Children ’ s Vulnerability and Legitimate Authority Over Children

Children ’ s vulnerability gives rise to duties of justice towards children and determines when authority over them is legitimately exercised. I argue for two claims. First, children ’ s general vulnerability to objectionable dependency on their caregivers entails that they have a right not to be subject to monopolies of care, and therefore determines the structure of legitimate authority over them. Second, children ’ s vulnerability to the loss of some special goods of childhood determines the content of legitimate authority over them. My interest is in the so-far little-discussed goods of engaging in world discovery, artistic creation, philosophical pursuits and experimentation with one ’ s self. I call these ‘ special goods of childhood ’ because individuals, in general, only have full access to them during childhood and they make a distinctive and weighty contribution to wellbeing. Therefore, they are part of the metric of justice towards children. The overall conclusion is that we ought to make good institutional care part of every child ’ s upbringing.


Introduction
This article brings together two different issues in just childrearing, shows that they are related, and argues that they have a common solution. Both have been receiving increasing philosophical attention, but have so far been treated separately. The first problem is that children's dependence on some adult's care makes them generally vulnerable to abuses of power. Building on previous work, 1 I argue that children's general vulnerability entails that they have a moral right to not be subjected to monopolies of care, meaning that they have a moral right to receive care from more than one source -'one source' being one adult or group of adults whose interests are closely intertwined. This is a claim about the structure of legitimate authority over children, and it states that authority ought to be divided between several authority holders. The general form of the argument is: 1. Monopolies of care are, pro tanto, illegitimate. 2. Current child-rearing practices give parents a monopoly of care over children. 3. Parents' monopoly of care cannot be justified by appeal to the children's interests. 4. Therefore legitimate authority over children must avoid monopolies of care. 5. The best way to avoid monopolies of care in current societies is to supplement parental care with good institutional care.
The second problem is generated by the fact that children are vulnerable to adults not only for the provision of goods that are generally necessary for flourishinglike food or safetybut also for the provision of the so-called intrinsic, or special, goods of childhood. Children are able to derive great value from certain goods that can be enjoyed either exclusively or mainly during childhood. These goods make a distinctive and weighty contribution to the quality of children's lives and, it has been argued, are part of the metric of justice towards children. 2 Children's vulnerability to the loss of these goods partly determines the content of legitimate authority over children. The general form of the argument is: 6. Children are owed adequate access to the special goods of childhood. 7. Therefore the exercise of legitimate authority over children must give children adequate access to these goods. 8. The best way to give all children adequate access to these goods in current societies is in good institutional settings.
The common conclusion of the arguments is that: 9. We ought to make good institutional care part of every child's upbringing.
I discuss each argument in the next two sections.
Objectionable Dependency and the Structure of Legitimate Authority Over Children Children are incapable of defending their own interests. To survive and thrive, they unavoidably depend on, and therefore are vulnerable to, adults for the satisfaction of many needs: some material, some emotional 3 and intellectual, and some having to do with children's development into autonomous individuals able to make authoritative decisions about their lives. 4 Because they lack full autonomy, children are justifiably subject to paternalismadults have a moral right to exercise authority over them. Moreover, adults also have a duty to exercise authority over children since children's flourishing depends on paternalistic adult behaviour. But children's pervasive vulnerability means that authority over them must be exercised in ways that best protect their morally weighty interests. 5 Some believe that children's dependency and hence vulnerability are in themselves regrettable. 6 However, children's dependency makes possible the unique intimacy and wholehearted love between children and their rearers. Children's dependency includes their interest in continuity in care and in close knowledge of their needs; hence, loving a child well requires from the rearer a long-term commitment to be part of the relationship and an unusually high level of attention to the child's need. These features of good parental love are, in turn, likely to make it unusually deep and robust. This may render children's vulnerability valuable all things considered. Yet, the current level of dependency is not necessary for preserving intimacy and wholehearted love in childrearing. Because children's dependency involves significant risks 7 we ought to minimise dependency as far as it is compatible with ensuring that children's interests are met. Children's enjoyment of some non-parental caring relationships need not erode the love and intimacy between children and parentsas we know from numerous cases when children have been lucky to have enduring close relationships with adults other than their parents. Robert Goodin's account of objectionable dependency 8 is particularly helpful for explaining why justice requires a radical change in the way in which we organise childrearing. In his analysis of dependency as a ground for moral obligation, Goodin argues that it is not dependency as such 9 but the risk of abusing power over the dependent that is intrinsically undesirable. According to him, dependency relationships exhibit a particular combination of features that, together, generate an unacceptably high risk of abuse: first, the relationship is asymmetrical in terms of parties' power over each other; second, the dependent party has a vital need for the resources provided by the other party; third, the superordinate party exercises discretionary control over those resources; and, finally, the relationship in question is the only source of such resources for the dependent party. 10 The monopoly generated by the combination of these features is morally unacceptable. Goodin identifies exploitation as the specific problem of monopolies of power. In applying his criticism to parenting, I do not take a position on the question of whether parental monopolies of power are exploitative. 11 If relationships are morally unacceptable when they display certain features, then childrearing arrangements that exhibit these features are illegitimate (unless such degree of dependency was necessary for advancing children's interestwhich, I shall argue, it is not).
Parenting as we know itthat is, the status quo as well as parental rights throughout history and different societiesexhibits all these features. And yet, only the first three must apply to all forms of childrearing. Firstly, children have a vital need for care. Secondly, asymmetry of power in the parent-child relationship is desirable because without it adequate care for children may be impossible. If children had the same level of power as the people who care for them they could resist receiving the care that they need. Therefore, asymmetry of power between children and their parents is legitimate.
Thirdly, some discretionary control of the parent over the child characterises parent-child relationships as we know them. This feature may be necessary in order to preserve the intimate nature of parent-child relationship which is valuable to both parent and child: if all the interaction between parent and child was conducted according to public rules meant to protect the child's interests, parents' own needs and desires could never find expression in the relationship. 12 This, in turn renders legitimacy to some parental discretion in their use of power in relationship to the child. In addition, a certain degree of discretionary use of the power that child-rearers have over children cannot be eliminated as long as children are raised by human beings, whose needs and desires are also at stake in the relationship with children, rather than robots. 13 Because it would be undesirable to have children raised by robots (for reasons that go beyond children's need for intimacy), it would be undesirable to strive to eliminate all discretion from childrearing. It is true that in many contemporary societies there are policies and institutions dedicated to the protection of children which limit the discretionary use of parental power. Their main function however is to prevent, detect and put an end to parental neglect and abuse. They are safeguards only against the very worst forms of abuse and they are highly fallible safeguards. Moreover, child protection policies try to mitigate the worst consequences of power abuse but cannot, nor should they, fully eliminate the discretionary control feature and, with them, various failings of care. 14 Even if childrearing was more strictly regulated, child-rearers would have to decide on numerous details about how to do things with and for them, decisions in which children, for good reason, have no authoritative say.
The monopoly of care over children is neither unavoidable nor desirable. Rather, it is a consequence of giving parents the power to exclude other adults from forming and maintaining caring relationships with the child. In some countries, parents are permitted to be the only source of regular and reliable care for their children. This is the case in countries where there is no compulsory attendance at care-giving or educational institutions: that is, everywhere for very young (pre-school) children, and wherever home-schooling is legal. In other countries, legislation limits the parental right to exclude all others from the childbut only after the child reaches school age and is required to enrol in (some kind of) school. Even in these cases, parents typically can exercise control over who are the individual people who provide the child with care. In countries where school attendance is compulsory, parents can choose schools or, under the more restrictive legislation, avoid particular schools by moving house. Moreover, parents can and often do control their children's caring relationships with others. A parent has the last say with respect to which people associate with their children, and their legal right to exclude particular others does not depend on how these associations benefit the child. For instance, a parent has the legal right not only to fire a long-standing, caring nanny, but also prevent herif the parent so wishesfrom continuing to see the child and cultivate their relationship. The same applies to parental control over caring relationships with friends, neighbours and members of the extended family. 15 Now, it is true that most children have two parents, rather than one, and this can mitigate, to some extent, the monopoly of care. 16 However, many caveats apply. First, significant numbers of children are raised by single parents. Second, even in legally intact families some children in fact receive hand-on care mainly or only from one parent. Third, cohabiting parents who raise children together display, at least in theory, a unity of interest that make them, as a couple, hold the monopoly of care over the child. Children cannot exit, even temporarily, the relationship with one parent in order to seek better care from the otherat least as long as parents function as a unitary parenting couple. Co-parents may and do sometimes correct each other's mistakes in the relationship with their children; but they can also bein virtue of their unity of interestmore likely than other individuals to enable or at least protect each other's shortcomings as caregivers. To sum up, existing childrearing practices display a monopoly of care.
Goodin indicates two strategies for addressing morally objectionable forms of dependency. One is to ensure that parties can, if necessary, defend themselves against each other. This is possible only when one of the first two features can cease to apply that is, when basic needs are no longer involved or the relationship can become symmetricalsuch that each side can either withdraw or else 'give as good as it gets in any fight'. This clearly cannot be the case with parents and children. The other strategy is 'to try to forestall the threat of exploitation rather than merely trying to defend against it' by 'depriving superordinates of discretion in the disposition of needed resources. That is by far the best way to deprive dependencies of their moral sting'. 17 What should be dismantled in order to render children's relationships with those who give them care morally acceptable is the monopoly feature, since 'as long as the subordinate party can withdraw without severe cost, the superordinate cannot exploit him'. 18 As long as they depend exclusively on the care of their parentsor any other unique providerchildren cannot withdraw without paying exceedingly high costs. Even when they have several adults to whom they can reliably turn to for care, children need not be able to fully exit the relationship with their own parents (and, especially young children, are unlikely to wish to do so.) But they would nevertheless gain the possibility of temporary exit: when significant breakdowns of care occur in a relationship, even when such breakdowns do not amount to serious abuse or neglect, the child has a fall-back option in place. Therefore, the presence, in the life of the child, of several caring adults who are independent from each other and who can both support and supervise each other, would make children's dependence on their adult caregivers less objectionable. 19 There are several ways in which we can think about particular policies that would create non-monopolistic ways of providing care for children. Because the monopoly of care and power that parents have over children is protected by parents' legal right to exclude others from the child, this particular legal right has to go. This entails stripping parents of their right to control children's association with other adults who can and do provide care for the child. But this measure would not, in itself, ensure that all children could rely on sources of care additional and alternative to parental care. Goodin's account supports the more demanding conclusion that we ought to ensure that all children are engaged in caring, long-lasting and secure relationships with more than one adult or group of adults belonging to the same family.
The parental monopoly of care could only be vindicated if it was necessary for the optimal protection of children's interests. This, I assume, is implausible. The most extreme failings of careabuse and neglectare easier to discover when children receive some continuous care from other adults than their parents, especially if these adults are trained professionals and the interaction takes place in a public setting. It is possible, and desirable, to educate parents in order to reduce current levels of abuse and neglect. In an ideal world, they would be entirely eradicated. But my argument, like Goodin's, is premised on the assumption that people will always be fallible with respect to abusing their powerno matter how well-trainedand that, in order to protect the vulnerable, we ought to put in place mechanisms that prevent or mitigate such failures. Mandatory attendance of adequate institutional care is such a mechanism, one that significantly advantages the worst-off children.
Further, children are likely to benefit more generally from having long-term caring relationships with adults other then their parents. Not only can they draw on richer caring resources but, plausibly, exposure to different caregivers can help correct the effects of more minor failings of care 20 and increases the chances that all children have at least one competent caring adult in their lives. Empirical research confirms that frequent and stable interaction with caring non-parental adults is beneficial in terms of children's behaviour and general resilience. 21 In particular, we know that attendance of good quality day-care is beneficial for children over age one. 22 Finally, having longterm caring relationships with several adults increases a child's chances to be cared for by someone who loves the child. 23 And if the non-parental care is provided by trained individuals on public settings it can significantly increase children's access to other special goods of childhood (as I discuss in the next section.) The best way to ensure that all children are engaged in multiple caring relationships is therefore to require all children above one to regularly attend good quality caregiving institutions such as day-care and kindergarten and to ensure that schools provide careby employing social workers who must make themselves available to children for emergenciesas well as academic training. To properly provide children with enduring caring relationships these institution must employ enough well-trained staff and ensure low turnover.
My proposal faces two worries. 24 First, if children have a moral right to receive care from more than one adult, then it seems that we collectively are under the duty to ensure that sufficient adults provide such care. But this duty, especially if it is enforceable, seems very demanding. Second, if in order to respect children's rights we were to deprive parents from a monopoly of care they may lack sufficient incentive to have children.
It is true that my views concerning just childrearing are demanding; the first problem is merely an instantiation of the general problem that our duties towards children are very demanding. For instance, it is likely that orphans have the moral right to an upbringing as good as that of other children, which requires committed and loving child-rearers; yet, this is only possible if we are collectively under a duty to provide such child-rearers to the orphans. 25 The duty to provide all children with non-parental caregivers may be enforceable, at least as long as there are legitimate ways of enforcing it. For instance, it may be possible to nudge sufficient people, through financial and non-financial incentives, to work as non-parental caregivers. This, again, seems to be the same problemalbeit on a different scaleas ensuring, non-coercively, that adults work as caregivers when some children become orphans.
The over-demandingness worry is somewhat alleviated by the fact that we haveor, at least, can achievecontrol over the number of children who come into existence. Which prompts the related worry that my suggested practice of childrearing could dissuade people from having children. This worry may or may not be warranted, but it is not a good reason to resist my proposal. Surely, people's unwillingness to procreate or parent under certain conditions cannot show that children do not have the moral rights which, if enforced, would disincentivise their being brought into existence. People may have become less willing to have children when they were forbidden to require their minor children to work for a wage or to apply corporeal punishments to children. But these surely could not be (even weak) reasons to think that children do not have moral rights to be free from corporal punishment or waged labour; nor were they reasons not to implement the respective reforms.
This analysis of children's vulnerability would apply equally to any group of individuals who are dependent on other people's care, who cannot voluntarily exit all caring relationships and whose caring relationships will either unavoidably or desirably contain some degree of discretion (for instance, adults whose mental impairments make them subject to legitimate paternalism). In the next section I turn to a second way in which children are vulnerable, which is unique to them.

The Goods of Childhood and the Content of Legitimate Authority Over Children
Children depend on adults in order to enjoy what they are entitled to by justice; adults' authority over children can be legitimately exercised only if the adults in question provide children with what is owed to them. As Colin Macleod and Lars Lindbolm have argued, standard theories about the metric of justicethat is, about the things the distribution of which is a matter of justicedo not work well in the case of children, to whom it is improper to ascribe full moral responsibility. 26 It is likely that, instead of resources, or opportunities for wellbeing, or capabilities, we owe children those goods that will make their lives go well while children and as future adults. In this case, perfectionism in childrearing need not be a problem. 27 The difficult questions concerning the metric of justice for children, then, are which goods contribute to a good childhood, which goods contribute to a flourishing adulthood and how to weight them against each other when necessary trade-offs have to be made between these two categories of goods.
Recent developments in thinking about the nature and value of childhood help with these questions. Several philosophers have been drawing attention to the fact that childhood is not mere preparation for adult life; 28 rather, they think that children can enjoy a number of 'special goods of childhood', understood as goods that are valuable whether or not they contribute to children's development. In particular, Macleod has argued that we owe children those goods that make their childhoods go well. 29 I defended the view that children have privileged but non-exclusive access to a number of such goods and this can make children's lives go better, in some important respects, than adults' lives. 30 The special goods of childhood of interest here are the experiencing of caring affection from adults whom the children can trust and love wholeheartedly and unstructured time during which children engage in fantasy play, experimentation and undirected exploration of the world and of their minds. The enjoyment of both kinds of goods is made possible by some of the features of childhood that create children's general vulnerability analysed in the previous section: that is, their pervasive dependency on adults. Without it, neither the quality of the intimate relationship between children and caregivers nor children's necessary leisure time would exist.
This view is not uncontested. Some believe children's dependency on adult care and their being subject to justified paternalism mean that childhood as such is, all things considered, bad for children. 31 If so, then we should help children overcome this state as soon as possible. According to Sarah Hannan, childhood is an intrinsically undesirable state because children necessarily suffer from various bads, 32 some of which coincide with the facts that make monopolies of care over children problematic: they are unable to meet their needs and are subject to legitimate domination in almost every area of life, 33 they are relatively small and weak, bad at means-ends reasoning and prone to form incautious attachments (these last features make their exit from relationships difficult and costly.) Other bads to which children are subject, as mentioned by Hannan and others, concern their lack of full moral agency: children do not have a fixed practical identity, a 'self of their own'. This, of course, might be the very reason why authority can be legitimately exercised over them 34 the reason why we would not want to do away with all childrearing, even if it raises the above-discussed worries of illegitimate use of power. If the picture of childhood as involving distinctive bads were both true and complete, it would support the belief that legitimate authority over children must be exercised in order to help children leave childhood behind, the sooner the better.
I will argue to the contrary. First, this picture of childhoodwhile possibly widespreadis incomplete and hence not entirely correct. It underestimates those valuable abilities that children possess to a higher extent than average adults. Indeed, these abilities are so important that we value them very highly in both children and adults and we value them even when they do not lead to remarkable accomplishments. The exercise of these abilities constitutes some of the central goods of childhood, necessary for the full flourishing of children. It is crucial to see that the existence of these abilities in children is made possible, at least in part, by some features which also generate childhood bads. In this sense, many of the goods and bads of childhood are two faces of the same coin.
This issue is important both for understanding children's specific vulnerability and for reaching conclusions about the content of legitimate authority over children. If childhood were bad for individuals in ways that are not mitigated by significant goods of childhood, it would be fair to conclude that being in the state of childhood is, other things equal, making the individual worse off. We usually think that being vulnerable and being the worse off come in the same package: vulnerability makes people worse off, and being amongst the worse off usually makes one more vulnerable. Yet, I am saying here that children have privileged access to some very important goods. For this reason, childhood is likely to be a period when our lives can go better, in some respects and possibly overall, than they can go during adulthood. Where does this leave us with respect to thinking about children's vulnerability? Is it possible to continue seeing children as the most vulnerable members of society, and hence in need of special protection, while at the same time believing that, as such, children's lives go well in unique waysthat, indeed, childhood may be one of your best times? I will address this issue after having clarified the childhood goods at stake.
Over the past three decades, several distinct bodies of literature started to challenge the conventional image of children as being mere 'unfinished adults'. 35 This literature comes from philosophy with children, neuroscience, developmental psychology and even art history. According to several scholars working in these fields, children possess, to a much higher degree than usually acknowledged, and to a higher degree than average adults, certain epistemic abilities as well as artistic and philosophical abilities.
Results from developmental psychology as reported by Alison Gopnik indicate that even quite young children are in fact very good at theoretical reasoningwithin the purview of their limited experience of the world. Looking at children's early learning, Gopnik and others came to the conclusion that children's uniquely fast learning is in part explained by their ability to form hypotheses and then test them in light of experience. 36 On this account, toddlers and even babies 'think, draw conclusions, make predictions, look for explanations, and even do experiments'. 37 To use Gopnik and her co-authors' own catchy metaphor, they are 'scientists in the crib'. Beyond the crib, we know that older children's unusual scientific creativity enables them to occasionally make scientific breakthroughs even in times of highly specialised knowledgeproduction. 38 In a similar vein, Gareth Matthews criticised Jean Piaget's standard theory of children's cognitive development for assuming rather than proving children's imperfect rationality. 39 One of Matthews' charges to Piaget is that the latter failed to understand children's questions as philosophical questions. But Matthews', and other philosophers', experience, is that childrenat least pre-pubertyare more capable of asking deep, competent philosophical questions, than adults who do not go on to be professional philosophers. This is because children are more curious, less conventional in their thinking, and more imaginative, making it possible to dream up alternative worlds. Matthews' critics have objected that children's philosophical pursuits do not display the same range of values that makes for good philosophysuch as systematicity, and tenacious engagement with an issue and the ability to see one's reasoning to its last conclusions. 40 However, this can be grantedand, with it, the conclusion that children are not particularly good at philosophical accomplishmentswithout denying that they posses the uniquely philosophical ability of asking the right question.
Not only Matthews but also art historians 41 and artists themselves (for instance, Pablo Picasso) attribute to children artistic abilities that are, on average, superior to those of adults. The claim here is that children are, in general, better able to produce aesthetically worthwhile drawings than adults who are not professional artists. It is an open question whether adults could (be educated to) retain children's superior abilities. Even if they could to some degree, age-specific neurological development may limit the retention of these abilities (as I explain below). Moreover, the unavoidable responsibilities of adult lifeincluding the responsibility to care for the next generationwill probably always limit adults' opportunities to exercise these abilities to the extent to which they can retrain them. 42 To sum up, although children on average rarely come up with compelling scientific or philosophical answers, let alone theories, and rarely create works of art that are considered of highest merit, they are, in comparison to the average adult, more philosophicallyand scientifically-minded and artistically creative. In Matthew's words: For many people the art or philosophy of their childhood is never equaled, let alone surpassed, by the art or philosophy of their adult lives. If painting or doing philosophy has any non-instrumental value for them, it is their child art and their child philosophy that have such value. 43 Neuoscience explains this discrepancy between children's and adults' abilities by the fact that, in children, the prefrontal cortex is not fully developed, leading to a lower ability to inhibit information and to focus on particular tasks. The disadvantage, for children, is that they have much lower executive abilities than adults; yet, the same neurophysiological feature makes them more open to new experience, curious and creative. Gopnik writes that '[t]o be imaginative, you want to consider as many possibilities as you can, even wild and unprecedented ones . . . In learning, you want to remain open to anything that may turn out to be the truth.' 44 This seems an overstatement, since the process of learning involves eliminating some possible answers in favour others, 45 yet this does not distract from the point that children's openness is conductive to distinctive and valuable learning abilities.
If this account of childhood is correct, then it is not true that children are as bad at ends-means reasoning as they have been traditionally believed to be. Rather than a lack of rationality as such, it is their very limited experience of the world in combination with their lack of emotional self-control that justifies the view that children are insufficiently competent decision-makers and that they are subject to legitimate paternalism. Further, this account shows that children who exercise the creative abilities outlined above have privileged access to some goods that we consider highly valuable in children, as well as in adults' lives: philosophical, scientific and artistic pursuits. Most of us think such pursuits are very valuable whether or not they lead to corresponding accomplishmentsindeed, that they are important enough to make lives that contain them good, in some important respects, on an objective understanding of wellbeing. Interestingly, the goods of childhood discussed here seem to be made possible by the very factors that generate the bads of childhood outlined by Hannan: children's neurological immaturity, in combination with legitimate adult authority over them, meant to protect them from the disastrous consequences of this immaturity. But children's lack of control over their lives at the same time relieves them of much of adult responsibility, thus freeing their time and attention for creative pursuits which adults do not, and cannot, have the same leisure to enjoy. 46 Finally, children's lack of a stable self is often taken to be a source of childhood bads, because it, too, legitimises paternalistic behaviour towards children. Christine Straehle identifies this as the source of children's vulnerability. 47 Yet, the same fact is likely to generate one of childhood's special goods: the ability to experiment with different selvesexperimentation which, plausibly, has intrinsic as well as instrumental value. 48 Children, then, have privileged access to goods that make their lives go particularly well in some respects. The above discussion of the goods and bads of childhood is compatible with several positions and I do not, in this article, commit to any in particular: maybe the bads and goods of childhood balance each other out, such that children are not, qua children, either better or worse off than adults. Or maybe children are, qua children, either better off than adults oras Hannan thinksworse off. I believe that the following analysis of the children's moral right to the special goods of childhood does not hinge on a particular conclusion on this matter. The reason is that, even if the special goods of childhood are not valuable enough to make up for the bads of childhood, some of them have high developmental value: Gopnik and her coauthors believe that children's experimental play is the basis for their future ability to innovate. More generally, the importance of unstructured play for healthy development is sufficiently uncontroversial for paediatricians to issue guidelines meant to protect it. 49 There is similar consensus concerning the importance of children's experiencing trusting and loving relationships with adults who provide care to the children. 50 Therefore, the quality of one's life qua future adult depends on the enjoyment of these goods during childhood. Trying to artificially speed up children's development would, in this case, backfire. 51 Children's adequate access to such goods depends on many factors that are outside their control: most importantly, on being given sufficient time unburdened by specific demands, when they can explore and experiment freely. They also depend on adults ready to encourage their scientific curiosity and artistic creativity and to discuss philosophy with them. (For instance, the emerging field of doing philosophy with children suggests that children tend to be unusually curious, imaginative and ready to ask fundamental questions but also lack the experience and grit necessary to enjoy good philosophical explorations; therefore they greatly benefit from competent adult guidance.) And children depend on adults for access to the material basis of these pursuits.
Because, unlike adults, they are not in control of their time, social interactions and of material resources, children are at the mercy of adults for the enjoyment of these goods. Take, for instance, the example of pretend play that allows one to exercise one's artistic imagination and have fun. This good, I assume, is available to both children and adults although adults may have a harder time accessing it. Adults can and do decide, at will, to engage in pretend play: for example, they can decide to re-enact a day in the Second World Waras it sometimes happens in Englandwith relatively modest resources: their free time on a weekend and a couple of funny hats and handme-down clothes. By contrast, children need adults' permission to play in an unstructured way (rather than study or exercise), to play with playmates (rather than alone), to play in particular ways (even when these are disruptive of adults' comfort), and to use whatever materials they need. Parents themselves may have the will to provide children with the necessary basis needed to exercise children's abilities, yet lack the necessary resources: they may be excessively busy making ends meet (and hence lack the time to supervise them or engage in exploratory conversations), or lack the necessary education, or have little choice but to live in urban environments lacking in public spaces where children can play together. Or they may have enough time, education and infrastructure to provide for their children, but assumecorrectly or notthat, in order to achieve success in overly competitive social circumstances their children must be trained, as early as possible, in the skills that will enable them to be ahead of their competitors: first in educational, and later in working, environments. 52 These skills may, but do not necessarily have to, coincide with the skills that children develop while discovering the world, playing, drawing, dancing and asking metaphysical questions. If a parent raises a child in an overly competitive environment in which all other children are expected to show early academic achievements, it may be rational for that parent to deprive her child of the special goods of childhood. This is especially true if admission to particular schools (or pre-schools!) depends on such achievements.
This last point suggests that making the goods of childhood generally available is a collective problem that affects even parents who wish to, and can, provide their children the special goods of childhood. Each parent may be right to think that, as long as other parents' children spend increasing amounts of their time in training for future competition, it is too risky to give one's child too much unstructured play. Yet, all parents may have reason to wish that enough of their child's time is protected from goal-oriented activities which are allegedly making them more competitive. If so, then the only way of ensuring that all children enjoy unstructured time is to partly socialise their rearing in institutions that provide them with enough such time. Universal attendance at good institutions can undermine to some extent the rat-race of making one's child as competitive as early as possible, partly by diminishing parental influence over their children.
There are other reasons why the special goods of childhood are best provided in communal contexts. First, not all children have easy access to enough other children in their private homes. Perhaps children playing on their own can realise some of the good of imaginative play, but, I assume, playing with other children and engaging in friendships with children is part and parcel of a good childhood. 53 Second, because the special goods of childhood are also more likely to be available to children if they benefit from long-term engagement with adults who are themselves interested in, and knowledgeable about, science, the arts and philosophy. Third, compared to other forms of non-parental care, well-run childcare institutions are more likely to distribute more equally children's access to at least one loving and trustworthy adult.
Therefore, there is a common way to address the two kinds of children's vulnerability. Because children are vulnerable to their caregivers, justice requires that we avoid monopolies of care and ensure that all children have robust access to some non-parental long-lasting caring relationships. Partial, mandatory attendance of daycare centres, kindergartens and schools 54 is the best way to meet this demand. Moreover, it is also the best way to provide all children with the special goods of childhood that are owed to them.
The discussion about children's vulnerability usually refers to the ways in which being a child makes one, at least potentially, worse off. 55 And yet, I suggest, childhood is also a source of advantage because children have privileged access to important goods. This may raise a worry: assuming that the special goods of childhood make children's lives go better than adults' lives, how can children be entitled to them by justice?
Here it is worth distinguishing between two sources of vulnerability. One can be especially vulnerable because one is at higher risk than others of not enjoying an important good that everybody is capable of enjoying equally: a junior person in many companies is more vulnerable than more senior persons to losing her job. Alternatively, one can be especially vulnerable to the loss of a good that only some can enjoy: a talented and passionate pianist is more vulnerable then others to the evils of rheumatism. The second kind of vulnerability can only affect those who are, in some sense, better off.
Children are vulnerable in both ways, and both kinds of vulnerability generate weighty claims to protection. The first kind, the one typically discussed in relation to children, is quite straightforwardly a matter of justice. But additional explanation is required to see why the second kind of vulnerability generates a claim of justice. In the case of the pianist there are consequentialist reasons (like excellence and utility) to protect her exceptional ability. These are not, at least not obviously, reasons of justice. The pianist is very lucky, in the first instance, to have her wonderful and unusual ability; she is, in this respect, more advantaged then the rest of us. It would be unfair to give priority to the protection of her abilities when faced with claims to protection from people who are worse off to start with. The case of children may be thought to be similar, if they enjoy wonderful abilities that are unusual in adults and that make their lives go better in some important respects than the lives of adults. I think there are some good answers to this challenge.
In certain respects the case of children is indeed similar to that of the talented pianist. The goods that children enjoy if they can fully exercise their epistemic and creative abilities merit protection on account of being forms of human excellence as well as for reasons of efficiency, or utility: The full exercise of these capacities is likely to bring joy to children and the adults in their lives. But children's possession of special abilities is also disanalogous with the case of the pianist because we all start life as children but we do not all have the privilege of possessing, at some point in our lives, excellent musical abilities. Thus, reasons of efficiency for giving special protection to children's enjoyment of the childhood goods discussed here are also likely to be reasons of justice: The enjoyment of these goods contributes to a good childhood and has significant developmental value. Moreover, most individuals who have lost their chance to enjoy the special goods of childhood as children have missed their only chance to enjoy them at their best. Hence, our lives go better overall if we enable access to them during childhood. And at least some philosophers will agree that a Pareto optimal equal distribution (according to whatever principle of equality one subscribes to) is more just than a Pareto inferior equal distribution. Justice may prohibit levelling down when there is no gain in equality for doing so. According to the lifetime perspective on justicea controversial yet dominant viewjustice requires some kind of equality between individuals over the entire course of their lives. Therefore, if the lifetime perspective on justice is correct, one need not worry that providing children with the goods of childhood is unfair even if it really results in children being better off than adults.

Conclusions
Children are vulnerable in several ways, both of which bear on the issue of just childrearing. Most of us agree that authority can be legitimately exercised over children, for their own good. Yet, the exercise of authority over another individual puts her at a significant risk that power over her will be abused. Indeed, most literature on justice in childrearing focuses on this very obvious and general kind of vulnerability in children. The usual conclusion drawn by philosophers is that parents are the main holders of legitimate authority over their children, and that limits of this authority must be rigorously defined and monitored with a view to advancing children's interests. Without denying this, I suggest a more radical restructuring of authority over children: to be legitimate, it must avoid monopolies of power in ways that I have sketched in the second section above.
But this does not say much about how authority over children ought to be exercised: children's access to which goods must be ensured by those who have authority over them? Undeniably, children have claims to goods such as food, safety, shelter, clothing, some emotional nurturing and education. Yet, as the recent discussion of the special goods of childhood suggests, there may be goods that are particularly important to children and which are not on this list. Some of them are more likely to have developmental value than others. I have argued that the existence of these goods partly determines the content of legitimate authority over children: those who control children's lives must provide children with robust access to them.
The discussion on both the structure and the content of legitimate authority over children led to the conclusion that justice in childrearing requires the establishment of institutions which provide children with the goods they are being owed and the revision of parental rights to ensure that parents cannot prevent their children from attending these institutions. In the first claim, I join my voice to the view already advanced by Macleod that schools ought, as a matter of justice, to provide children with play and artistic activities.
A final conclusion is that two opposed views on the question of the distinctive goodness and badness of childhood can both acknowledge that children are particularly vulnerable, yet yield different conclusions on how badly or well off children are in virtue of being children. The special goods of childhood may outweigh its bads, in which case rushing children through childhood would be objectionable because it would deprive them from proper enjoyment of these goods. Or the bads of childhood may outweigh its special goods, in which case we would be wrong to rush children though childhood because these goods have significant developmental value. It is fortunate that the debate on whether childhood is, as such, a desirable or undesirable state need