Children’s Capacities and Paternalism

Abstract

Paternalism is widely viewed as presumptively justifiable for children but morally problematic for adults. The standard explanation for this distinction is that children lack capacities relevant to the justifiability of paternalism. I argue that this explanation is more difficult to defend than typically assumed. If paternalism is often justified when needed to keep children safe from the negative consequences of their poor choices, then when adults make choices leading to the same negative consequences, what makes paternalism less justified? It seems true that ordinary adults have capacities enabling them to promote their interests in ways most children lack. This can explain why paternalism is more often justified towards children than adults. What is not explained, however, is why paternalism would be justifiable for children, but not adults, when neither possess the relevant interest-promoting capacities—exactly the cases when paternalism towards adults might be considered. I argue that this dilemma undercuts capacities-based explanations for the belief that childhood is distinctively relevant for the permissibility of paternalism. I then address defenses of both consequentialist and deontological versions of the capacities-based explanation. Absent this capacities-based explanation, I argue that the intuition that less demanding justificatory standards apply to paternalism when directed at children than when directed towards adults presents unresolved problems for egalitarians.

Paternalism towards adults is thought to raise grave moral concerns, at least among those with liberal and egalitarian views. Even the most consistent opponents of paternalism, however, generally make exceptions for children, or stipulate that their opposition to paternalism extends only to adults. Paternalistic management of highly personal choices, such as choosing friends, diet, media consumption, religious practices and bedtimes, is widely thought to be eminently justifiable for children, but morally intolerable for adults.¹ These beliefs correspond to a widely shared view that might be generalized as the following:

‘Childhood Relevance View’: Whether paternalism² is justified towards a person can depend merely on whether they are a child.³

This view implies that paternalistic coercion towards children is justified according to different, less demanding standards than those required to justify paternalism towards adults. A view with this structure would ordinarily present a problem for egalitarians who believe that people should be treated equally absent compelling justification. This problem, however, has a widely accepted⁴ solution:

‘Incapacities Explanation’: The childhood relevance view is justified because children lack relevant capacities possessed by ordinary adults.

In this paper, I argue that justifying paternalism through appeals to lack of capacities does not lend support to the childhood relevance view—surprisingly, such appeals actually undermine the view that childhood is itself relevant to paternalism. This is because the incapacities explanation depends on an intermediate conclusion: whether paternalism is justified towards a person can depend merely on the presence or absence of certain capacities (hereinafter the ‘capacities relevance view’). Explaining and defending this view, I argue, lends support to judgments that diverge from the childhood relevance view.

An application of the childhood relevance view is the tendency to discount children’s liberty interests and judge paternalistic interventions justifiable when they leave children better off. The claim that an adult is left better off by a paternalistic intervention, however, would rarely be considered sufficient justification: instead, the costs that the paternalistic intervention imposes on that adult’s interests, such as interests in liberty, dignity, bodily integrity or equality, would also be taken into account. The setbacks to those interests will often be thought to override whatever benefits paternalism confers on adults, or to support deontic constraints on paternalism (Dworkin 1972: 71), even though the same interests are rarely considered for children. It is also rarely thought compelling grounds for paternalism that an adult appears to lack relevant capacities (such as reasoning or will power) such that they will, absent paternalism, suffer a bad outcome, even though paternalism towards children is justified on similar grounds.

This type of thinking is implied in On Liberty where Mill asserts that paternalism is permitted for children, who “must be protected against their own actions” (Mill 1859: 22–23)—but not for adults, since adults, unlike children, are best situated to make decisions for themselves (Mill 1859: 136–138). Mill, however, maintained that paternalism remains impermissible for adults even if they are “equally incapable of self-government” as children (Mill 1859: 144). In this way, when considering children, substandard or insufficient capacities are thought to justify paternalism and liberty interests are not considered, but when considering adults, substandard or insufficient capacities are not thought to justify paternalism owing to adult interests in liberty.

The divergence between the childhood relevance view and capacities relevance view can be illustrated through the following hypothetical scenario:

Unhealthy Diet: Two adults, Abby and Betsy, and two children, Cathy and Dorothy, are all following a diet that puts them at risk for the same negative health outcomes. That risk can be eliminated by the same change of diet. Two interventions are available: a paternalistic intervention involving mild coercion and a non-paternalistic intervention involving an educational webpage. It happens to be known to a high degree of certainty that all four people will adhere to a change of diet if the paternalistic intervention is applied. It is also known that Betsy and Dorothy possess greater diet-relevant self-regulatory capacities than Abby and Cathy. As a consequence, if the non-paternalistic educational intervention is applied, Betsy and Dorothy will successfully change their diets and avoid the negative health outcomes, but the non-paternalistic intervention will have no impact on Abby and Cathy. Absent the paternalistic intervention, Abby and Cathy will remain at risk.

In the unhealthy diet scenario, the view that childhood independently affects the justifiability of paternalism might lead to the conclusion that it is morally permissible to apply the paternalistic intervention to one or both children, but not the two adults, assuming that other conditions for justifiable paternalism are met. Someone only concerned with how relevant capacities bear on the justifiability of paternalism, however, might think it wrong to paternalize towards Betsy and Dorothy, since a non-paternalistic intervention is sufficient to eliminate the risk they face, and might conclude that if it is justifiable to paternalize towards Cathy, it would likewise be justifiable to paternalize towards Abby. In such a case where we have more direct information about each person’s relevant capacities than the rough proxy of whether they are a child or adult, many would likely resist adopting the capacities-based conclusion, even though the childhood relevance view is standardly explained in reference to children’s incapacities.

The first part of this paper attempts to make sense of the capacities relevance view by considering how different types of capacities could be relevant to paternalism. In Part 1.A, I argue that capacities more frequently possessed by adults can be morally relevant to paternalism because they can obviate or diminish the rationale for paternalism. This sort of moral relevance, however, is situation specific, not categorical. In Part 1.B, I argue that judgments about the justifiability of paternalism based directly on capacities cannot provide a defense of the view that childhood is categorically relevant to paternalism. The second part of this paper considers defenses of several variants of the incapacities explanation. In Part 2.A I consider consequentialist views that children make systematically bad choices in a broad set of domains due to their lack of capacities. In Part 2.B, I respond to deontological defenses of the incapacities explanation that hold that children do not make morally meaningful choices at all, and therefore deontic constraints on paternalism applicable to adults do not apply to children. I argue that neither set of defenses can withstand scrutiny. In Part 3, I consider grounds for adopting the childhood relevance view that do not initially seem to require a categorical version of the incapacities explanation, such as the idea that childhood is relevant in practice because it is a good proxy for capacities. I argue that, without a categorical incapacities explanation, these non-categorical variants will not support the childhood relevance view. In Part 4 I argue that the apparent failure of the incapacities explanation points to broader implications about paternalism, some of which should be concerning to those with egalitarian commitments.

1. The Relevance of Capacities for Paternalism

Although it is widely agreed that paternalism is permissible towards children in instances where it would not be permissible towards adults,⁵ and that this is explained by differences in capacities, exactly how adult capacities modify the permissibility of paternalism remains underexplained. In this section, I will argue that if paternalism is sometimes justified for consequentialist reasons, a subject’s possession of capacities that could relevantly alter those consequences may reduce or eliminate the force of those reasons. Appealing to the lack of such relevant capacities as grounds for the permissibility of paternalism does not, however, support the childhood relevance view—instead it actually undercuts it. I will leave aside the possibility that certain capacities possessed by adults might provide deontic prohibitions on paternalism but will return to it later in this paper in Part 2.B.

Before proceeding further, it is useful to clarify what I mean by “capacities.” In legal and medical contexts, the terms “capacity” and “competency” are sometimes used interchangeably in potentially confusing ways (Moye 2013: 158). Generally the term “capacity” denotes functional abilities, such as performing certain tasks, exercising certain forms of reasoning, or making specific decisions, whereas “competence” denotes the legal status presumptively possessed by adults to act for themselves independently.⁶ Since this paper concerns the moral justifications for paternalism, and philosophers writing on the morality of paternalism and children tend to use the term “capacities” and not “competences,”⁷ I will use the term “capacities” as defined above and reserve “competence” for the legal status.

1.A How Can Capacities Modify the Justifiability of Paternalism?

To evaluate how certain capacities could defeat reasons motivating paternalism, it is useful to consider what minimum conditions must be met for any instance of paternalism to be plausibly justifiable. First, we likely must assume that people have interests that they do not subjectively appreciate. If a person’s interests are limited to those they themselves identify, then a paternalistic agent would rarely be justified in believing that they act in a subject’s interests over their objections.⁸ Second, because the entire moral impetus of paternalism is to benefit the subject being coerced, an intervention can only be justified on paternalist grounds if it truly benefits the subject.

These criteria imply that for paternalism to be justifiable, it must satisfy at a minimum what might be termed a ‘paternalistic calculus’: the benefits to the subject of a paternalistic act must outweigh any detriments to their interests.⁹ If an action harms a subject more than it benefits them, then the action would not be to their net-benefit and would therefore not serve properly-so-called paternalist purposes. Actions towards a person that are to their net-detriment might be justified on other grounds, such as self-defense, but not as paternalism.

An appeal to this paternalistic calculus is implied in the often used example of a parent requiring their child to eat their vegetables before they can eat dessert as justifiable in part because the health benefits conferred by the paternalistic intervention are substantial while the setbacks to a child’s liberty are trivial (Husak 2010: 108). It can also be inferred from the way most think about paternalism towards adults. Whether one thinks that paternalism is rarely or often justified, minor impositions that confer major benefits—such as laws threatening small fines for not wearing seat belts, when seat belts confer enormous safety benefits—will, all other things equal, always seem more justifiable than severe impositions that confer minor benefits—such as threatening imprisonment to compel someone to learn a skill that is superfluous to their ability to function in society.¹⁰ Force or threats are generally understood as pro tanto harms, at least for adults, so the coercion used to effectuate a paternalistic intervention can be understood as a cost to the subject’s interests to be weighed against the benefits the intervention aims to confer.¹¹

This balancing of costs and benefits points to a way that two types of capacities could modify the paternalistic calculus. Paternalism might be aimed at either coercing a subject into accepting a decision in their interests, or ensuring the decision is actually carried out, or both. If a subject has decision-making capacities that enable them to make the interest-advancing decision in question, or an equally beneficial decision, then any coercion used to induce them to adopt this decision is unnecessary. Likewise, if a subject has executive capacities that allow them to carry out a beneficial course of action, then any coercion employed for this purpose would also be unnecessary. To assert that children are unable to choose rationally is to allege that they lack decision-making capacities; to assert that children cannot take care of themselves is to allege that they lack executive capacities.

To illustrate by way of a hypothetical, imagine a town governed by a paternalistic official who has compelling evidence that three townspeople (A, B, and C) will develop type-II diabetes if they continue their sugar-rich diets. The official has the authority to issue injunctions mandating that a townsperson agree to change their diet, and the authority to add townspeople to a merchants’ database of people prohibited from buying sugar-rich foods.

Townsperson A erroneously believes that diet and type-II diabetes are unrelated, despite the official’s evidence-based efforts to persuade them otherwise. Their decision-making capacity is somehow impaired in this instance. The official might conclude that while using the injunction and database will humiliate Townsperson A, doing so would sufficiently advance A’s health interests that A would benefit overall.

Townsperson B, unlike A, has the decision-making capacity to review the official’s evidence, weigh the enjoyment of sugar-rich foods against the probable detriment to their health interests, and agree to change their diet without coercion. Townsperson B, however, lacks the executive capacity to carry out their decision. B would never eat sugar-rich foods if B had perfect impulse control, but given B’s akrasia, B will continue to eat sugar-rich foods against their better judgment. For Townsperson B, the official has no paternalistic reason to issue an injunction since this would cause humiliation without providing any benefits, but they would still have a paternalist reason to add B to the no-sugary-foods database. Given B’s judgment, B might agree to be added to the database, in which case the database intervention would be consensual not paternalistic, but they may not. Judging a decision to be correct does not necessarily imply either the ability to carry it out or agreeing to steps needed to do so.

Townsperson C, unlike A and B, possesses both the decision-making capacity to choose to change their diet given evidence or expert advice, and the executive capacity to restrain themselves from buying sugar-rich foods. C would find both the injunction and the database humiliating, and neither intervention would provide C any benefit, since C would decide to abstain from sugar-rich foods and carry out that decision on their own. As a consequence, C’s capacities alter the official’s paternalistic calculus: both interventions represent setbacks to C’s interests in avoiding humiliation without advancing C’s interests in maintaining their health, and therefore cannot be to C’s net-benefit. C’s capacities defeat the official’s reasons for considering paternalism.

If a person possesses decision-making capacities and executive capacities that enable them to achieve the benefit or avoid the harm motivating a paternalistic intervention, those capacities render that intervention unjustifiable even within a paternalist framework. Most adults have a larger set of decision-making and executive capacities than most children. As a result, there will be many more situations where paternalism cannot be justified in light of capacities possessed by typical adults than in light of capacities possessed by typical children.

1.B The Ways Capacities Modify the Justifiability of Paternalism do not Support the Childhood Relevance View

This explanation for how capacities can negate the reasons for paternalism explains why it makes sense for adults to be subject to paternalism in fewer cases than children. It does not, however, explain why an adult, but not a child, should be immune to paternalism when both lack the same relevant capacities. If decision-making and executive capacities are morally relevant to the justifiability of paternalism by enabling a subject to obtain the beneficial outcome sought by a paternalistic intervention, then they are relevant to particular situations, problems, benefits and harms, rather than properties of a person applicable to paternalism generally. Adults are in positions analogous to Townsperson C more often than children, and children are in positions analogous to Townsperson A or B more often than adults—but in cases where a particular adult and a particular child both lack the decision-making capacity to choose in their interests and/or the executive capacity to implement an interest-advancing choice, how could differences in their capacities permit paternalism towards the child but not the adult?

The question of whether it is permissible to paternalize towards adults tends to arise precisely in those cases where adults appear to lack the relevant decision-making or executive capacities to promote their own interests. The right to refuse medical treatment, for example, is paradigmatically the right to reject decisions that third-party medical experts believe to be essential to protect a person’s health interests. This implies that the right to refuse treatment is not based on the belief that adults choose best for themselves, but that adult’s health interests do not provide sufficient reasons to subject them to force or coercion.¹²

Consider the following case:

Infected Foot: Two patients, one 12-years-old, the other 42-years-old, both arrive at a hospital with identical foot infections. An examining physician determines that, if treated with antibiotics only, both patients have a 50% chance of survival, but if each patient’s infected foot is amputated, they will each have a 99% chance of survival. The physician recommends to both the 12-year-old and 42-year-old that they should have their infected foot amputated. Both patients make statements to the effect of “I recognize the risk to my life of using antibiotics only, but I’d prefer to take my chances rather than lose my foot.” Upon being questioned, both the 12-year-old and 42-year-old demonstrate basic numeracy, understanding of the materiality of death, the relative normalcy of living with a prosthetic foot, and neither have unusual athletic or professional motivations. The physician says to the 42-year-old “that is a poor choice, I urge you to reconsider, but it is morally yours to make as an adult” and says to the 12-year-old “that is a poor choice, but since you’re a child, I am not morally bound to respect it.”

If the 12-year-old asks what non-arbitrary grounds the physician has for thinking the 42-year-old’s decision demands respect but their own decision does not, and the physician offers the standard account that children “cannot rationally advance their interests” (Rawls 1999: 218) but the 42-year-old has “attained the capacity of being guided to their own improvement” (Mill 1859: 23), this will not be persuasive. The physician cannot claim the 42-year-old exercised capacities to choose according to their net-interests since both patients made the same ‘poor’ choice against the physician’s assessment of their interests. This poses a dilemma for the incapacities explanation for the childhood relevance view. The 42-year-old might have capacities enabling them to decide and act according to their interests in many situations where the 12-year-old cannot, but if those capacities do not enable a better decision or reasoning in this case, then why are they relevant? The 12-year-old likely has decision-making and executive capacities that enable them to promote their interests without paternalism in other cases too (e.g., keeping out of traffic without being restrained), including cases some adults struggle with (e.g., refraining from cigarettes). Adults are not alone in possessing some capacities to advance their interests.

2. Two Defenses of the Incapacities Explanation

In the previous section, I argued that, while the incapacities explanation is an initially appealing justification for the view that childhood is distinctly relevant to paternalism, it is only clearly successful at explaining why adults might have capacities-based defenses to paternalism in more situations than children. It does not provide a clear explanation for why paternalism towards children should be subject to less stringent justificatory standards than paternalism towards adults.

In this section, I will evaluate two ways the incapacities explanation might be elaborated and defended. First, it could be argued that children lack capacities that aid most or all decisions. Given facts about human development, childhood is categorically relevant because certain pervasive incapacities would lead to intolerable consequences for children absent paternalism. Second, it might be objected that I have erroneously conflated the moral significance of paternalistically overriding bad choices adults make with that of making choices for children who cannot choose at all. Paternalism towards someone who makes suboptimal choices disrespects their autonomy and represents a setback to their interest in choosing for themselves, but paternalism does not jeopardize equivalent interests when directed towards someone who does not make truly voluntary choices. Both the first set of broadly consequentialist arguments and the second set of deontological arguments are initially plausible when characterizing children in isolation, but fail when the same standards are applied to adults.

2.A “Poor Choices” Arguments: Childhood is Relevant Because Children Pervasively Lack Capacities Needed to Choose Well

Defenders of the incapacities explanation frequently advance some variant of the claim that children “lack some of the emotional and cognitive capacities required in order to make fully rational decisions,” exposing them to dangers that justify paternalism (Dworkin 1972: 76). The claim is generally not that children have no capacity to reason but that they reason poorly and suffer an “incapacity to make choices which protect and advance their own interests,” or “fail to arrive at good conclusions” (Schapiro 2003: 580).

Failing to reason well, of course, does not categorically distinguish children from adults. Even those who cite children’s deficiencies in rationality as justifying paternalism recognize significant overlap in the rational capacities of adults and children—at least adolescents—(Gutmann 1980: 354). Likewise, legally competent adults possess relevant capacities to differing degrees (Feinberg 1986: 30). More distinctive deficiencies in capacities characteristic of children must be identified to make this position plausible. Several candidates are often proposed: children cannot appreciate their future interests (Dworkin 1972: 76; Feinberg 1986: 325; Mill 1849: 148); lack impulse control (Dworkin 1972: 76; Goodin 1993: 233); and lack settled aims and preferences (Gutmann 1980: 340). We can divide these claims into ‘strong’ and ‘weak’ versions, where ‘strong’ versions assert that children lack these capacities altogether and ‘weak’ versions assert only that children’s capacities in these areas are distinctly limited.

Children’s inability to imagine themselves in the future and appreciate their future interests is among the most frequently cited incapacity. It is also intuitively plausible given children’s relative lack of experience over time. There are empirical grounds for thinking that the ‘strong’ claim is true of only very young children. Children as young as three “have the capacity for future hypothetical reference” (Atance and Meltzoff 2005: 342–43). Atance and Meltzoff presented subjects aged three, four and five with hypothetical future scenarios they might find themselves in, and asked their subjects to choose what items they would need to safeguard their future physiological needs. 74% of the 3-year-olds, 91% of the 4-year-olds, and 97% of the 5-year-olds chose correctly. When asked to verbally explain their reasoning, nearly all of the four and five-year-olds explained their thinking in terms of anticipated future states (Atance and Meltzoff 2005: 349–50). This suggests the ‘strong’ claim about future-oriented reasoning applies only to the youngest children, whereas the childhood relevance view extends to much older children.

The ‘weak’ claim, that children can imagine and plan for their futures, but cannot do this very well, is not very helpful for the childhood relevance view—since this does not provide a categorical distinction between child and adult capacities. Adults’ perceived failures to plan for their future needs effectively are well known and present some of the few cases where paternalism towards adults is often considered.¹³ L.A. Paul and others have also argued that our ability to make many major life decisions rationally is often complicated by the fact that they may be both personally and epistemically transformative. For example, it is impossible to know what it is like to be a parent before becoming one, and doing so often changes a person’s subjective values in ways they could not anticipate when making the choice.¹⁴

Likewise, children might have especially poor impulse control and capacities for delaying gratification, but child and adult capacities in these areas differ only in statistical norms, not categorically. Even very young children can delay gratification to varying degrees,¹⁵ whereas adults continue to struggle with temptation and impulse control, often failing to act on their second-order desires to protect their future interests (Frankfurt 1971: 133). The ‘weak’ claim regarding children’s impulse control, as with future-oriented reasoning therefore also fails to provide the categorical distinction needed to support the childhood relevance view.

It is debatable to what extent children’s aims and preferences are less settled than adults’ or that less settled judgment demand less respect (Goodin 1993: 238). A Pew Research Forum on Religion and Public Life survey found that roughly 50% of Americans changed their religion, many multiple times, and that while the frequency of religious conversion drops off dramatically as people age, more people change their religion as adults than as minors (Pew Research Forum on Religion and Public Life 2009: 3, 46). Adults’ political preferences also seem to change with new information, arguably a ‘learning effect’ giving more reason to respect those views (Goodin 1993: 241). Possessing views that are comparatively fixed might be a reason for challenging those views rather than leaving them alone, if this is interpreted as suggestive of less willingness, ability or opportunity to revise them in light of new information.

It might be tempting to concede that, although differences between the relevant capacities of adults and children are matters of statistical norms rather than categorical differences, this is sufficient to warrant presumptively permitting paternalism towards children but not adults. This position would have problematic implications. If a group of people who have capacities superior to average adults in a certain domain can be identified, would they not be similarly justified in paternalizing towards typical adults? Responding that there is a threshold above which no distinctions among average capacities are morally relevant and that threshold corresponds to average capacities at the age of majority requires an argument for why that threshold is the morally relevant one. Answering the complaint raised in the infected foot scenario requires more than identifying capacities adults typically possess in greater degrees than children; it requires explaining why the capacities and thresholds identified are apt and non-arbitrary.

If we reflect on what capacities might be relevant to broad sets of cases of possible paternalism without anchoring our consideration in favor of those that might support the childhood relevance view—we might arrive at a different list. One obvious capacity is consciousness: someone unable to gain consciousness cannot form or communicate decisions or take deliberate actions. Another is language: someone who cannot communicate with linguistic specificity might be able to make choices, but any choices they make will be partially or totally opaque to others, and they will have profound limits on their ability to demonstrate to others what they understand.

When someone arrives at a hospital unconscious and without any instructions, decisions to treat them are neither against their will nor in accordance with their will, because their will is unknown. This necessitates choosing on other grounds. Pre-linguistic infants present analogous necessities. Infants’ abilities to communicate are so limited that even a caretaker opposed to using paternalism needs to make many decisions without regard to whether they are consistent with their baby’s wishes or not. This is not due to a felt moral imperative that might prompt caregivers to sometimes disregard older children’s express wishes, but due to practical inabilities to ascertain what infants believe and want. We do not need to appeal to infants’ status as “children” to explain why we can justifiably act towards them without their consent in instances where an adult’s permission would be required. The inability to communicate with specificity makes caring for infants more closely analogous to caring for comatose patients or kittens than caring for older children who can explain what they understand and what they’ll agree to. The conventional grouping of infants and 12-year-olds together as “children,” and 19-year-olds and 42-year-olds alike as “adults” might nudge our intuitions in ways that would not otherwise make sense.

2.B “No Choices” Arguments: Childhood is Relevant Because Children Lack Capacities Necessary for Morally Meaningful Choices

If the incapacities explanation is not well defended by appealing to distinctly worse consequences of failing to paternalize towards children, might there instead be distinctively adult characteristics that place deontic limits on paternalism? One possibility is that adults have liberty interests in choosing for themselves that children lack. Mill writes that while “no one but [a non-delirious adult] himself can judge the sufficiency of the motive which may prompt him to [knowingly] incur [a dangerous] risk” (Mill 1859: 173) the state is justified in intervening to prevent accidents. For example, if someone is “attempting to cross a bridge which had been ascertained to be unsafe, and there were no time to warn him of his danger, they might seize him and turn him back, without any real infringement of his liberty; for liberty consists in doing what one desires, and he does not desire to fall into the river” (Mill 1859: 172–73).

People might ordinarily have powerful liberty interests in making deliberate choices for themselves in accordance with their will, but people also sometimes act in ways that do not carry the same moral significance and cannot be regarded as expressions of their liberty interests. Mill’s bridge case is one such instance: the man approaching the bridge thinks he is about to cross it and is instead about to plunge into a river. This fact about the bridge is material to the man’s decision: if he knew about the condition of the bridge, he would not step onto it. Seizing the man to prevent him from falling is, then, not a morally meaningful setback to his liberty in the way it would have been (according to Mill) if the man was aware of the risk and wished to take his chances. Someone who sleepwalks towards a cliff edge can be seized without infringing on their liberty because they do not have the basic awareness needed for their walking to count as a morally meaningful choice. If children’s actions could be shown to be the moral equivalent of those of a sleepwalker, then duties to respect liberty that might count as reasons against paternalism towards adults would not apply to children.

Along these lines, Tamar Schapiro argues that paternalism towards children is justified not because children are less proficient in making good choices, but because “children are incapable of making their own choices, whether good or bad” (Schapiro 2003: 575, 581). Paternalism is then permitted towards children because they are not truly “authors” of their actions and so “by interfering with a child’s action, we do not thereby violate her” (Schapiro 2003: 586). The “paternalistic attitude” that adults feel towards children is prompted by children’s undeveloped condition in contrast to that of a “developed human” with a “capacity for reflection” placing them “in a position to give themselves reasons of their own” (Schapiro 1999: 716, 722, 729). “[A]dult[s] qua adults [are] already governed by a constitution—a unified, regulative perspective which counts as the expression of [their] will” enabling them to speak in their own voices whereas for a child “there is no voice which counts as hers.” (Schapiro 1999: 729).

Schapiro explains that “[o]n Kant’s view, nonhuman animals are governed by nature’s law, a law that is embodied in the animal’s instincts” whereas humans are distinct in “our capacity to question nature’s authority” such that we “substitute a law of our own making” for “nature’s law.” This is a task which “finds completion in ‘personality,’ our capacity to act on a law which in no way derives its authority from instinct” (Schapiro 2003: 587–88). For Schapiro, the “concept of an adult” is simply that of a “full person” “who has completed the task of liberation from nature” such that adults are in “a condition of autonomy” where their actions are attributable to themselves in the “authorial” sense, rather than their instinct. In contrast, “a child is one who has yet to complete the same task,” leaving children as “not full persons” in conditions of “as-yet-incomplete liberation from nature’s rule” (Schapiro 2003: 575, 588–89). Adult paternalism towards children is then justifiable for Schapiro because children without paternalism are still subject to an alien rule by their instinctual “animal nature” (Schapiro 2003: 590–92).

Schapiro is careful to specify that what she defines are status concepts of “adults” and “children” and not biological kinds—but it is far from clear whether these statuses are accurately attributed to the people ordinarily called “adults” and “children”—a distinction commonly drawn according to age alone. It is also unclear if Schapiro’s concepts map even roughly onto the sets of people we refer to as “adults” and “children” in ordinary speech. Schapiro begins with the observation that adults feel distinctly entitled to paternalize towards children, recognizes that this requires justification, and proceeds to describe what children would need to be in relation to adults for this to make sense in Kantian ethics (Schapiro 1999: 715–18, 720, 726–32). If children and adults in the world really did match Schapiro’s status concepts, this would be a powerful argument for the greater permissibility of paternalism towards children. If, however, Schapiro’s models are tailor-made to justify paternalistic attitudes towards children,¹⁶ they may be only idealized images of children and adults that neither live up to in reality.

This possibility is suggested when Schapiro acknowledges that, strictly speaking, no one “is able to achieve autonomy on Kant’s view” and notes that “the applicability of the moral law depends upon our mapping these ideal concepts onto ourselves and one another…despite the fact that perfect realizations of autonomy are nowhere to be found,” leading her to consider whether Kant understood the distinction between adults and children as degrees of imperfect autonomy (Schapiro 1999: 723). Schapiro rejects this possibility: “if children are simply adults in a less cultivated form” this would “run counter to the intuition that adults have a special obligation to raise children, whether the children like it or not” (Schapiro 1999: 724).

The central problem with Schapiro’s “attributability” argument, and similar arguments, is that they rest on the claim that some actions (typically those of adults) morally belong to the agents performing them, whereas other actions (those of children) do not count as morally meaningful expressions of their will—but fail to provide viable standards delineating those actions that morally count this way from those that do not in a manner that maps onto adult/child distinctions. The standards are either too high, excluding most or all children, but on reflection also excluding most or all adults, or too low, including most adults, but also including all but the youngest children, or too vague.

Making decisions consistently according to Kantian ideals of autonomy is not a standard any adults will be able to meet. Widely cited research in moral psychology suggests few people ever reason this way. Lawrence Kohlberg developed an account of six stages of moral development, beginning with two preconventional levels, interpreting morality according to hedonic and prudential standards, through two conventional levels, concerning pleasing others and respecting social rules and culminating with two postconventional or “autonomous,” “principled” levels, the lower of which Kohlberg termed “the social-contract legalistic orientation” and the higher “the universal-ethical-principle orientation” (Kohlberg 1973: 631–32). During a 20-year longitudinal study of moral judgment development, Colby and Kohlberg decided to drop their Kantian-influenced universalist sixth stage of development because none of their subjects, including “upper middle class” college-educated 36-year-olds adopted “stage 6” judgments (Colby et al. 1983: 5). By age 36, only 62% of Colby and Kohlberg’s subjects adopted a “stage 4” ‘law and order’ orientation towards conventional fixed rules and fewer than 10% ever adopted “stage 5” postconventional rights and standards-based judgments (Colby et al. 1983: 46). If truly voluntary action requires reflectively adopting one’s own principles of self-governance, this is a standard only a small portion of adults meet. On the other hand, if morally meaningful action only requires giving rationales for moral judgments—the youngest children in Colby and Kohlberg’s studies did that.

If “we conceive of children as only partially free from the governance of instinct” (Schapiro 2003: 590) and define adults as those completely freed of instinctive governance, then all humans would seem to be defined as children not adults. As Kahneman famously described, instinctive, automatic, unreflective judgment and decision is a pervasive and normal feature of adult psychology (Kahneman 2012). Major life decisions are often made without articulable reasons but on the grounds that they just feel right. While a “no instinct” standard will exclude all adults, a view that behavior counts as volitional in morally meaningful ways just as long as the actor exhibits any “ordering of impulses” (Schapiro 1999: 730) would seem to include all but the youngest children.

Other standards fail to distinguish adults from children neatly because their application to real cases is underdefined. Rawls, for example, writes that “while individuals presumably have varying capacities for a sense of justice, this is not a reason for depriving those with a lesser capacity of the full protection of justice. Once a certain minimum is met, a person is entitled to equal liberty on par with everyone else” (Rawls 1999: 411–12). Rawls seems to regard some low-threshold sense of justice as morally necessary for liberty without providing enough details to determine who meets the standard and why.

If we consider what capacities might really make the difference between choosing badly and not making morally meaningful choices at all, without concern for whether these separate children from adults, the capacity for consciousness and language would be obvious candidates. Factual misunderstandings and mistakes are another, as in Mill’s bridge case. Hallucination and psychotic delusions might also cause such total misunderstandings that a person acts in ways that do not reflect what they intend. The ways children tend to lack understanding are less comparable to these cases and more similar to making factual mistakes in the way ordinary adults do.

3. Defending the Incapacities Explanation Without Relying on Categorical Differences Between Child and Adult Capacities

If the incapacities explanation will not provide a satisfactory justification for the childhood relevance view categorically, it might nonetheless be argued that age-based classifications are useful and necessary proxies for capacities, and rule-making and administration requires establishing broad classifications (Feinberg 1986: 326). Such a response might acknowledge that although there is no generally relevant division of capacities between children and adults, since we must make policy decisions regarding people as classes, it is reasonable to make them informed by the reality that adults will tend to be relevantly capable in far more domains than children. In a non-ideal context of limited information, we might best approximate minimum necessary levels of paternalism by paternalizing towards children in many cases where we would not paternalize towards adults.¹⁷ An inquiry into each person’s exact capacities is administratively impossible in practice, so even if individuated evaluation and treatment is ideal, there is no injustice in categorically treating adults as presumptively competent and children as presumptively incompetent for many questions of paternalism.

While real-world individuals, governments, courts and institutions lack perfect knowledge, they can and do often develop significantly more individuated knowledge about specific people. There are also proxies for capacities that are more accurate proxies than age. When considering how to treat patients refusing medical treatment, or people who, absent intervention, consistently eat unhealthily, or who risk death from of alcohol abuse, real world actors often have more precise information to rely on than just the subject’s age. When creating rules that treat people differently, using broad proxies that are over and under-inclusive is normally thought to be less justifiable than using more direct information.¹⁸ The knowledge that, regardless of what is typical for their age, particular people refuse medical treatment, eat junk food exclusively, or binge drink, can provide more direct indications that they could not ensure their welfare in those areas. In this way, while administrability concerns might excuse some age-based categories in paternalistic laws, they do not resolve the dilemma: such concerns cannot provide an explanation for why, when it is actually known that an adult is unable to adhere to a healthy diet, it is still felt morally wrong to coerce them into following one, but it would be felt morally right to coerce a child for the same purpose.

Even when considering adults as a class, there are many cases where it seems to be the norm for adults to be unable to choose in favor of their welfare interests, but most think it their right to make the ‘wrong’ choice regardless—even while simultaneously believing that children should be prevented from making the same wrong choice. For example, many believe that most adults cannot self-regulate occasional cigarette use so as to avoid adverse health effects, but, in line with the childhood relevance view, it is still felt reasonable allow adults to damage their welfare with cigarettes while prohibiting children from doing the same. To use another example, the fact that most adults who go on diets to lose weight ultimately regain that weight (Anderson et al. 2001) suggests that most adults weighing more than they would like lack the executive capacity to lose weight when they want to. So, although it is widely believed that adults are usually unable to lose weight, and that failing to lose weight can harm severely overweight adults, the idea that this situation might be remedied through coercively controlling adult diets would strike most as an unjustifiable afront to adult liberty and dignity. The case of a paternalistic agent closely managing a subject’s diet is, however, frequently cited as a paradigmatic case of justified paternalism where the agent is a parent and the subject is their child, (e.g. Husak 2010: 108) so it is not the situation or remedy but the subject that is objectionable.

The objection that age-based proxies for capacities are administrative necessities in the real world is also somewhat undermined by existing practices of individuated assessment of individuals’ specific capacities—but opportunities for such assessments are generally only afforded to people above a certain age. Driving a car is both a ubiquitous and high-risk activity, but governments still make individuated determinations about who has the capacity to drive safely through exams. Determining that an elderly or intellectually disabled adult lacks the capacities to be legally competent to direct their own affairs likewise requires an individuated adjudication through guardianship hearings, and guardianship law allows for guardianships limited to specific, individuated incapacities (Cassidy 2015).

These observations would seem to suggest that it is not that age is used as a proxy for capacities because it is the best proxy, but because most take for granted that the adult/child division is a normatively permissible basis for classification. Replacing age-based classifications with more precise proxies would likely strike many as objectionable if doing so would argue for more paternalism towards adults or less towards children. This implies that justifications from administrability only work to support the childhood relevance view if a categorical incapacities explanation is already assumed.

The possibility that parents may have special obligations to paternalize towards their (typically less capable) children might also be thought to provide a basis for the childhood relevance view. Husak, for instance, suggests that part of what makes a parent’s paternalistic regulation of their child’s diet justified is that parents occupy an especially appropriate role for paternalizing towards their children (Husak 2010: 108). If these special duties terminate on adulthood, then adults might not be proper subjects for paternalism because no one has a duty to paternalize towards them.

Many scholars have written extensively on parental duties to children, prescribing a wide variety of duties and corresponding rights (e.g. Alstott 2005; Gutmann 1980; Kolodny 2010; Eekelaar 1991). Others have written extensively on the interests of parents in exercising control and influence over their child’s upbringing (e.g. Altman 2011; Brighouse and Swift 2006; Schoeman 1980; Yaffe 2018: ch. 6)—interests which are distinct from paternalism in the pure sense considered in this paper, but which implicate authority over children in ways often termed ‘paternalistic’ even if the interests vindicated run in part to the parent rather than the child. A full examination of this literature is beyond the scope of this paper. Whatever special duties parents owe to their children, the particularities of the scope and duration of parental duties are likely to be at least in part socially, legally and historically contingent.

That parents do in fact have legal and social obligations to paternalize towards their minor children but not to their adult offspring (however much they may want to) is a legal and social reality. This reality does not answer the question of whether or not parents should have these special obligations to paternalize to their underage children, but not their adult offspring, or what the extent and limits of these obligations should be. We can acknowledge that parents do have social, legal, and felt obligations to paternalize to their children without viewing this as a reason for thinking that they ought to have these obligations—to argue otherwise would be to adopt a naturalistic fallacy.

There may be reasons to think that parents are especially well-suited to paternalize towards their children, such as the presumption of affection and knowledge of a child’s needs. However, there may also be reasons to think that parents often have conflicts of interests and needs that do not align with those of their children.¹⁹ At best, arguments that appeal to parents’ special duties answer the question of who should paternalize towards children, not when paternalism towards children is permissible.

Special parental duties must be thought to terminate at adulthood if they are to support the childhood relevance view—otherwise the distinction would not be between children and adults but between those with parents and those without. Childhood is thought to be categorically relevant to paternalism whether or not a child has available parents, though, and the belief that these special obligations duration is limited to childhood seems to presuppose the incapacities explanation. This would suggest that appeals to special parental duties do not provide much justification on their own unless their duration is explained on non-capacities-based grounds.

4. Some Possible Implications

If the incapacities explanation for the childhood relevance view is more difficult to defend than it initially appears, this has a number of possible implications. This section will consider some of those implications and suggest why they might be difficult to reconcile with other widely held commitments.

4.A The Primary Harms of Paternalism are Found in Coercion Itself, not Loss of Choice

If neither consequentialist nor deontological accounts of adults’ distinctive decision-making capacities neatly divide adults from children—what is at stake in paternalism? Why is it so profoundly objectionable to adults? All choices are already limited by the social, environmental and physical constraints on our actions, and our desires are shaped by our circumstances and unchosen features of our psychology, such as framing effects, adaptive preference formation, and sublimation.²⁰ Yet, we still feel that we have a powerful interest in being able to make choices, and a sense of injury when choices are taken away from us.

If adults’ choices are neither distinctly interest-promoting or distinctly voluntary, then the harm in paternalism might not arise primarily from losing choices but simply from being subject to force or coercion. The subjective experience of anguish, humiliation, frustration and/or degradation at having one’s will deliberately over-ridden by another is itself experienced as a profound afront. This is most vividly felt with regard to impositions on one’s body. A person may experience these injuries attendant to coercion and force without losing out on the ability to exercise choices in an especially autonomous or beneficial way. If the chief harm of force and coercion is subjective anguish and degradation, not the loss of an objective capacity to make reasoned choices without constraint, then coercion and force implicate interests that do not depend on decision-making capacities. This undercuts the sense that coercion does not substantially set-back children’s interests—even if these interests are not always dispositive, paternalism carries ‘costs’ common to both children and adults.

4.B Weighing Child and Adult Interests Unequally

If the question of whether or not an instance of paternalism is thought to be justifiable rests in part on how the costs and benefits are weighed, then the childhood relevance view might imply a tendency to weigh these costs and benefits differently for adults and children. This should worry egalitarians who insist that each person’s interests must carry the same weight absent special justification otherwise.

First, the non-welfare interests being set back by coercion, such as interests in subjective autonomy, bodily integrity, avoidance of humiliation, and so on, may be valued more highly for adults than children. This seems to be implied in accounts that view paternalism for children as ‘cost-free’ in the sense of conferring benefits without incurring costs—whereas for adults, the humiliation, degradation, and loss of status equality attendant to coercion must be weighed against its benefits. This may involve a lapse of empathy for children in failing to take adequate account of their experience of humiliation and anguish from force or coercion. However comparatively unfree children are given natural inabilities and socio-legal constraints, they, like adults, have a range of freedoms to act within their constraints and, like adults, experience resentment when those constraints are narrowed.

Another possibility is a heightened concern for the welfare interests of children as compared to those of adults. It may be that when children are harmed in certain ways, it is felt to be a graver tragedy than when adults suffer the same harms. As such, it might be felt tolerable to allow an adult to choose against their best interests and suffer the consequences, but felt intolerable for children to do so because their suffering is less tolerable to us. This would be consistent with the (possibly apocryphal) practice of saving ‘women and children first’ in disasters, the tendency of the media to report child deaths separately from adult deaths, the greater willingness to provide government assistance to deprived children than equally deprived adults, and the way statistically minimal risks to children are often regarded as more politically or morally dire than statistically greater risks to adults (Stickler et al. 1991: 522–28). Some of these cases may have other explanations, such as adults having greater ‘responsibility’ for their circumstances and ability to help themselves, but not all of these cases can be explained this way.

A third explanation is that people tend to more frequently judge the justifiability of the treatment of children according to consequentialist standards while they more often judge the justifiability of the treatment of adults according to deontological rules—perhaps analogously to Nozick’s concept of “utilitarianism for animals, Kantianism for people” (Nozick 1974: 39–41). There are many instances where it is often thought that even if someone makes a decision that is demonstrably bad for them, their right as an adult to decide creates a duty of non-interference that trumps others’ altruistic desires to intervene for their benefit, even if that person is left worse for it (Mill 1859: 144). It is far less common to think that a child’s right to decide on anything creates a corresponding duty to respect that decision when it results in demonstrably bad consequences. Likewise, people often perceive positive or negative outcomes as morally decisive for questions concerning how to treat children in cases where outcomes are not sufficient to determine how to treat adults. It is, for example, a cliché when debating the permissibility of a parenting practice to assert “my parents did it, and I turned out okay,” suggesting a shared understanding that the appropriate moral measure for how children should be treated is not based on their contemporaneous rights, but on what consequences the treatment has for their future selves. There is no widely shared sense that the morality of an act towards an adult is overwhelmingly determined by its impact on their distant futures.

These three explanations of the thinking that might lead to the childhood relevance view are mutually compatible, and some combination might account for the intuition. It is plausible that people comparatively overvalue certain welfare interests in children, while comparatively undervaluing other interests. It is also plausible that people think about some set of these comparatively overvalued adult interests as rights that preclude competing considerations in a deontological sense, while evaluating how children should be treated according to consequentialist considerations (evaluating their long-term welfare as hugely significant but their contemporaneous desires as unimportant).

All three accounts, however, amount to extending unequal concern for the symmetrical interests of children and adults. In the first two accounts, some interests are treated with greater respect when possessed by adults, other when possessed by children. In the third account, some adult interests are treated as duty creating rights while few if any child interests are accorded such status. Each of these accounts is potentially in tension with the widely held moral principle that people’s interests should be accorded equal consideration that a general commitment to egalitarianism or liberalism favors.²¹ We should then have reason to think that the childhood relevance view could be a case of common judgments about specific moral cases involving children failing to cohere with widely shared general principles of justice. Seeking an internally consistent moral framework would seem to require either rejecting beliefs central to egalitarianism or rejecting different standards for paternalism with regard to children and adults—unless an alternative explanation can be found.

4.C Treating Persons like Less Than Full-Persons

If Schapiro is right that the intuition that paternalism is permitted towards children implies that children are regarded as less than full persons able to provide their own reasons that demand respect, and there is no categorical distinction that explains how someone with Kantian ethical commitments can carve out a category of less-than-full-persons—then the childhood relevance view would seem to run the risk of doing a grave disrespect towards children. If part of the intuition about children and paternalism is that children can be treated according to consequentialist ethics with reasons supplied by others, while adults must be understood as generating their own reasons, this seems to be indicative less of thinking that adults’ reasons are in their best interests, as much as a belief that adults are entitled to value that which is contrary to their interests as ordinarily understood. Genuine self-sacrifice, for example, provides a paradigmatic case of choosing other values at the expense of one’s own welfare interests. This is often regarded as noble for adults, almost never for children. People often value a variety of moral, relational, aesthetic, religious or political commitments above their own welfare interests, even when taking into account the hedonic satisfaction of promoting one’s values in the world.

If the common intuition is that adults are morally entitled to choose to vindicate these other values at the expense of their welfare interests, but children are not so entitled and may be coerced for their benefit at the expense of their values, then this intuition would also require an explanation if it is to be made compatible with extending equal concern and respect to children. To assume without explanation that adults’ value judgments are worthy of more respect than those of children would seem to be at least facially incompatible with this egalitarian commitment.

If it is argued instead that adults’ value judgments are reasoned and children’s value judgments are not reasoned, this would seem to exclude from consideration the large range of adult value judgments that reflect personal desires, judgments and preferences without reference to reflectively determined principles. It is also difficult to imagine a standard that will neatly separate the most compelling reasons that children provide from the least compelling reasons that adults provide. Adults are rarely thought to require such reasons or to owe fully rational explanations for choosing for themselves to anyone, even when they undermine their best interests as judged by themselves and others. The ability to meet high standards of rationality and deliberation is treated as if morally relevant when explaining why children are not entitled to their liberty interests, but philosophers, politicians and jurists are rarely so exacting about what standards of rational decision making must be met when considering the importance of adults’ interests in liberty.

This is not to say that we ought to accept Schapiro’s model, however. We might be able to respect persons as persons by trying to aid them in cultivating their reflective capacities to form and pursue values rather than taking their stated position at any given time as their final word. This might be an essential part of caretaking and education guided by the assumption that children are persons who must grapple with the question of what is important and why.

We might also be justified in sometimes interfering with people paternalistically when the stakes are high, the imposition minimal and we have good reason to think that their acts or omissions do not reflect thoroughly considered judgments. Most think it justifiable to prevent other people, even adults, from impulsive suicide. Doing so does not treat them as a mere means rather than an end since they are the intended beneficiaries. It also need not disrespect their capacity for self-governance in the global sense. Since most people planning suicide change their mind on reflection, taking a firearm away from someone preparing to kill themselves would seem like a proportional paternalistic intervention with a low potential for degradation and a high likelihood of enabling their considered values. My argument does not imply that paternalism should never be considered, but that the reasoning explaining why certain instances of paternalism are justifiable ought to be consistent.

4.D Objections to Implications

Here, I will briefly address two objections that might have arisen at this point. First, it might be argued that it would be a mistake to think that the childhood relevance view implies a lesser regard for children’s interests in avoiding humiliation. This is because children and adults are degraded under different circumstances: it is humiliating to treat an adult as a child but not a child as a child. What would be degrading for an adult is not degrading for a child.²² Usually, however, if something is regarded as degrading for one person but not for another, the reason for this depends on their differing subjective views or experience, such as the presence or absence of consent, desire, or humiliation. Children frequently desire and agree to activities that would be humiliating for an adult, but these activities would be humiliating largely because these are subjectively unwanted for adults. Often, children expressly do not want to be treated like a kid.

If an act is unwanted and subjectively humiliating for an adult, this usually counts (in both public and academic discourse) as a very strong reason against performing that act. This is even more the case if the act is intended for the adult’s own benefit rather than the good of society or a third party. In contrast, that something is unwanted and subjectively humiliating for a child is frequently not regarded as an especially compelling reason not to coerce them into it, if it is thought to be for their own good. What a child wants can, for example, be socially dismissed with affirmations such as ‘if you’re the parent, you decide.’ It likewise goes without social reprobation to regard a child or adolescent’s “embarrassment” as amusing non-harms, or even further evidence of their immaturity and unreasonableness. This is true even in cases where it would be reasonable for an adult to likewise find the same acts unwanted and humiliating.

Second, it might be argued that valuing children’s welfare more highly than adults is not necessarily a problem for egalitarian commitments. Children literally have more to lose than adults from death because they have longer remaining life expectancies (Liao 2010: 173). If this accounted for the intuition consistently, however, then the degree of concern over a person’s death should be linear. Instead, concern for people’s welfare interests seem to sharply drop off at the onset of adulthood. The heightened welfare consideration extended to children could be interpreted as cases of treating children better than we treat adults. Less charitably, this could be interpreted as thinking of children as precious possessions rather than equal persons. The thought that we should be especially protective of children because they represent the future, for example, glorifies children in an abstract way but also objectifies children as vehicles for social interests rather than persons with their own lives to live.

5. Conclusion

The view that it is generally permissible to paternalize towards children in cases where it would not be permissible to paternalize towards adults is implicit in nearly every domain of social, political and family life. It is a premise that the structure of our society relies on, without which much of it would be unjustifiable—this is most obviously seen in child–parent relations, but also in education, economy, civic life and the scope of state authority. Pervasive disparities in rights and treatment demand justification. This paper’s primary contribution is to argue that the chief justification for this disparity is not as defensible as generally assumed.

The belief that childhood is relevant to the standards according to which paternalism might be justified is standardly explained in reference to children’s lack of relevant capacities possessed by adults. This paper argues that it is true that capacities adults are more likely to possess are relevant to the justifiability for paternalism. Doing so, however, also led to the discovery of a substantial dilemma. Capacities reduce the reasons that motivate paternalism when they enable a person to enjoy a benefit or avoid a harm that might otherwise prompt paternalism—but when particular adults lack the capacities relevant to the same benefits or harms that prompt paternalism for children, an adult’s other capacities do not explain why the benefits and harms to them should not likewise justify paternalism.

The paper next tested whether this dilemma could be resolved by prior authors’ consequentialist and deontological incapacities-based explanations of the childhood relevance view. I argue that while initially compelling, neither consequentialist nor deontological accounts hold up to scrutiny because they do not identify workable and non-arbitrary grounds for judging paternalism towards children according to less demanding justificatory standards. Arguments appealing to administrability and special obligations can supplement a successful incapacities explanation, but they do not supply an independent defense of the view.

If the childhood relevance view is not as clearly justified as generally believed, this, I argued, would have a number of concerning implications. First, the harm paternalism might pose to adults has been improperly construed as depriving adults of autonomous or beneficial choices. Absent capacities that make adults’ choices distinctively autonomous or interest-promoting, and in light of the reality that choices are always constrained, a more plausible principal harm is the subjective experience of humiliation and anguish of having one’s will over-ridden for one’s own good. Children can experience this harm too. This then implies that the childhood relevance view relies on relatively undervaluing the injuries to children attendant to coercion, relatively overvaluing children’s welfare interests, or conceptualizing children in consequentialist terms while understanding adults as possessing deontic rights, or some combination thereof. It also implies a failure to recognize the moral relevance of children’s values, which, if unjustified, should be of great moral concern. Without the incapacities explanation, or a viable alternative, different standards for justifying paternalism for children are very difficult to reconcile with widely-shared commitments to equality and equal concern and respect among persons.