LEAP 5 (2017) Reply to Critics1 L I A M SH I E L DS University of Manchester ABSTRACT In this paper I reply to the response articles in this issue, which discuss my recent book. In addition, I restate the main claims of the book and clarify some of the key distinctions and arguments. Keywords: distributive justice; sufficientarianism; autonomy; education; global justice; parents’ rights 1. INTRODUCTION In this paper I reply to the critical responses to the arguments in my book that appear in this issue. I am very grateful to all of the contributors for their detailed, thoughtful and forceful criticisms of the arguments and to the editors of the issue for their hard work in putting it together. I will not be able to reply to every aspect of their responses in this piece, so I will focus my discussion on the few that I have replies to. I know that I will continue to ref lect on these criticisms and I hope I will find fully adequate ways to confront them, but for now this is my immediately reply. I shall begin by restating some of the central claims of the book and, in particular, set out one distinction that is not made in the book but is important for my replies, and then I engage with each of the responses, beginning with those that respond to earlier parts of the book and ending with those that respond to later parts of the book. 2. RESTATEMENT OF MAIN CLAIMS The book defends the claim that the prospects for sufficientarianism are 1 I am very grateful to Pierre-Etienne Vandamme, Lasse Nielsen and David A xelsen for helpful feedback on an earlier draft. D OI : 10. 310 0 9/L E A P. 2017.V 5.18 211 LEAP 5 (2017) Reply to Critics good and better than has been thought because sufficiency principles have an indispensable and extensive role in our thought. It does so through first articulating the central commitments of sufficientarianism. These are the necessary and sufficient conditions that must be satisfied by some principle for it to count as “sufficientarian”. The plausibility of these, and only these, principles is what determines the prospects for sufficientarianism. I have taken as my starting point the idea that sufficientarianism is concerned with the concept of sufficiency, the idea of having enough, and its important role within an account of distributive justice. If sufficiency is to have an important role in an account of distributive justice it must have normative significance. In other words, it must make a difference to our reasons of distributive justice. In particular, securing enough cannot merely be instrumentally valuable. If it were, then it would be possible to fully state the demands of justice without ever referencing sufficiency. I therefore characterized sufficientarianism as endorsing two claims. The Positive Thesis: We have weighty non-instrumental reasons to secure at least enough of some good(s). The Shift Thesis: Once people have secured enough there is a discontinuity in the rate of change of the marginal weight of our reasons to benefit them further (2016: 34-35). For short-hand I refer to the shift of sufficiency as a change in the nature of our reasons, but for reasons to do with distinguishing it from prioritarianism the technical definition is important. This definitional statement of sufficientarianism may appear to contrast with the commonly used upper-limit definition of sufficientarianism, which combines the positive thesis with the negative thesis. The negative thesis holds that once enough has been secured there are no distributive reasons that apply to benefits and burdens. Appearances are a little deceptive here though. For those who endorse the negative thesis, the particular shift in our reasons, once sufficiency is achieved, is a shift from some set of distributive reasons to no distributive reasons. As such, the negative thesis specifies a particular shift and so those who endorse the negative thesis offer one type of sufficientarian view, on my conceptualisation. The definitional statement of sufficientarianism that I presented above allows a rich variety of views to count as sufficientarian. The reason to celebrate this is that, the negative thesis has attracted a good deal of forceful criticisms and so if the negative thesis were a definitional claim, the prospects for sufficientarianism would appear to be poor. I conclude that if the prospects for sufficientarianism are to be good and better than has been thought, then it must be because there are some more attractive positions that reject the negative thesis. 212 Liam Shields LEAP 5 (2017) In the book, I defend as plausible and theoretically and practically significant several sufficiency principles that endorse a non-instrumental concern with distributions once the threshold is met. Each of these views rejects the negative thesis and insists that different distributive reasons applying to supra-threshold benefits and burdens. Because of this let’s call these views versions of positive shift-sufficientarianism, a label I do not use in the book but it will be helpful to use here. To defend a positive-shift sufficientarian principle, one must defend a shift as specified by the shift and positive theses and reject the negative thesis. I set out two ways of defending a shift. First, one could defend the existence of justice-relevant satiable reasons. By their very nature, satiable reasons cease to confer their weight on claims once they are satisfied. One way this can happen is if a person has enough of something. So, if we have satiable reasons that cease to confer their weight on claims once that reason is satisfied, the point where that reason is satisfied is likely to cause a shift in our overall reasons to benefit that person, but it does not entail nor imply that there are no other reasons that apply to the distribution of benefits and burdens thereafter. It merely entails that at least one reason that did apply before does not apply after, thus changing the overall set of reasons that apply. Second, one could defend a relative change in the weight of our reasons at the point of sufficiency. This can occur with insatiable reasons. Imagine a uniformly diminishing insatiable reason, one that confers weight on claims for one unit more of some good but confers less weight the more of the good that is possessed, and imagine another non-diminishing insatiable reason, one that confers equal weight on claims for one unit more of some good, regardless of how much someone has. Imagine the non-diminishing reason confers weight of 5 onto any further unit of the good. The diminishing reason could outweigh the non- diminishing reason, be weightier than 5, when low amounts are possessed, but the non-diminishing reason will, as that reason diminishes, be decisive, as greater amounts are possessed. This cross-over point is a shift or change in our reasons to benefit someone and represents a non- instrumental sufficiency principle. By these two methods for defending a shift I aimed to show that principles of sufficiency could be defended as plausible and had an important role to play in theoretical and practical debates. In Chapter Three, I argued for a principle of sufficient autonomy, by appeal to the satiable reason we have to secure the social conditions of freedom, the conditions under which belief can be freely held. In Chapter Four, I explained that this principle should play an important role in theoretical debates in helping us to correct a defect in theories of equality of 213 LEAP 5 (2017) Reply to Critics opportunity, especially those that take meritocracy as an important component, and it should help us to justify compulsory schooling to those who would opt out on welfarist grounds. In Chapter Five, I argued for a principle of adequate upbringing, by use of the value clash method. I argued that parents have a relevant interest in parenting, but one that is often outweighed by a child’s interest in the quality of her upbringing as part of her life as a whole. According to this argument, a good enough parent, that is a parent who is good enough to retain the right to rear her children in the face of superior alterative parents, may fall short from the best alternative custodian by no more than the significance of her own interest in parenting. I argued that this position is theoretically important, since it elucidates the most plausible way of thinking about the good enough parent and it strikes an important departure from minimalist abuse and neglect thresholds and demanding best custodian views. I argued that this position is practically important because it helps us to determine the good enough upbringing threshold and directs us to think carefully about the relative quality of alternatives to parental care prior to severing a relationship. Finally, in Chapter Six, I sought to show how the shift-sufficientarian conceptual apparatus made possible new and plausible positions in the debate about the fundamental demands of global justice. I argued that the shift in reasons once one group has secured enough could help explain both compatriot partiality, the belief that we can be required to do more for compatriots than non-compatriots, and state is arbitrary, the belief that the factors that determine compatriot status are morally arbitrary. If we relax these statements of intuitions slightly, we find that they need not conf lict. They can be reconciled. And one way of reconciling them would be to introduce a shift in our reasons to benefit people at the point of sufficiency. So, once people have secured enough, our reasons of compatriot partiality may be sated, or our cosmopolitan reasons may be sated. So, once we have secured enough, we may transition from holding a cosmopolitan position to a statist one or vice versa. 3. REPLY TO VANDAMME Pierre-Etienne Vandamme characterizes my position in certain ways that I would like to discuss in order to clarify and hopefully strengthen my arguments. First, Vandamme characterizes my position as both agnostic and partial. My characterization of sufficientarianism does leave open the possibility that we may be concerned with inequality or priority to the worse off even once the threshold has been met. The shift thesis is 214 Liam Shields LEAP 5 (2017) compatible with a wide-range of views about how to distribute supra- threshold benefits and burdens. As the view I defend does not commit us to any specific recommendations, or lack thereof, once the threshold is met Vandamme characterizes this position as agnostic. “Shields rejects the principles of equality that fail to take into account the discontinuity introduced by the sufficiency threshold, but he does not provide a justification for not adopting a form of sufficiency- constrained egalitarianism. And this might be explained by agnosticism towards residual inequalities” (Vandamme, in this issue). Vandamme also notes that my commitment to the shift thesis renders my favoured views partial. Unlike most sufficientarians, who endorse a particular distribution of sub-threshold benefits and burdens and an attitude of indifference to supra-threshold benefits and burdens, my favoured view does not provide a determinate answer to the question of what to do with supra-threshold benefits and burdens on its own, though I do think some principle does apply. Therefore, it is true to say that my view is partial. I think that a single sufficientarian principle could, at most, be part of a full account of distributive justice. While it is true to say that the conceptualisation of shift-sufficientarianism, the combination of the positive thesis and the shift-thesis, is agnostic, it is intended only to identify which views are and which views are not sufficientarian. Any conceptualisation of sufficientarianism should be compatible with a wide variety of specific accounts of sufficientarianism. In other words, it should be agnostic, to some extent. It should allow for different currencies, different placements of the threshold(s) and different guidance for how to deal with both sub-threshold and supra-threshold distributions of benefits and burdens. So, the conceptualisation is certainly agnostic about that and partial in that it insists that other principles must be included alongside sufficiency principles in a complete theory of justice. However, there is one important way in which the conceptualisation it is not fully agnostic. What it means to count as a sufficientarian, on the shift- based understanding, is that the distributive reasons that apply to supra- threshold benefits are not the same as the distributive principles that apply to super-threshold benefits. This still leaves open a w ide range of possible combinations, but it rules out one v iew of how to distribute supra- threshold benefits and that is, the same as sub-threshold benefits. Such a v iew would not contain a normatively significant threshold, one that makes a difference to our reasons. More generally, I am sceptical of indifference as the correct attitude to take to supra-threshold benefits and burdens. I believe that the indifference objection will always have force against the negative thesis within the 215 LEAP 5 (2017) Reply to Critics circumstances of justice. The specific sufficiency principles that I defend in the book do endorse a positive shift, that is, they endorse the shift thesis but deny the negative thesis. However, in the book at least, I am silent, if not agnostic, on what principle(s) should apply to supra-threshold benefits and burdens. But the fact that I endorse a positive shift, means that I also accept a partial view. I am sceptical about sufficientarians that endorse the negative thesis, which is why I recommend the positive shift. However, Vandamme is correct to note that I am quite silent on what this should be, I simply doubt being agnostic or offering a partial view amounts to a problem. Being agnostic is essential to a broad conceptualisation and being partial is the only way that sufficientarians can be at least moderately plausible. Second, Vandamme characterizes my position as pragmatic and concerned with rules of regulation, rather than fundamental moral principles when he says that “What Shields seems to be looking for is a clear rule of regulation, and this pragmatic motivation might partly explain his non-selection of luck or outcome equality as the primary or secondary principle of justice” (Vandamme, in this issue). Although Vandamme notes a number of reasons that could be used to defend sufficientarianism that are themselves pragmatic, such as reasons of urgency, reasons of feasibility or reasons of modesty, these are not reasons that I believe should figure in a task like mine. My concern with sufficientarianism is to see whether it is justifiable simpliciter rather than justifiable to currently existing people or justifiable as a means to an end. I can see why there might be good pragmatic arguments from these bases for sufficientarianism, but such pragmatic arguments are too contingent to provide a secure grounding for principles of justice. For example, if, as is likely the case, the most feasible option is the status quo some pragmatic reasons would not support sufficiency, but that would not seem to be a good reason to favour the status quo. But there is another sense in which my view could be pragmatic. Vandamme says I seem to endorse the view that principles of justice should directly guide action, which appears in my claim that if principles “had little significance in terms of policy implications […] then it could not have an extensive role in our thought” (Shields 2016: 10-11) characterizing my project as seeking rules of regulation and not fundamental principles of justice. I can understand why the quoted passage would lead someone to think I was seeking rules of regulation, but I am not. It is important to recall that my objective is to assess the prospects for sufficentarianism and this turns on the extent of their role in our thought. I think that if a principle was true and sound, but nevertheless has no important policy or practical 216 Liam Shields LEAP 5 (2017) implications it wouldn’t much improve the prospects for that principle, even though it would have an indispensable role in our thought. For example, imagine that in order of lexical priority our first principle is sufficiency (basic needs), our second principle is equality, our third principle is priority to the worse off and our fourth principle is efficiency. Now imagine a fifth principle of individual desert is added. According to this arrangement of principles, individual desert would make a difference only in those cases where, the first, second, third and fourth principle were satisfied as far as possible in lexical order. Because of this, we would not expect the principle of individual desert to play much of a role in our thought nor about practical debates, even if it is strictly required in a full description of the demands of justice. Concluding that some principle has an indispensable role in distributive justice would be too trivial a conclusion on its own, I think. So my concern with the practical and theoretical significance of a principle of justice is not a concern with a principle having clear policy implications here and now. Rather, it is a concern with its place within a theory of justice and its capacity to help us to understand practical debates. The latter can be achieved by derivation. Indeed, all rules of regulations are derived from, and in this sense account for, the action- guidingness of the fundamental principles they are grounded in. 4. REPLY TO NIELSEN Lasse Nielsen’s response to the arguments of Chapter Two advances an argument in favour of retaining the upper-limit character of sufficientarianism, against my suggestion that sufficientarians do better by opting for, what I have here called, a positive shift. Nielsen thinks that by defining sufficientarianism as I do I allow that non-sufficientarian reasons could outweigh the reasons we have to achieve sufficiency, but this would betray the strong sufficientarian conviction that securing enough is paramount. He also objects that the shift sufficientarian position remains vulnerable to the indifference objection, which partly motivated its creation. I shall respond to each point in turn before engaging with a further objection. Nielsen suggests that to be worthy of the label sufficientarian one must believe that sufficiency is the only thing that matters or that it is the most important among a plurality of considerations. The label sufficientarian would cease to capture a wide range of positions that give a fundamental role to sufficiency that could not be avoided in fully describing the principles of justice. But it would also limit sufficientarianism to a number 217 LEAP 5 (2017) Reply to Critics of positions to which there are already good objections.2 Those who hold that only sufficiency matters or give lexical priority to sufficiency, which is the f lip side of my allowing other reasons to outweigh reasons of sufficiency, will implausibly always favour helping the badly off by tiny amounts at the expense of helping the well-off by huge amounts. For this reason, we should wish to look for positions that use the attractive and common sense idea of sufficiency in more plausible ways. In setting out the shift-thesis I believe I have articulated the central idea in a way that is compatible with more plausible views. If we took Nielsen’s suggestion and applied it to other views too, then we would see the problems more clearly. For example, if we restricted the label egalitarian to views that held that equality is the only or most important demand then egalitarianism would always be vulnerable to levelling down. If we apply this to prioritarianism and restrict the label prioritarian to views that held that priority to the least advantaged was the only or most important demand then all prioritarian views would be vulnerable to a waste-based objection too, as it would only include absolute prioritarianism. I think this approach would impoverish our understanding of different ideas that can be useful in distributive justice. It would restrict use of these labels to views that were quite implausible, necessitating the creation of more labels. This is just to say there are good reasons to be pluralist and this comes from the problems there are with monist views. Moreover, my characterization includes those monist views in any case. Nielsen is right to say that my own characterization of sufficientarianism doesn’t fully avoid a version of the waste or indifference objection because the objection attaches itself to lexical priority, which is compatible with a rejection of the negative thesis, and not merely the negative thesis itself. But the difference between my position and upper-limit sufficientarianism is that I don’t have to endorse a claim that has this implication, such as lexical priority or the negative thesis. I set out a way that sufficientarians can be distinctive without endorsing lexical priority. Overall, I suppose some of the disagreement between Nielsen and myself is that I do not find lexical priority plausible. I do not think there exists a disadvantage (however tiny) such that ameliorating it is more important than any other benefit. I do not know what further to say about this, though I think the discussion of the illusion of numbers discussion in Nielsen’s paper is relevant to it, so I shall now turn to that. In Nielsen’s response, he develops a point about the illusion of numbers, which I think give expression to an idea that underpins the suspicions that 2 W hether these are good objection is, obviously, a matter of dispute between me and Nielsen (and many others). 218 Liam Shields LEAP 5 (2017) relational egalitarians have for luck egalitarians and others, and expresses a suspicion about outlandish numerical counter-examples to upper-limit sufficientarianism. I think that my particular use of numbers provides an ideal case for his objection, but I think the numbers are forceful even when the differences are lower than those I state. The point of using very large numbers is to simply exaggerate the point to make the implausibility of indifference as clear and as forceful as possible. We could describe a case to illustrate that point instead, without using numbers. Either it can be modelled using numbers – in which case the illusion is not one – or it cannot be modelled using numbers – in which case it is unclear how people can be said to be better or worse off. It appears, however, that the upper limit sufficientarian thinks numbers matter below the threshold but not above it and that seems odd to say the least. There is much more to say about this, but a final brief remark will explain my caution in accepting it. The structure of the move made by Nielsen in the discussion of illusion of numbers is to deny that there are numbers so big that they can represent different levels of advantage, but it seems to me that the underlying sufficientarian position he endorses is insensitive to the fact of the matter. It should not matter to the sufficientarian position whether it is possible to have huge inequalities once enough is secured or not. The position states that even if massive inequalities are possible, they do not matter. So Nielsen’s suggestion that we deny the possibility of these inequalities does not provide a defence of that claim any more than a denial that slavery would maximize aggregate utility is a defence of utilitarianism. I am sure there is much more to say about this on both sides. 5. REPLY TO HUSEBY In his response to Chapter Three, Robert Huseby identifies several ways in which the principle of sufficient autonomy is not clearly specified. In my reply I will aim to provide some clarification in those areas. The first area that Huseby identifies as needing clarification concerns the satiability of the principle of sufficient autonomy. On one understanding autonomy is itself satiable, which is to say that you can get enough autonomy and once you have enough you cannot get any more autonomy. On another understanding the principle of sufficient autonomy is satiable in that the changes or improvements in autonomy it calls for can be fully met, even when it is possible to get “more” autonomy. Huseby states that “If satiable in this way… the principle of sufficient autonomy now looks like a high- threshold sufficiency principle that conforms to the negative thesis” (Huseby, in this issue). 219 LEAP 5 (2017) Reply to Critics Huseby is right. If one cannot get more autonomy than sufficient autonomy, then the position I defended would be vulnerable to the main objection that motivates my argument. I am happy to clarify that my view is that one reason to promote autonomy is the conditions of freedom and that with respect to the promotion of autonomy it is satiable. Once we have enough autonomy to be free we might need more things to be free (though not more autonomy) and we may have reasons to obtain more autonomy (to be happy). So I think that you can get more autonomy, or the related features, once enough autonomy is secured. Anticipating this response, Huseby claims that “This might be perfectly reasonable, but the level would have to be specified.” (Huseby, this issue). But I wasn’t sure why this particular view had any more burden of explanation than any other. Why for example, doesn't an upper-limit principle of sufficient autonomy also have to explain where the threshold is? My position is that I don’t think it has to be specified more than saying that in order to enjoy the social conditions of freedom one must be sufficiently autonomous and to point to gains in terms of autonomy, perhaps valuable options, that wouldn’t make you more free. As it applies to belief formation, one needs a certain amount of autonomy but not full autonomy. The level doesn’t have to be specified for it to be true, vagueness is an acceptable feature of moral principle. This is one reason why the vagueness objection that has been levelled at sufficientarianism is not one I consider in the book. Huseby also urges me to clarify the link between social conditions of freedom and autonomy. Huseby works through several ways of understanding what I have said at various points. When Huseby says “if autonomy is a part of what constitutes the conditions of freedom (or if it is a condition of freedom in itself ), then it could be the case that autonomy can be sated with respect to the conditions of freedom” (Huseby, in this issue) he describes my view. Autonomy can be sated with respect to the conditions of freedom, but autonomy is not sated conceptually, at that point, you can get more autonomy. Nor is it sated normatively. There may be other reasons to promote autonomy. The social conditions of freedom include sufficient autonomy. Sufficient autonomy is not the only aspect of the social conditions of freedom and so being sufficiently autonomous, is not sufficient for the social conditions of freedom, but it is necessary. Huseby goes on to point to a particular problem with this understanding, “If autonomy is a part of what constitutes the conditions of freedom (or if it is a condition of freedom in itself ), then it could be the case that autonomy can be sated with respect to the conditions of freedom. Sufficient autonomy just is autonomy sufficient for the realization of 220 Liam Shields LEAP 5 (2017) (sufficient) conditions of freedom. In my view, however, this interpretation squares badly with Shields’ presentation of the principle, according to which there are supposed to be weighty, non-instrumental, satiable reasons to provide peoples with sufficient autonomy” (2016: 45). As I understand Huseby’s point, it is that the non-instrumental character of the principle of sufficient autonomy is threatened by its being sufficient for the social conditions of freedom. To clarify I don’t think sufficient autonomy is sufficient for the social conditions of freedom, there are other conditions, but the “sufficient” in “sufficient autonomy” is a level determined by what is required, if other conditions are met, for the social conditions of freedom. In other ways, the whole justificatory basis for the principle of sufficient autonomy is that it contributes to the realization of the social conditions of freedom. It therefore looks instrumentally valuable. If it were instrumentally valuable, then it would not support the prospects for sufficentarianism as I have characterized them. Instrumental sufficiency principles can be omitted from a complete description for the demands of justice. However, I think that the link between the principles of sufficient autonomy, as the autonomy component of the conditions of freedom, has a tighter link than an instrumental principle might. This is because nothing else could help us realize the conditions of freedom in its place. Sufficient autonomy is not substitutable. One way of characterizing this link is in terms of the constitutive value of sufficient autonomy. The commitment to the social conditions of freedom, always and every where, includes a commitment to sufficient autonomy because they are so linked. For this reason, a complete description of the principle of justice could not omit reference to sufficient autonomy. 6. REPLY TO MILLS In responding to the arguments of Chapter Three, Chris Mills makes two points about the principle of sufficient autonomy' which states that individuals should secure enough autonomy to secure the social conditions of freedom. First, he states that the principle is too thin and will fail to protect us from all violations of autonomy, in particular he is concerned that the principle I offer relies on a distinction between coercion and external threats, capturing only the latter and not the former. Second, that constitutive views of autonomy and welfare can be defended and won’t have a threshold. With regards to the first point, that the principle I offer relies on a distinction between coercion and external threats, capturing only the 221 LEAP 5 (2017) Reply to Critics latter and not the former, the principle that I put forward is only supposed to offer a partial defence of autonomy. My aim is to show that there exist sufficientarian shifts, and so all I need to do in this chapter is show that autonomy has one such shift. I don’t need to say these are the only or even the most important violations of autonomy. So then Mills could emphasize his remarks about how our views might not have been arrived at freely when there is self-deception. That freedom to set and pursue our ends can be thwarted or frustrated by our own self- deception. In reply, I would say that the requirement to deliberate, and be disposed to deliberate, with others seems sufficient for avoiding some kinds of self-deception at least. This focus generates the requirement that citizens are: “(a) well-informed, (b) able to give reasons for one’s views, and (c) disposed to exchange reasons and participate in a public deliberative process with others.” (Mills, in this issue) These attributes would provide good protection against self-deception through being ill-informed or unref lective. However, it might not avoid the problem entirely. There may be some forms of self-deception that are consistent with sufficient autonomy, and if they too frustrate our freedom, particularly freedom in belief formation, that would be a problem. One avenue sketched by Mills seems attractive. I am tempted to say that some forms of self-deception themselves are not obviously a concern of justice. Not in the purest case of self-deception at least. Where the social background or particular policies or laws encourage self-deception, it is not clear that the deception is really self-deception rather than something else. I follow Mills when he is mapping the possible positions I could take to the point where he characterizes my view are being concerned primarily or exclusively with interpersonal threats. While it is true that sceptics will respond by “denying the downstream relationship and arguing that our autonomy is threatened by more than a mere loss of freedom” (Mills, in this issue) my view is not incompatible with other additional justifications for a concern with autonomy and while I have not yet developed an account of what they are I could possibly adopt them and thereby explain these cases too. Mills’ second point is that a non-instrumental constitutive value of autonomy as a pre-condition for welfare could be defended and could be governed by a prioritarian principle. He states that “If you are a uniform prioritarian about welfare, then constitutive welfarism allows you to: (a) distinguish between qualitatively different disadvantages, and (b) appeal to some reasonably fine-grained metric of well-being in order to distribute autonomy without necessarily appealing to sufficientarian reasons” (Mills, in this issue). 222 Liam Shields LEAP 5 (2017) The first thing I would like to say in reply is that prioritarian welfarism is not incompatible with my view. The idea that we have reasons to promote autonomy that are grounded in welfare and that are uniformly diminishing in moral importance is consistent with thinking that there is an overall shift caused by our reasons grounded in the social conditions of freedom. It is only if prioritarian welfarist reasons were the only reasons to care about autonomy, that this would be a rival to the principle of sufficient autonomy. But that sort of monist view would be implausible and would fail to meet criterion a). If our only reasons to be concerned with autonomy are to do with well-being, then there would be no qualitative difference between violations of autonomy. One possible way around this would be to give the account of autonomy a special place within well-being, so that violations of it were different from violations of well-being simpliciter. Mills suggests that autonomy might be a pre-condition for well-being in his discussion and I discuss that below as Danielle Zwarthoed develops this point further. 7. REPLY TO ZWARTHOED In her response to Chapter Four, Danielle Zwarthoed advances two arguments. First, Zwarthoed argues that at least some instrumental accounts of autonomy, where autonomy is causally necessary for welfare, can justify mandatory autonomy enhancing education, thus denying parents the right to remove their children from aspects of civic education. This point runs contrary to my argument that because instrumental accounts of autonomy cannot justify mandatory autonomy enhancing education and intrinsic accounts of autonomy can, we should endorse an intrinsic account like the principle of sufficient autonomy. Second, Zwarthoed argues that the requirement of talents discovery, which holds that individuals have an entitlement to sufficient opportunity to know and develop their native talents, does not fit well with Rawls’ principle of fair equality of opportunity, as I claim, because that principle points us towards the development of different talents than does the principle of talents discovery. I shall respond to each point in turn. In advancing the claim that instrumental arguments can justify mandatory autonomy enhancing education Zwarthoed considers whether autonomy is a necessary pre-condition of well-being. If it is, then mandatory autonomy-enhancing education would follow from this instrumental argument. In my response I put forward some reasons for doubting that autonomy is a pre-condition for well-being and that autonomy as a pre- condition could justify an intuitively plausible level of mandatory 223 LEAP 5 (2017) Reply to Critics education, particularly at a level that would address the practical disagreements around mandatory education. Regarding autonomy as a pre-condition, it seems odd to say that some people haven’t lived good lives simply because they are not autonomous. Indeed, I think that this view implausibly commits its holder to the view that childhoods cannot go better or worse or cannot make your life go better or worse. Consider two more plausible roles autonomy might play that are constitutive of well-being. Autonomy can be said to amplify our well-being in a way that means our successful pursuit of objectively valuable ends is much greater when that pursuit is autonomously chosen. Autonomy can be said to enable us to reach high levels of well-being that it is not possible to reach non-autonomously. Rather than being a pre- condition, autonomy might more plausibly be an amplifier or the lifter of a cap on well-being. But once we reject the pre-condition account of the value of autonomy we cannot make the causal claim and so the account ceases to be instrumental. Moreover, we lose its ability, on its own, to ground mandatory education, since there is going to be a trade-off between the kind of well-being that a person can get from living a non-autonomous, traditional, way of life. I suppose that persons can live f lourishing lives in such communities. Others may deny this, but, this denial is implausible and uncharitable to those who argue against autonomy enhancing education. The strongest point those from traditional communities have is that these children are currently on a path that leads to f lourishing. My argument does not deny this, it simply insists that f lourishing is not the only thing that matters. Being free also matters, and it matters a great deal. But even if there is some way around this problem, even if autonomy really is a pre-condition for well-being, the kind of education for minimal autonomy that is a pre-condition is not going to support compulsory education beyond a very minimal level. The reason for this is that the higher the threshold is set the more implausible its implications. In the case of autonomy, it gets implausible because it implies that very many people live lives of zero well-being. So the view is only plausible if the threshold is set fairly low, but this may be set too low to ground an intuitively plausible account of mandatory autonomy education. If you look at the court cases Yoder v. Wisconsin and Mozert v. Hawkins, the traditional communities are not asking for their children to be exempt until they are almost teenagers. At which point I think it is plausible to think they have enough autonomy to have met the pre-condition for well-being, but it is not plausible to think that they have enough autonomy to be making free choices. In her second argument, Zwarthoed takes issue with my account of the 224 Liam Shields LEAP 5 (2017) requirement of talents discovery. That account states that each person should have sufficient opportunity to identify and develop their native talents. I claim that this account is a pre-requisite for any plausible version of Fair Equality of Opportunity, which is concerned with equalizing the prospects of those with equal native talent and ambition (Rawls 2001: 42-44). I argue that the requirement of talents discovery is attractive partly because it fits well with Fair Equality of Opportunity. If it did not, I take it, that would be a reason to be suspicious of it, if not reject it. Zwarthoed argues that the fit is not good. This is because the sorts of talents that the requirement of talents discovery focus on are not the same as Fair Equality of Opportunity, so there is a tension between the two. She claims that the requirement of talents discovery will focus on talents requirement for the conception of the good, while Fair Equality of Opportunity will focus on the talents required for economic positions. While I agree that Fair Equality of Opportunity might naturally focus on talents for acquiring economic positions I think this would be included in any conception of the good planning. So, without conception of the good planning, an adequate range would include a focus on marketable talents, but this would not be the exclusive focus. Moreover, the grounding of Fair Equality of Opportunity is in self-realization, which itself is grounded in the two moral powers, so I don’t think that we can say that economic talents would have any significance for Rawls except insofar as they are conception of the good talents (see Taylor 2003; 2004). The fit then, with Rawls and with his explicitly stated grounds of Fair Equality of Opportunity is good. 8. REPLY TO GHEAUS In her response to the arguments of Chapter Five, Anca Gheaus raises some very important issues in relation to my account of when parental rights over particular children can be re-allocated. I will respond to two of the counter-arguments she provides. First, Gheaus argues that what I take to be a unique advantage of my version of the dual-interest view, its ability to explain why we need not re- allocate wherever there is a better custodian available, can be had by both the child centered and dual-interest views. I believe her argument, however, begs the question. Gheaus concedes that one way that my rivals might respond, by appeal to the child’s interests in continuity of care, would be question begging. The reason for this is that this type of case is not the one that separates my view from the rest. If children do have a very strong interest in continuity of care, then it is hardly likely that there is a better alternative custodian, though this isn’t necessarily the case. My concern is 225 LEAP 5 (2017) Reply to Critics with showing that the dual-interest view I defend can explain that even if there is an alternative custodian who would in fact do a better job in terms of the child’s interests, we would not usually be justified in re-allocating rights to her under certain conditions. Gheaus then presses what she takes to be a more decisive argument, which is to appeal to the need for the right of parent to be securely held, something which she finds in the work of Vallentyne, a child centered theorist, but which can also be adopted by dual-interest theorists, since they are concerned with the child centered reasons and other reasons too. “There is, however, a reason why a change in custody away from adequate parents is impermissible even when the child would really be better off with extraordinarily good parents. This reason is advanced by some child-centred theorists (Vallentyne 2003). Children‘s interests are well served if, once acquired, the right to parent is securely held – that is, immune to custody change, as long as the parent is at least adequate” (Gheaus, this issue). This example looks structurally identical to the question begging case. Securely held rights would have to be grounded in an interest that children have for Vallentyne to endorse it. The details of the interest are not so important, the fact that it is a child’s interest suggests that in this case, the interests of children are being best promoted by maintaining a secure attachment, which, if severed, would leave the child to live a worse life, even if the alternative custodians would have done a better job excepting the costs of severing this attachment. So, again, this is not a case where rival dual interest or child centered views do succeed in explaining why the best custodian should not get the child. The objection from secure attachments relies on an assumption that my view identifies as the best parent, the parent who would do the best job if there were no costs of severing the relationship. But that it not my position. My position is that the best parent(s) would do the best job from now, taking into account all the relevant costs, which include the costs of separation. One could argue that this makes the view uninteresting because all children and parents value secure attachments, making redistribution unjustified in all but the most extreme cases. Of course, this wouldn’t be an objection to the soundness of my view, but even if secure attachments are very important to children, this interest will likely vary in its strength when applied to particular attachments. For example, in the early years it may be possible to remove a child from a current parent without creating large costs and without jeopardising the possibility of the establishment of a secure attachment with the new carers. As the child ages, and their initial attachment and relationship develops, the costs could be much larger and 226 Liam Shields LEAP 5 (2017) therefore harder to outweigh. I believe our views should be sensitive to this case and mine is. The second major issue is explained by Gheaus as follows, “Most of us now believe that children are our moral equals except from the fact that their lack of full autonomy makes paternalistic behaviour towards them permissible (indeed, required.) If so, then exercising authority over children must be justified by appeal to their consent or by appeal to their own interests but not, usually, by appeal to the interests of those who exercise the authority. Children cannot give valid consent. Therefore authority over them cannot be denied to those likely to advance their interests as much as possible for the sake of advancing the interest of other prospective authority-holders” (Gheaus, this issue). According to Gheaus, as someone who believes that children have full moral status I am committed to the presumptive principle that authority over them can only be justified by reference to their own interests. The only acceptable exceptions are where there are very strong overriding reasons, such as equality. But, as a sufficientarian, I do not believe in equality, so I cannot avoid being committed to the presumptive claim that authority over children can be justified only by reference to children’s interests, thus making me a child-centered and not dual-interest theorist. Indeed, all those who reject equality and endorse the full moral status of children, should think this, if Gheaus is correct. This is a very interesting and intricate argument, and so first a few clarifications are required. I do believe that equality in distributions does matter and that is part of what motivates me, if not motivates the view that I defend throughout the book. Also, I don’t think that I am committed by this particular argument to saying that children have full moral status if that means they have the moral status of fully competent adults. I do find it plausible to think that children may have a moral status that is different, I do not know what “full” means here, but if all it means is that children can have basic rights, then I do agree with it, but I don’t see why it follows from that that we cannot make decisions that affect them or yield authority over them in the interests of others. Having said all of this, I don’t think these points suffice to produce an adequate response to Gheaus on my part. Instead, my response will question the claim that an agent with moral status cannot have someone else wield authority over them in the interests of others. Gheaus argues that dual-interest theories violate the sound (prima facie) principle that one cannot claim legitimate authority over someone possessing full moral status by appeal to one’s own interests. On my view, 227 LEAP 5 (2017) Reply to Critics it is possible for the interests of parents not only to matter, and to be decisive in tie-breaks, but to out-weight the interests of children. So, one might think my view violates this principle in the worst way and not only in some way. Gheaus describes two ways of resisting the presumption. One way of showing the presumption can be overridden is by appeal to equal opportunity to f lourish and the way that parenting contributes to f lourishing can override that principle. But in order to advance this argument one must accept that parenting is non-substitutable and that justice requires equality of opportunity to f lourish. But I reject the non- substitutability of parenting. So I cannot avoid this problem, it seems. There’s plenty of ambiguity in the so-called “sound principle”. In particular, it could be interpreted as prohibiting authority over someone simply because it is in the interests of the prospective holder of authority, or it could be that no one can have authority over someone in anyone’s interests but those of the person over whom authority is held. These two interpretations have quite different extension. Moreover, if the principle Gheaus describes, and borrows from Vallentyne, is true then it is unclear that, for example, democratic institutions, where authority is exercised, at least in some significant life-affecting decisions, in the interests of all. So the idea that no one can have authority over me in anyone’s interests but my own must be false on either interpretation. I agree with lots of the examples Gheaus gives about adults having control over other adults, e.g. romantic partners. But the problem with drawing conclusions from that example this is that someone must have control over children, no one else need have control over the bodies of competent adults. That is a morally relevant distinction that can explain different treatment. Moreover, insofar as Gheaus accepts this she accept that children’s status is different from adults. This is just to say I think the “sound principle” is questionable. I think the cases I describe are still the best to show why the parents’ interests can justify shortfalls from the best custodian in some cases because they illustrate the following. While some of the child’s interests are very weighty, such as their interest in avoiding neglect and abuse, their interests are not all weighty. Those less weighty interests can be outweighed by the interests of parents. Moreover, I think cases where parents make decisions that have costs for their children, like taking a different job, re- locating, etc. can, in some cases be justified. Gheaus’ view cannot explain this except if she appeals to an artificially robust distinction between the interests that are relevant to decisions parents make once they have the rights and the interests relevant to the decision to give a particular parent the right. There should be continuity between these, not least because 228 Liam Shields LEAP 5 (2017) what a parent would do with the rights is a determinate of how well they will do it and so a sharp discontinuity here cannot be justified. 9. REPLY TO HARB AND AXELSEN Siba Harb and David Axelsen’s response to Chapter Six has two parts. Part one of their reply argues that there is nothing new in one of the distinctions I make while part two provides an interesting development of one of the lines of thought in the chapter. As the second part is mainly complimentary I won’t engage with that much, except to say “thanks”. I will use this reply to try to explain why the first distinction is of importance. I argue that the achievement of sufficiency could trigger a shift in the content of our obligations to compatriots and to foreigners, and not merely their stringency. By this I mean that once the sufficiency threshold is met for at least one of these groups, then we may move from being statists to being cosmopolitans. That means we move from thinking that the content of our obligations to compatriots and foreigners is different, to thinking that they are the same. The authors discuss the example I give in the book about moving from being a statist to a cosmopolitan by reference to the satiable reasons that can justify a threshold. “Now, according to Shields, reasons that can justify the existence of a threshold are ones that are satiable. The reasons that Shields explores in previous chapters are basic needs and autonomy. Both are satiable in the sense that they do not provide a normative basis for benefiting above a certain level (the threshold) (Shields 2016: 34-37). Surely, however, reasons of basic needs and autonomy apply universally; everyone shares the trait that gives rise to such reasons. And, indeed, when theorists in the global justice debate, be it statists or cosmopolitans, hold that we owe basic needs to foreigners, they hold that we do so because they are human beings, not because they are foreigners. It is difficult to see, then, how one can be statist about content below the threshold” (Harb and Axelsen, this issue). Now, I do not claim that only satiable reasons can justify a non- instrumental threshold, but I claim that some do, and these may very well be the most plausible such reasons that justify thresholds. Reasons that focus on basic needs and autonomy, once met, no longer apply. These reasons however appear to apply universally and so would not be plausible candidates for the sorts of reasons that can explain why the content of our obligations can change. 229 LEAP 5 (2017) Reply to Critics The concern that the category of shift sufficientarian content approaches to global justice does not contain a plausible member, is a serious one. To respond I will develop the statement in the book explaining this view. I state that “Possible examples include a view whereby we owe sufficiency to non-compatriots and equality to compatriots but once non- compatriots have enough, we owe equality to all” (Shields 2016: 187). This is all too brief, so I am grateful for the opportunity to expand here. The line of thought set out in the quoted sentence described a view whereby we pursue equality among compatriots and sufficiency globally. These are two of our duties of justice. How they are to be weighed is a further question and concerns stringency. This position remains too vague to test its plausibility comprehensively, though a good deal of the plausibility test will be met by the implications of such a view and whether they can explain intuitions like compatriot partiality and state is arbitrary. So consider a possible example. Imagine our set of reasons include a commitment to global sufficiency, perhaps specified by basic needs or autonomy, and a commitment to local or state- wide equality of opportunity, as well as a global commitment to helping the worse off in the form of a priority principle. At this point our reasons are mixed. Some are cosmopolitan and others are statist. Whether our outlook is statist or cosmopolitan, I think, depends on how far we have gone in meeting our reasons of justice. If some of our reasons (cosmopolitan or statist) can be sated, then once enough is secured, the content of our remaining obligations is different for compatriots and non-compatriots. This would mean that the satisfaction of global sufficiency means our outlook is statist. This might appear to be a superficial shift since the totality of our reasons remains the same: a mix of cosmopolitan and statist reasons. It’s just that some no longer apply. Alternatively, the achievement of global sufficiency might bring into existence new reasons. For example, once everyone is sufficiently autonomous perhaps political equality matters between compatriots but not globally. Perhaps also equality of opportunity matters between compatriots but not globally. The last example would be a more thoroughgoing account of a shift in the content of our obligations 10. CONCLUSION I am very grateful to the contributors for pressing me on these and other areas, and I hope I will be able to more fully appreciate those contributions as I continue to think about them. 230 Liam Shields LEAP 5 (2017) BIBLIOGRAPHY Rawls, J., 2001: Justice as fairness: A restatement. Cambridge, MA: Harvard University Press. Shields, L., 2016: Just Enough: sufficiency as a demand of justice. Edinburgh: Edinburgh University Press. Taylor, R. S. 2003: “Rawls's defense of the priority of liberty: a Kantian reconstruction”, Philosophy & Public Affairs 31: 246-271. Taylor, R. S. 2004: “Self-realization and the priority of fair equality of opportunity.” Journal of Moral Philosophy 1: 333-347. Vallentyne, P. 2002: "Rights and Duties of Childbearing." Wm. & Mary Bill Rts. J. 11: 991 – 1009.