wants is tied up with her sense of self-worth; and the importance of achieving what she wants is, for her, independent of her own experience.¹²

Jaworska thinks that many persons with dementia retain the capacity to value even when they have lost other important cognitive capacities, such as a grasp of their life’s narrative, or the ability to recognize previously familiar people. She provides extensive empirical evidence to support this claim. Since valuing underlies autonomy, she concludes that we often have reason to respect the decisions of those with moderate dementia, even if they sometimes need help in carrying them out:

the essence of the capacity for autonomy consists in the ability to lay down the principles that will govern one’s actions, and not in the ability to devise and carry out the means and plans for following these principles.¹³

Liberal Common Grounds views can render intuitively correct verdicts in one set of cases—those in which we judge that the individual is clearly competent. For example, Jaworska’s view will find most mentally mature adults competent to consent. Further, it can explain why: most mentally mature adults are valuers. Furthermore, in some cases of borderline competence, a liberal view will give the intuitively correct result. For example, it will judge that Medhi can give consent to donate blood and—outside the medical field—that an eight-year-old can trade pocket money for a toy or agree to a hug.

The challenge for liberal Common Grounds views is that they appear inclusive to the point of negligence. For example, since Medhi is a valuer, it seems as though Jaworska’s view would imply that he should be allowed to decide for himself about phase 1 clinical trial participation (albeit perhaps with considerable outside help in his decision-making). This, as we have noted, is implausible. Liberal Common Grounds views fall at the same hurdle as their more demanding counterparts: they fail to meet the desideratum of rendering the correct verdict in cases where it is intuitively clear.

We can derive a critical insight from the forgoing analysis. More liberal Common Grounds views focus on protecting the exercise of autonomy when it is present. They prioritize the power to consent, but at the cost of allowing people with impaired decision-making capacity to make potentially dangerous decisions. More demanding Common Grounds views protect people who lack decision-making capacity from harm, but at the cost of not allowing individuals to exercise the limited autonomy that they have. Common Grounds views struggle to balance these goals when dealing with individuals who have marginal autonomy—leading to verdicts that are counterintuitive in one direction or the other.

ii.c risk-adapted common grounds views

One might object that the counterintuitive implications we describe arise only because we treat the threshold for competence determinations as fixed. It is natural to think instead that judgments of competence should be sensitive to the level of risk that a decision involves. This is the strategy followed by Allen Buchanan and Dan Brock. They write:

No single standard for competence is adequate for all decisions. The standard depends in large part on the risk involved, and varies along a range from low/minimal to high/maximal. The more serious the expected harm to the patient from acting on a choice, the higher should be the standard of decision-making capacity, and the greater should be the certainty that the standard is satisfied.¹⁴

On this view, Medhi probably does have the capacities necessary to give valid consent to the low-risk blood donation, but not the capacities to refuse medically indicated, life-sustaining treatment in favor of a risky and potentially life-threatening experiment. It seems that if we allow the threshold for competence to vary depending on the stakes of the decision, then we can avoid the counterintuitive implications described above, and yet the right to consent for oneself and the right to demand that other people respect one’s consent decisions can still have the same basis.

But risk-adapted Common Grounds views also face serious challenges. Where non-risk adapted views fail to meet the first desideratum, risk-adapted ones fail to meet the second. They cannot offer plausible explanations of their verdicts because their analysis has little to do with competence itself.

Competence is about one’s ability to do something. Since some things are harder to do than others, it is plausible that competence is task-specific. Someone can be competent to juggle three balls but not four, or competent to juggle but not to somersault. But one’s ability to do a task does not seem to vary with how risky the task is. It is not that juggling four balls is more dangerous, but that our juggler keeps dropping the balls, which makes her unable to juggle four even though she can juggle three.

The point generalizes to one’s ability to make decisions. Some decisions are more difficult to make than others, but they are not more difficult in virtue of the harms associated with them, but in virtue of how complex the considerations are on either side.¹⁵ Moreover, there are many examples of decisions with high-risk options that are simple and examples of decisions with low-risk options that are complicated.¹⁶

Consider a young woman with mild cognitive impairment who has been diagnosed with acute appendicitis. Immediate appendectomy is recommended. For simplicity, assume that her options are to agree to or refuse the surgery. Agreeing is a low-risk choice; declining is a very high-risk choice. Suppose further that her cognitive impairment is such that she is judged incompetent to consent to the life-saving operation due to the great harm associated with refusal. Nevertheless, she is judged competent to make low-risk medical decisions for herself, such as whether to undergo an additional X-ray to confirm her diagnosis. The decision about surgery is straightforward. Whenever asked, she affirms that she wants to receive the surgery because she thinks it will save her life. The X-ray decision is complicated, since it involves weighing the small chance of catching a misdiagnosis and then the risks and benefits associated with that result against the very small long-term increase in cancer risk and the co-pay her health insurance will require. In fact, despite lengthy discussions, she is unable to understand the long-term risks or make comparative assessments of the options available to her. When asked though, she says that she wants to do the X-ray too, because it sounds like a good idea. In fact, let us suppose, given her values and given the very low probability that the X-ray will provide extra information that changes the course of clinical care, it would be a prudential mistake for her to undergo the X-ray.

This patient is able to understand what is at stake in the high-risk decision, but not in the low-risk decision. Further, she makes the prudentially correct decision for the right reasons in the high-risk case, but not in the low-risk case. In what sense, then, is she able to make the latter decision but not the former? It is because risk-adapted views do not give us a plausible analysis of what competence is that they cannot explain the difference our intuitions track here: namely, that some decisions with high-risk options are simple, whereas other low-risk decisions are complex.

A proponent of a risk-adapted view might deny their analysis renders an ad hoc verdict here. They might suggest that the reason why the patient is able to make the high stakes decision is because this is a choice with asymmetric risks.¹⁷ Agreeing to the surgery is a low-risk choice; declining is a very high-risk choice. On a risk-adapted view, it therefore seems as though the threshold for the patient to be competent to refuse the surgery should be higher than the threshold for her to be competent to agree to it.¹⁸ Some patients will therefore be judged competent to agree but not competent to decline. For example, perhaps someone who is very poor at taking long-term harms into account might be judged competent to agree but incompetent to refuse surgery.

This is also a counterintuitive result. It seems odd to say that a patient is capable of agreeing but not capable of refusing the surgery. If she is capable only of agreeing, then in what sense does she have the ability to make a decision? And, in the context of consent, in what sense would her ability to make a decision ground a right to make that decision, given that there is only one choice she can make that merits being respected by others?¹⁹

No plausible answers are forthcoming from risk-adapted accounts. They would be able to explain this counterintuitive verdict if the high risks associated with refusal were generally harder to comprehend, appreciate, or reason with than the low risks of acceptance. However, as we have shown, risk no more tracks complexity than complexity tracks risk. Since there is no reason to think that the threshold for competence should rise as risks increase, risk-adapted Common Grounds views are ad hoc.

ii.d summary

We have argued that liberal Common Grounds views are too liberal, that demanding Common Grounds views are too demanding, and that risk-adapted Common Grounds views fail to capture what competence is. A natural response to this critique might be to say that we have merely demonstrated the need to develop a non-risk-adapted, moderate Common Grounds view. This would be a view that is not quite as liberal as the one suggested by Jaworska nor quite as demanding as Grisso and Appelbaum’s.

This is a reasonable response. We have argued that the component rights that comprise the kind of competence implicated in consent transactions require justification. There is either one justification from which the right to waive rights through consent and the right against paternalistic interference can be derived or each has a separate justification. We have argued that existing views that attempt to derive both rights from a single ground fail to meet at least one desideratum of a plausible account of competence. In addition, we think there are compelling and independent reasons to believe that these component rights have separate normative foundations. We therefore leave articulating a more compelling moderate Common Grounds view to its proponents, and turn now to those reasons. Once a Separate Grounds view is outlined, it should be clear why this alternative is preferable to an imagined non-risk-adapted, moderate Common Grounds view.

iii.a what capacities ground the power to waive a right through consent?

In answering this question, it helps to consider why being able to give (and refuse) consent is valuable to beings like us. A natural answer lies in how it facilitates autonomous action: an individual who can give consent is thereby able to direct their own life by changing their normative relationships to others. Moreover, directing one’s life in this way is often very valuable. It allows us to pursue important human goals, including those related to intimate relationships, personal property, and medical care. In brief, being able to direct one’s own life by giving or withholding consent is prudentially good. Each of us therefore has an interest in realizing this good. We think that the strength of that interest is sufficient to ground the power that allows individuals to realize that good.

Before we can derive the capacities that ground the power to waive a right from this insight, a clarification is in order.²⁰ There is a specific good of directing one’s own life through autonomous action. It is distinct from the good of directing one’s own life in the sense of simply getting what one wants. Having one’s desires frustrated is, along at least one dimension of wellbeing, typically bad for one. For example, it is probably somewhat bad for a toddler who is prevented from making a noise with an annoying drum that she is frustrated in doing what she wants. The good that we are describing is different. Directing one’s life on the basis of reasons derived from one’s values is a distinctive good, and it is that good that the power to consent facilitates.²¹ So, to realise the good of directing one’s own life through autonomous action, one must be capable of valuing some interest, project, relationship, or end in the sense that it is a source of reasons for action.

What is this capacity to value? Perhaps the most thorough and influential analysis is offered by Samuel Scheffler. On his view, valuing some X involves:

1. A belief that X is good or valuable or worthy,

2. A susceptibility to experience a range of context-dependent emotions regarding X,

3. A disposition to experience these emotions as being merited or appropriate, and

4. A disposition to treat certain kinds of X-related considerations as reasons for action in relevant deliberative contexts.²²

Building on Scheffler’s analysis, Nandi Theunissen distinguishes simple valuing from final ends valuing. The former can be intermittent and unsystematic, while the values that form the normative principles upon which we direct our lives—our final ends—are typically stable and enduring.²³

Though the details of their views differ, the implications of the Scheffler/Theunissen account are similar to Jaworska’s: a valuer is someone capable of having rational attitudes (condition 1) and capable of being (stably) disposed to treat those attitudes as reason-giving and fitting of certain emotions (conditions 2, 3, and 4).

Only persons who possess these capacities could realise the distinctive good of directing their own lives. Unlike the toddler, the valuer can distance themselves from their immediate motivations, can occupy a reflective standpoint, and can judge that some interest, project, relationship, or end is worthwhile. It is generally agreed—no matter what your specific account of autonomy—that autonomous agency requires the ability to direct one’s life on the basis of reasons in just the way described.²⁴ Moreover, the person capable of valuing is stably disposed to believe and feel that what they value is worthwhile, and their emotional and behavioural responses are merited by its worth.

These marks of valuing explain why restricting Donna’s intimacy with Henry seems normatively different from restricting the toddler’s access to the drum. Before her cognitive decline, Donna was a valuer. She loved Henry and valued their relationship. As her disease progressed, her affection and behaviour towards him remained stable across time and context. Unlike the toddler’s mercurial whims, unintelligible desires, and sometimes irrationally strong or inappropriate emotional reactions, Donna was reliably delighted by Henry’s visits, always affectionate towards him in public, and frequently desirous of intimate relations in private. Had she wavered unpredictably, there would be few marks of agency to attend to, and managing her wellbeing would be more akin to managing the toddler’s.²⁵ And yet, her dispositions were not only stable, but fitting—she responded to Henry in the way in which one is often disposed to when one loves one’s spouse. Her behavior was contextually appropriate too: she held Henry’s hand in public and only sought sexual relations in private. Her stable and fitting emotional and behavioural dispositions are marks of valuing, and if these were motivated by the worth she placed in that relationship, then restricting her access to Henry did not just frustrate her desires; it prevented her from directing her life in accordance with her values.

The central question, then, is: was Donna still a valuer? In answering this question, the capacities not identified by standard analyses of valuing are just as relevant as those that are. Nothing in the views outlined implies that a valuer must have a coherent sense of their life as a narrative whole, the ability to describe their values eloquently, or the capacity for detailed means-ends reasoning, complex risk-benefit calculations, or even comparative assessments of different deliberative options. The fact that Donna has deliberative or cognitive frailties—impairments that make it difficult for her to implement principles derived from her values without her caregivers’ assistance—does not reveal her inability to direct her own life on the basis of reasons. She may still “know how she wants her life to be governed”²⁶ and her values could still constitute her normative principles of action.²⁷ Of course, we have too little information about this actual case to make a definitive judgement, but the suggestion that she might still be capable of valuing is also consistent with the neuropathology of Alzheimer’s disease. The path of neurological destruction in regions of the hippocampus responsible for language, memory, and abstract thought is “always several steps ahead of the pathologies in the areas most likely to affect the capacity to value.”²⁸

A valuer is able to direct their life on the basis of reasons and this means they can realize a very important prudential good.²⁹ Changing one’s normative relationships to others through consent is one form this good can take. This is why we think that the capacity to value is the central capacity that underlies the power to waive rights through consent. Standard analyses of valuing imply that this capacity might be quite minimal. Of course, we accept that those with more developed or sophisticated capacities are likely to get more good out of directing their lives in accordance with their values. But the question here is not: how much good can one achieve by directing one’s own life? The question is: at what point is one capable of achieving to some degree the good of directing one’s own life? It is at that point that the threshold ought to be set.

iii.b what capacities ground the claim against paternalistic interference?

We have argued that the power to waive one’s rights by giving consent is grounded in the capacity to value. The reason for this is the contribution to individual flourishing made by directing one’s own life according to one’s values. This component of competence is therefore ultimately grounded in considerations of the agent’s own well-being. Could the claim against paternalistic interference be justified in the same way? Just as it is good to autonomously direct one’s own life, it is widely thought to be bad for someone capable of autonomous decision-making to have their life directed by others.³⁰ Perhaps this setback to the individual’s interests is what makes it wrong for others to interfere.

The claim against paternalistic interference cannot be justified in the same way. While there is prudential value in directing your own life and disvalue in having your own life directed by others, these are just components of well-being. This means that while the value to the valuer of directing their own life will count against paternalistic interference, other components of well-being could count in favor of it, and sometimes these will be more weighty when all things are considered. Someone can exercise control in ways that are very self-destructive, making the overall impact of autonomous decision-making on their well-being negative. If the prudential value of consent and disvalue of heteronomy were the only grounds for a claim against paternalism, then the paternalist would be permitted to interfere with someone’s consent decision as soon as they had good reason to think that the net effect on the individual’s well-being of respecting those decisions would be negative. Appealing to the value of making one’s own decisions without interference would then be futile: the paternalist would have already factored that in.³¹

If considerations of well-being are not sufficient to ground the claim against paternalistic interference, then what is its basis? The natural answer is, again, autonomy. But this time it is not the value of autonomy in the sense of the contribution to well-being made by someone acting autonomously; it is the value of autonomy as grounding a demand to respect the autonomous agent as the author of their own life.³² While we think this answer is correct, as far as it goes, it does not help with cases in which someone’s agency is in question. As we have seen, there are different accounts of the capacities that constitute autonomy and these capacities can be possessed to differing degrees. At what point does someone move from being unable to demand respect for their agency to being able to demand it?

We suggest that this threshold lies at the point where someone has the capacity to be responsible for their actions.³³ At the point where someone can be responsible, they can be blamed for harms to others—that is, be “held responsible.” But they can also “take responsibility” for harms to themselves.³⁴ By doing so, they can make it so that these harms do not count in the moral calculus of the would-be paternalist (or at least not to anything like the same degree).³⁵

To see this, consider that sometimes the fact that an agent performed a self-harming act means that it is permissible to allow them to suffer the associated harms. We take no view on what other conditions must be met for this to be the case, except to note that it is only true when the person is responsible for the action. If they were not responsible—for example, because the action was involuntary—then other parties should treat preventing or rectifying the harm just the same as if it resulted from natural causes. Now, most people do not think that being responsible for a self-harming act is always sufficient to permit others not to assist. For example, a struggling swimmer still deserves help from a lifeguard, even when they knew about the risks. But being responsible can be sufficient in conjunction with other conditions. In particular, it can be sufficient when the agent knowingly and voluntarily refuses assistance. By refusing assistance, they accept the consequences. And if they accept the consequences, we think that those harms should normally be bracketed off in the moral calculus of the would-be paternalist.

In such circumstances, the harm is not sufficient to provide legitimate grounds for interference when the agent has refused assistance. The agent can say, in effect, “I’ll take responsibility for my own well-being.”³⁶

This concludes our positive argument. We have argued that competence comprises two rights: a power to alter one’s rights—in the cases we have discussed, by waiving them through granting consent—and a claim against paternalistic interference with one’s self-regarding decisions. These component rights have separate grounds. The power is grounded in the value for the consent-giver of controlling their own life. This value can be realized by those individuals who have the capacity to value. The claim against paternalistic interference is grounded in the capacity to be responsible for one’s own actions. Since it is plausible that someone can have the capacity to value but lack the capacity to be responsible, it is plausible that someone can have the power to consent but not have a claim against paternalistic interference. We now show how this entails some intuitive verdicts and makes sense of some otherwise puzzling phenomena.

iii.c applications

To show the potential practical payoffs of a Separate Grounds view, we must make some assumptions. Thus far, we have argued that the component rights implicated in consent transactions have separate grounds, but we have not argued for a specific Separate Grounds view. As we have already noted, competence is a necessary but not a sufficient condition for valid consent. Different conceptions of valuing and responsibility—together with different views of the other necessary and internal conditions for valid consent—will result in distinct accounts of competence.³⁷ In order to illustrate the important implications our analysis could have for consent cases involving persons with marginal autonomy, we now adopt some substantive positions on each of these.

A fleshed-out account of what internal capacities are needed to give valid consent will need to specify not just the criteria for being a valuer, but also what must be understood to give valid consent, and what act constitutes a token of consent. For the purpose of illustration, we accept Scheffler’s analysis of valuing. We adopt a minimal account of the understanding requirement, according to which the person giving consent must understand only that they are giving consent, how to token consent, and to what they are giving consent (that is, what rights they are waiving).³⁸ Competence therefore requires that someone be able to understand these facts. Finally, we assume that to give valid consent one must be capable of communicating one’s will.³⁹ To emphasize: these specific views are not essential to holding a Separate Grounds view. They matter, though, because they affect what a Separate Grounds view will imply for practice.

To arrive at a specific view of what grounds the claim against paternalistic interference we need one more assumption—this time about responsibility. It is commonly held that responsibility as accountability, which is the type of responsibility relevant to our discussion, comprises two necessary and jointly sufficient conditions—a control condition and an epistemic condition.⁴⁰ The control condition says that to be responsible for an act the agent must have acted voluntarily. The epistemic condition says that to be responsible for an act the agent must have been aware of certain facts (since ignorance is sometimes an excuse). Common components of this epistemic condition are that the agent knows (or should know) what she is doing, what the likely consequences are, that there are alternatives to performing the act, and the normative status of the act.⁴¹ Assume that the epistemic condition does indeed have something like this set of components.

Consider now what capacities are needed in order to work out the normative status of an act, e.g., whether it is prudentially warranted or morally permissible. To work out the normative status of an act, it is necessary to be able to reason about what to do—that is, to apply one’s values to the available facts and decide on that basis which alternative to pick. If you could not come to the right decision for the right reasons, then you cannot be held responsible for not doing the right thing. But to do this, an agent must be able to: understand the information relevant to the decision (or the implications of their ignorance), appreciate how that information applies to their situation, reason in the light of their values, and come to a decision on the basis of reasoning. This entails, as seems plausible to us, that the capacities needed to be responsible for one’s actions—and so to possess a claim against paternalistic interference—are more substantial than those that ground the power to waive one’s rights against interference.

Having sketched a possible Separate Grounds view, we can now show some of its implications for cases. In the blood donation case, a Separate Grounds view of this sort implies that Medhi has the capacities necessary to waive his rights and give valid consent. However, he does not have the more complex cognitive capacities necessary to be responsible for his actions. There is no paternalistic reason to prevent Medhi from donating blood, and so there is nothing in this case that could justify another person overriding his rights waiver. Our view implies that he ought to be permitted to make this decision without interference. Indeed, even if donating blood were slightly contrary to Medhi’s interests, the value to him of making his own decisions on the basis of his values could tip the scales against paternalistic interference. While Medhi also has the capacity to give valid consent to study participation, in that case there are clear paternalistic reasons to interfere. Since he is not capable of being responsible for his actions, he is not immune from interference. Given the complexity of the decision and the serious risks involved, someone else ought to help make the decision about whether trial participation is a good option for him.

Similar points apply to Donna Rayhons who, on the basis of the media reports, may well have retained the ability to value. If so, she would have been able to waive her rights through consent and so it is possible that her husband could permissibly have sexual contact with her. However, she, like Medhi, likely lacked the capacities grounding a claim against paternalistic interference. Whether sexual contact that she welcomed was permissible would therefore depend on what was in her interests. She might need help to make a decision that took her preferences seriously and balanced the benefits to her of intimacy with someone she loved against the potential harms of engaging in sexual relations.⁴² For example, if she welcomed the sexual contact and it made her happy, she should probably be allowed to agree to it. However, if sex was initially welcomed, but left Donna distressed, then Henry should decline her future invitations to sex.⁴³ And, were he to continue to accept them, her caregivers might be obligated to intervene in the couple’s meetings to prevent harmful, albeit consensual, sexual activity.⁴⁴

These cases illustrate the value of consent to persons with marginal autonomy, even when they lack a claim against paternalism. The ability to give and refuse consent allows Medhi and Donna to achieve a good that those who lack autonomy cannot realize. The value of consent does not just reside in an interest in non-interference. It also resides in an interest in allowing other people to interact with us in specific ways—whether that be allowing a medical professional to draw our blood or a spouse to be physically intimate with us. Even if the only consent decisions that were verdictive for persons with marginal autonomy were those that were also in their interests, the degree of control this affords is likely to be especially valuable to persons like Medhi and Donna who often have very little control over their own lives.

Consider now a second set of cases, applying our view to the developing capacities of children. In buying and selling goods, one needs to exercise autonomy rights to alter one’s claims. Intuitively, some children can engage in some of these commercial transactions, and some cannot. If a two-year-old offers a shopkeeper money, the shopkeeper probably ought to do nothing, except look around for the toddler’s parents. The toddler cannot buy anything from her themself. On the other hand, a typical eight-year-old can transact. If they come into her shop to buy a $10.99 toy, then, normally, she can sell them the toy just as she would with an adult. Now, consider the same eight-year-old who attempts to buy a $8,000 state-of-the-art gaming system. Presuming the shopkeeper does not know the child already, she should not sell them the system. She should not sell them the system, even if they show her the receipt indicating that they just emptied their own savings account for the $8,000 and so the money is definitely theirs. Selling them the gaming system would not violate their rights, but it would be unscrupulous—it would take unfair advantage of the child. The shopkeeper, like everyone else, has a duty not to exploit persons with marginal autonomy.

All these verdicts seem very intuitive to us. They are also neatly explained by a Separate Grounds view, like the one we have described, that posits a substantial gap between what is needed to be a valuer and what is needed to be responsible for one’s actions. A typical eight-year-old, unlike a toddler, is competent to engage in some financial transactions—they have values, they know what it means to exchange money for goods, and so on. Like Medhi and Donna, they can achieve the good of altering their normative relationships with others—a good that could not be realised without the capacity to consent. They are not, however, able to take responsibility for their decisions. Exchanging a small amount of money for a toy is unlikely to substantially impact the child’s welfare one way or another (despite what they might think!). Even if they make a prudential mistake, the importance for them of making their own decision is likely to be great enough that paternalistic interference would not be warranted. It is these consent decisions—the ones that range from the prudent to the slightly imprudent—that showcase the role and value of consent even when there is no claim against paternalism. When the would-be paternalist’s calculus does not support interference, the ability to consent affords a person with marginal autonomy decision-making authority that is akin to a fully competent person’s. But when the calculus supports interference, consent (or refusal) is not verdictive. In this instance, emptying their savings account is something that might have a large effect on their well-being. Paternalistic interference is warranted and policies and practices that prevent children from engaging in large financial transactions without supervision are justified.

Similar judgements are reflected in medical law and practice. Mature-minor laws recognise a child’s right to consent to treatment that is in their best interests when they meet understanding and voluntariness conditions for valid consent.⁴⁵ For example, in England and Wales this is reflected in “Gillick” competence assessments, which evaluate a child’s maturity, understanding, and appreciation, and in the Fraser guidelines, which govern the sexual healthcare of minors.⁴⁶ For example, children under the age of sixteen who are Gillick-competent can consent to receive contraception and abortion services without parental knowledge or involvement.

And yet, were a Gillick-competent fourteen-year-old to refuse an appendectomy after suffering a ruptured appendix, doctors in England and Wales would be permitted to override their decision to prevent them from dying from a life-threatening infection.⁴⁷ Here, again, our Separate Grounds view can make sense of phenomena that Common Grounds views struggle with—namely, laws that treat young persons’ consent and refusal asymmetrically when the risk-benefit profile of these options differ. As a valuer, a fourteen-year-old might be capable of consent and refusal, and if they consent to life-saving treatment that is in their interests there is no paternalistic reason to interfere with their decision. However, a fourteen-year-old is not capable of being held responsible for their decisions. This is what explains why their refusal of the very same life-saving care can be overridden if the harms of non-interference outweigh those of paternalistically forcing treatment. Of course, the decision to interfere needs to take seriously the benefit to this valuer of making their own decisions, and much of the time this will count against overruling them. But not in this case—not when a valuer who is incapable of being responsible for their actions refuses surgery that is required to prevent peritonitis and potentially life-threatening sepsis.

iv.a the view seems to imply that some children can give valid consent to sex

Many older but preadolescent children have the capacity to value. Many can surely understand what would be needed to permit sexual acts, at least on a minimal account of the understanding requirement. Since valid consent is often thought to be a right-making feature of permissible sexual activity, it might appear that our view would permit adults to have sex with older children if the children consent. This would be troubling.

However, the Separate Grounds view does not have this implication. Competence—as we have emphasized throughout the paper—is a necessary, not a sufficient, condition for valid consent.

Adult-child sexual relations are profoundly wrong. There are two, often mutually reinforcing, explanations for why: the consent explanation and the harmful exploitation explanation. In every case of adult-child sexual relations, the Separate Grounds view can avail itself of the correct explanation of why they are wrong.

Take the consent explanation: Even a child who met both the competence and understanding conditions for valid consent would only succeed in waiving a right against sexual trespass if their consent were also sufficiently voluntary. This condition is not met in actual cases of adult-child sexual relations. These abusive relationships are built and maintained with the kind of coercion and deceptive emotional manipulation that constitutes sexual grooming. As we have argued elsewhere, control of this sort undermines the voluntariness of consent.⁴⁸

Of course, one might object that even if it is very unlikely that a child could give valid consent to sex with an adult in any actual cases, it is not impossible. However, we ought to be extremely wary of such thought experiments. They ask us to imagine a case in which a child voluntarily agrees to and proceeds with sexual contact with an adult after no grooming, no deceptive manipulation, and no coercion. This case bears so little resemblance to any actual case that we doubt that intuitions about it can render trustworthy moral verdicts.

Another way to see the problem with these sorts of imagined cases is to recognize that they can also be raised against Common Grounds views. It is conceptually possible that a child could meet the competence, understanding, and voluntariness conditions on any theory of consent. After all, we can simply stipulate hypotheticals in which a child could meet, say, the four abilities for competence. Of course, such thought experiments are never raised because once the stipulated hypotheticals are this divorced from reality, many agree that our moral intuitions would be so unreliable that they could not play an adequate role in our moral theorizing. However, if reality posed no constraint on such theorizing, then we quickly see that the implications of such outlandish thought experiments are far worse for Common Grounds views. After all, only on those views does valid consent entail a claim against paternalism. Incredibly, this implies that were a child competent to consent and were they to meet the other conditions for validity, it would be a wrongful rights violation to interfere with their valid consent to sex with an adult.

This brings us to the harmful exploitation explanation of the wrongfulness of adult-child sexual relations. We, like a number of others working on sexual ethics, think that valid consent is not sufficient for morally permissible sex.⁴⁹ We also think that consensual sex can be wrongful when it is exploitative. In any realistic situation in which a competent adult has sex with a child, the adult will be taking advantage of the child and thereby exploiting them. Moreover, these acts are deeply harmful and the child does not stand to benefit from them. The validity of consent is far from the only moral issue.⁵⁰

Our Separate Grounds view also allows us to make sense of some sensible legal distinctions relating to the age of consent. Many consent laws governing sexual relations have close-in-age exceptions. For example, in Italy exceptions are granted for consensual sexual activities between minors if (i) neither is under the age of 13 and (ii) the age gap between the sexual partners does not exceed three years.⁵¹ It is not obvious that mutually desired sexual relations between two 13-year-olds involves bodily rights violations. Sexual relations might be ill advised—and there might well be all-things-considered paternalistic reasons to interfere with some of it—but it is not obviously wrongful. Many 13-year-olds are valuers, can understand enough to waive their rights, and do so voluntarily. Contrast this with sexual relations between a 13-year-old and a 40-year-old. This will be wrongful, but the explanation of why does not need to refer to the 13-year-old lacking the power to consent. The wrong will be explained in virtue of likely impediments to voluntariness—due to grooming, coercion, and deceptive manipulation—and the 40-year-old’s harmful exploitation of the minor.

iv.b the view dilutes the value of consent

A central tenet of the Separate Grounds view is that the power to waive a right through consent does not include a claim against paternalistic interference. But without this claim one might wonder what the value of consent is. Many would say that its value lies in protecting us from unwanted interference. However, Separate Grounds views imply that consent cannot secure this good alone. Its achievement requires a separate right: a claim against paternalistic interference. This might leave one thinking that what does the work we tend to associate with the value of consent is a risk-benefit calculation by a third party. The worry, in short, is that Separate Grounds views dilute the value of consent.

Were the value of consent to lie exclusively in our interest in non-interference, this worry would have force. When a valuer lacks a claim against paternalistic interference, they cannot secure independence. In fact, those valuers who have just passed the threshold are likely to have such weak prudential interests in directing their own lives that their decisions may frequently be overridden. Fortunately, the value of consent is not exhausted by this interest. Its value also lies in our interest in redrawing rights and duties in ways that advance our goals—such as by allowing another to cut our hair, invest our retirement funds, or even operate on us. When a person lacks the capacity that grounds the power to waive their rights, they cannot permit or refuse these interactions. Even if the 4-year-old wants their ears pierced or does not want to go to the dentist, they are not a valuer. They lack the capacities necessary to achieve this value of consent. It makes no sense to factor in their interest in directing their own lives into this decision-making. The same cannot be said of the 14-year-old. Even if we are just considering prudential consent decisions, the 14-year-old can achieve this good without needing a claim against paternalism.

Not only is this degree of control likely to be valuable to the 14-year-old, their control does not just extend to prudent decisions. Lacking a claim against paternalistic interference does not license all and any interference by third parties (and, likewise, having that claim right does not forbid all interference). It is important not to conflate “no claim against paternalistic interference” with “no claim against interference.” For anyone, including those who lack both of the component rights of competence, some justification must be given for stopping them from doing what they prefer. For those who are valuers, the justificatory threshold is higher, since it must take into account the value to them of directing their own life. This applies to both consent and refusals. Insofar as making their own decisions is in their interests and no other moral justifications for interference apply, they should be left alone.⁵² This leaves valuers who are not yet responsible protected from much unwanted interference. Since they have the capacity to consent they can redraw the normative landscape to achieve goods that further their interests, preferences, and values.

vi.c the relationship between the good of making one’s own decisions and the power to consent is ambiguous

We have claimed that there is a distinctive prudential good that is realized when someone directs their life on the basis of reasons derived from their values. It is this good that is facilitated by the power to consent and so grounds the power. However, the relationship between the good and the power could be interpreted in two ways, both of which might seem problematic.⁵³ On a strong interpretation, a valuer’s consent is only valid if their consent decision is in fact in line with their values. That interpretation appears inconsistent with how consent works, since it would seem to rule out weak-willed or foolish decisions. On a weak interpretation, a valuer’s consent can be valid even if it is not in line with their values, but then it is not clear how the consenter achieves the good of directing their life based on reasons derived from their values.

In response, we would emphasize again the central insight of our analysis: that there are two separate rights implicated in consent transactions—not one. This is the constitutive claim of a Separate Grounds view and it stands independent of our further views about what a valuer is and exactly how the capacity to value relates to the prudential value of making one’s own decisions. However, we also happen to have a view on this question which, again, can help illustrate the payoffs of a Separate Grounds view. We think that having and applying values to your life is a distinctive good even if, on occasion, flaws in reasoning result in a consent decision that is, in fact, not in line with your values. In addition, we think that a valuer achieves more good if they direct their life based on reasons that successfully reflect their values.

If our substantive assumptions here are correct, then a Separate Grounds view that makes them could explain the value of scaffolded decision-making. If it is better that a valuer directs their life based on reasons that reflect their values, supported decision-making can help people like Donna and Medhi achieve this greater good by helping them identify, implement, and achieve their goals. It would also explain why it is worse to paternalistically interfere with a consent decision that is closely aligned with the valuer’s values than it is to interfere with one that is not.

Moreover, if this specific Separate Grounds view were correct, it would give us the benefits of both threshold and scalar approaches to competence. Given the functional role capacity assessments must play in practice, our goal is to provide the best account of a threshold view. This is a goal we share with Common Grounds theorists. However, the response we just gave implies that when the threshold is met, some valuers can get more good from the capacity to direct their lives in accordance with their values. The threshold is set at the lowest point at which a valuer can get some good from directing their own life based on reasons derived from their values. But the more good the valuer can achieve, the higher the bar for justified paternalism, especially when autonomy-enhancing interference could scaffold good decision-making instead.