A Paternalistic Account of Stakes1 Søren Flinch Midtgaard, Aarhus University (midtgaard@ps.au.dk) Abstract Authors broadly supportive of the luck egalitarian project have recently emphasized the need for an account of stakes or a structure of pay-offs. Moreover, they have sketched various substantive accounts. Their effort can, in part, be seen as a response to misgivings about luck egalitarianism’s apparent acceptance of cases of absolute deprivation. Joining this campaign, the paper presents a novel paternalistic account of stakes. By doing so it aims to redraw the battle lines between friends and enemies of luck egalitarianism, both groups aiming to avoid paternalistic considerations. However, the latter arguably ground an account of stakes that is superior to competitors especially with respect to its ability to solve a trilemma facing such accounts. Keywords luck egalitarianism • stakes • paternalism I. Introduction The criticism levelled against luck egalitarianism by Elizabeth Anderson and others is by now familiar and apparently hard-hitting.2 The charge that luck egalitarianism in some circumstances countenances deep inequalities, and even declines assistance to people in dire straits, seems especially worrisome.3 It appears to call in question the view’s ability to outline a way in which the state may treat individuals with equal respect and concern, as Ronald Dworkin memorably phrased the arguably fundamental 1 I am grateful to David Axelsen for very useful written comments. I am also indebted to the audience at the Society for Applied Philosophy Annual Conference 2015 at the University of Edinburgh, 3-5 July. 2 See, esp., Elizabeth Anderson ‘What Is the Point of Equality?’, Ethics, 109 (1999), 287-337. 3 Cf. Anderson ‘Point’, p. 300. 1 abstract egalitarian requirement to a satisfactory theory of distributive justice.4 The objection has called forth an array of luck egalitarian responses.5 Especially one of these concerns me here. It stems from the works of a number of authors―including Paula Casal, Serena Olsaretti, Zofia Stemplowska, Peter Vallentyne, and Andrew Williams6 belabouring, in different ways, a common theme. The theme is this. A satisfactory luck egalitarian theory should be seen against the background of a plausible account of stakes that delineates the kind of consequences or disadvantages that may be attached to certain features of persons, such as their genuine choices, for which they can appropriately be held 4 In, for example, Ronald Dworkin, Taking Rights Seriously (London: Duckworth, 1977), pp. 180, 272-3; Sovereign Virtue: The Theory and Practice of Equality (Cambridge, Massachusetts, and London, England: Harvard University Press, 2000), pp. 1-2, 11. 5 For an overview, see, e.g., Carl Knight ‘abandoning the abandonment objection’, Res Publica (2015), DOI 10.1007/s11158-015-9273-2, 1-17; Kasper Lippert-Rasmussen, Luck Egalitarianism (London, New Delhi, New York, Sydney: Bloomsbury Publishing, forthcoming), chap. 7. 6 Paula Casal ‘Why Sufficiency Is Not Enough’, Ethics, 117 (2007), 321-323; Serena Olsaretti ’Responsibility and the Consequences of Choice’, Proceedings of the Aristotelian Society, CIX, Part 2 (2009), 165-188; ‘Rescuing Justice and Equality from Libertarianism’, Economics and Philosophy, 29 (2013), 43-63; ‘Scanlon on Responsibility and the Value of Choice’, Journal of Moral Philosophy, 10 (2013), 465-483; Zofia Stemplowska ‘Making Justice Sensitive to Responsibility’, Political Studies, 57 (2009), 237-259; ‘Responsibility and Respect: Reconciling Two Egalitarian Visions’, Responsibility and Distributive Justice, ed. C. Knight and Z. Stemplowska (Oxford and New York: Oxford University Press, 2011), pp. 115-135; ‘Rescuing Luck Egalitarianism’, Journal of Social Philosophy, 44 (2013), 402-419; ‘Scanlon and Voorhoeve on Substantive Responsibility’, Journal of Moral Philosophy, 10 (2013), 488-507; Peter Vallentyne ‘Brute Luck, Option Luck, and Equality of Initial Opportunities’, Ethics, 112 (2002), 529-557; Andrew Williams, ‘Liberty, Equality, and Property’, The Oxford Handbook of Political Theory, ed. J. S. Dryzek, B. Honig, and A. Phillips (Oxford and New York: Oxford University Press, year), pp. 488-506. 2 responsible;7 and pace critics of luck egalitarianism, and sometimes, perhaps, pace certain paradigmatic luck egalitarian pieces,8 those consequences need not include the kind of disadvantages that get the indicated misgiving about luck egalitarianism’ egalitarian credentials going in the first place. In brief, the mentioned authors point out that an account of stakes or an answer to the question of stakes is part and parcel of a satisfactory luck egalitarian theory. Providing such an account may be an important step towards answering the important “deep disadvantages countenancing” objection to luck egalitarianism. Although I share the view of the mentioned authors that delineating the kind of consequences that individuals can appropriately be asked to bear may offer a pertinent reply to Anderson and others I believe that a promising way of answering the question of stakes has not been sufficiently explored. The suggestion is to draw on paternalistic considerations. The latter, I take it, are, roughly, reasons for restricting people’s liberty against their will in their interest or to prevent grievous harm from befalling them.9 Paternalistic considerations might justify both preventing harm and promoting good. 10 I focus on the former because of its special relevance to the luck egalitarian view on responsibility for disadvantages. It is this aspect of the view that triggers Anderson’s objection and makes exploring the question of stakes relevant.11 By appealing to paternalistic considerations my account challenges one of 7 See, Serena Olsaretti ’Responsibility’. 8 See Stemplowska ’Rescuing Luck Egalitarianism’ on the textual evidence of the commitment of luck egalitarians, including G.A. Cohen, Richard Arneson and Dworkin, to the luck egalitarian view attacked by Anderson and others. 9 Cf. Arneson ‘Mill versus Paternalism’, Ethics, 90 (1980), 470-489, at p. 471. 10 Cf. Feinberg’s distinction between so-called “negative paternalism” and “benefit-promoting paternalism.” See, his Harm to Self (New York and Oxford: Oxford University Press), p. 8. 11 Cf. Olsaretti ‘Reponsibility, p. 169. 3 the lemmas of the trilemma widely recognized to be facing accounts of stakes, 12 namely the problem of paternalism. Second, my account effectively tackles the other two lemmas of the trilemma, namely the problem of exploitation (roughly, A imposing avoidable costs on B13) and insufficiency (roughly, people experiencing disadvantages below a certain threshold). In a nutshell, the paternalistic account of stakes handles the trilemma in the following way. It challenges the paternalistic lemma on the ground that disallowing hard paternalism, that is, the prevention of voluntary self-regarding harm, in all cases is implausible. Especially, cases in which great goods are at stake, or where great potential harms are involved―harms accounted for by reference to the intrinsic well-being-related value of autonomy―and where the harms in question are preventable by limited infringements of the person’s autonomy, call for hard paternalistic measures. Moreover, situations of insufficiency of the kind at issue in debates about the harshness of luck egalitarianism fall within the set of cases mentioned. The paternalistic rationale for securing sufficiency is uniquely sensitive to the concerns brought forth by Anderson’s objection, these concerns arguably revolving around the badness of the grievous harm that may befall inconsiderate and imprudent agents. Finally, the paternalistic account, by virtue of favouring an internalizing strategy, that is, one that makes imprudent agents themselves bear the costs of their conduct, bypasses the exploitation lemma. Alternative accounts of stakes handles the trilemma in less convincing ways. Especially, they fail to cater to sufficiency and/or to non-exploitation. 12 See, Stemplowska ‘Making’, p. 248; Williams ‘Liberty’; Paul Bou-Habib ‘Compulsory Insurance without Paternalism’, Utilitas, 18 (2006), 243-263. 13 Cf. Bou-Habib ’Compulsory Insurance ’, p. 253 4 The paper proceeds as follows. First, I address the question of stakes and elaborate on the trilemma facing substantive answers to it. Secondly, I canvass central proposed answers to the question of stakes pointing to the mentioned respect in which they appear to be wanting. Thirdly, I present and defend my paternalistic account of stakes. II. The question of stakes and egalitarian constraints According to Olsaretti, “The question of the ground of responsibility asks ‘Which are the features of people that we can … hold them responsible for (so that they are liable to bear some costs on the basis of those features)?’ The question of stakes, by contrast, asks ‘Just what costs should attach to whatever features constitute the justifiable grounds of responsibility?’”.14 The distinction echoes Thomas Scanlon’s distinction between ‘responsibility as attributability’ and ‘substantive responsibility’.15 The question of stakes concerns the latter kind of responsibility. Whilst luck egalitarians have been preoccupied with debating various grounds of responsibility―with relevant candidates including courses of actions or patterns of preferences voluntarily chosen by the agent or such courses or patterns with which she identifies in an appropriate way―they have, according to Olsaretti, neglected addressing the question of stakes. 14 Olsaretti ‘Responsibility’, 170. 15 Olsaretti ‘Responsibility’, pp. 167-168; T. M. Scanlon, What We Owe to Each Other (Cambridge, Massachusetts, and London, England: The Belknap Press of Harvard University Press, 1998), p. 248. Actually, Olsaretti takes ‘substantive responsibility’ to form a subset of ‘responsibility as liability’ where the latter includes liability to moral praise and blame (or moral responsibility) THIS IS OPAQUE. See her exchange with Ben Colburn: Ben Colburn ‘Debate: The Concept of Voluntariness’, Journal of Political Philosophy, 16 (2008), 101-111; Serena Olsaretti ‘Debate: The Concept of Voluntariness―A Reply, Journal of Political Philosophy, 16 (2008), 112-121. 5 In terms of the paradigmatic case of the imprudent motorcyclists, Bert, riding his motorcycle without wearing a helmet, to enjoy the wind in his hair, and without having taken out health insurance,16 luck egalitarians unanimously agree that there are some relevant features of this person that we may hold him responsible for (for example, that he has ‘freely adopted’ his ‘negligent and reckless character’17), that is, they believe that we may point to a relevant ground of responsibility. Moreover, they have a tacit understanding of the kind of consequences that Bert and his likes are liable to bear on the basis of these features. Those consequences may include him being left unassisted by the roadside. However, as Olsaretti points out, the tacit assumption of proponents and opponents of luck egalitarianism alike, regarding consequences, belies the complexity and the open-endedness of the question involved. If we look at the case more closely, the consequences we may be inclined to attach to Bert’s conduct are not self-evident at all. It is not, for example, obvious that they include being left unassisted by the side of the road. And if not, in what circumstances are bypassers required to assist him, and to what extent do the costs of his treatment fall on him (and how should those costs be calculated)? As Olsaretti convincingly shows, a number of such questions arise, and are largely unaddressed by luck egalitarians and their critics. Her basic, and very important point, is, I take it, that holding people responsible for some of their features and believing that they are liable to bear some 16 See, Marc Fleurbaey ‘Equal Opportunity or Equal Social Outcome?’, Economics and Philosophy, 11 (1995), 25-55, at p. 40. I take imprudence to designate, roughly, assuming a risk the probability and seriousness of which is disproportionate to the importance of the goal for which it is assumed (the risk in question being given independently of eventual compensatory arrangements or collective measures of assistance). 17 Fleurbaey ‘Equal Opportunity’, p. 40. 6 costs on the basis of those features, presuppose, or is wanting in lieu of, an account of the precise nature of those consequences. The luck egalitarian view includes an important constraint on stakes. This derives from its commitment to a principle of equal opportunities: inequalities condoned by luck egalitarians are those, and only those, that arise from choices against the background of options of equivalent value.18 So had it not been the case that, for example, Bert has had a ‘normal and balanced upbringing’19 and/or enjoyed the option of taking out health insurance on a par with others, the resulting inequalities would have been tainted from the point of view of egalitarian justice. Similarly, consider a system of opportunities in which the stakes of two people’s comparable choices vary. For example, the stakes of one individual’s prudent or imprudent conduct are membership in one of the two top castes, whereas the stakes of another person’s comparable form of conduct is membership of the two lower castes.20 Luck egalitarians condemn this system because the stakes involved are unequal. In this sense, the luck egalitarian view embodies an egalitarian constraint on stakes. However, the constraint implied by the principle of equal opportunity does not deliver a satisfactory or complete account of stakes. An elaboration on the caste example mentioned above goes to show this. Imagine that the caste system is corrected for its inequality of opportunity and that people now face an array of options that is equal in terms of the prospects it offers for ending up in a given 18 Olsaretti ‘Responsibility’, pp. 177-8; ‘Rescuing Justice and Equality from Libertarianism’, Economics and Philosophy, 29 (2013), 43-63, at p. 58-9; Stemplowska ‘Making’, esp. p. 237. 19 Fleurbaey ‘Equal Opportunity’, p. 40. 20 See Serena Olsaretti ‘Rescuing Justice and Equality from Libertarianism’, Economics and Philosophy, 29 (2013), 43-63, at p. 58-9. The case is Brian Barry’s. 7 caste.21 Still, the stakes involved seem unpalatable. It seems unacceptable to let anyone end up in a low caste despite the fact that he has faced an opportunity equivalent to others of ending up in one of the higher castes. The luck egalitarian commitment to a doctrine of equal opportunities, as this brings out, merely caters to comparative concerns of justice, but, apparently, we also have non-comparative or absolute concerns of justice according to which some disadvantages should not be borne by anyone, irrespective of such being the outcome of a system of equal opportunities.22 Catering to these non-comparative concerns is the first challenge facing a satisfactory account of stakes. That is, it should rule out attaching unpalatable costs to people’s choices, however stupid or inconsiderate. We want to avoid the problem of insufficiency. What constitute unpalatable consequences or situations of insufficiency? In the debate between friends and enemies of luck egalitarianism, the boundaries of insufficiency are not stated clearly, and the extent to which a satisfactory account of stakes adheres to this threshold at all times is disputed. However, judging from the examples brought up in the debate I believe it fair to say that the paradigmatic case of insufficiency is of the kind where people experience serious injuries or where they get themselves into emergencies (which may or may not involve them being injured) where they stand in the need of others’ assistance or help.23 Thus we may say that situations of insufficiency are characterized by the kind of 21 Olsaretti ‘Responsibility’, at pp. 178-9. 22 Cf. Williams ‘Liberty’, pp. 499-500. 23 Anderson ‘Point’, p. 295 (focusing on an ‘uninsured driver’ left to ‘die by the side of the road’); Bou-Habib ‘Compulsory Insurance ’, pp. 252, 257 (focusing on the potential injuries of the reckless biker Helmut or of one about to embark on ‘a near-suicidal gliding mission across the [Grand] [C]anyon’); Olsaretti, p. 172 (discussing Bert, an imprudent uninsured biker, not wearing a helmet, experiencing a serious crash); Casal ‘Sufficiency ’, pp. 322-323 (debating uninsured driving, climbing, and smoking which may involve the death of persons engaging in such conduct); Feinberg, Harm to Self, p. 141 8 predicaments that ordinarily triggers the duty of rescue. The paper conceives of the notion of insufficiency along these lines. As pointed out by Williams, sufficiency may be secured either by measures of internalization or externalization. The former strategy lets the expenses involved in securing sufficiency be borne by the imprudent agent him or herself. It does so by restricting her liberty in various ways, taxing her conduct, compelling her to take out insurance, or outright prohibiting the conduct that threatens sufficiency. The latter, externalizing, strategy countenances cost displacement, funding sufficiency or assistance through a scheme to which all have contributed equally. It requires others, that is, to finance the costly projects of the imprudent―projects that are reasonably avoidable to him or her. In brief, it lands us with the problem of exploitation. Luck egalitarians, it seems, should be especially concerned with avoiding this problem. It is arguably a fundamental requirement of equality that the costly projects of some should not be allowed to trespass on the equal share of others.24 This appears to speak strongly in favour of the cost internalizing strategy. However, choosing this strategy may land luck egalitarians with a problem that is equally pernicious as the problem of exploitation, namely the problem of paternalism.25 It arises because the policy in question appears to impose restrictions on individuals’ voluntary self-regarding choices which liberals traditionally believe (focussing on the ‘injury or death’ of a biker); Stemplowska ‘Making’, p. 249, 252 (focussing on bungee jumping accidents and a climber getting stranded and might be left to die); Williams ‘ Liberty’, pp. 499, 501 (focussing on drivers blinded in an accident or biker Bert bringing ‘disaster upon himself’). 24 See, e.g., Andreas Albertsen and Soren Flinch Midtgaard ‘Unjust Equalities’, Ethical Theory and Moral Practice, 17 (2014), 335-346; Dworkin, Sovereign Virtue; Cohen, On the Currency of Egalitarian Justice, and Other Essays in Political philosophy, ed. by Michael Otsuka (Princeton and Oxford: Princeton University Press, 2011), esp. chap. one and five. 25 Bou-Habib ‘Compulsory Insurance’. 9 amounts to an unjustifiable affront to autonomy.26 Some consider the problem surmountable, that is, they believe that compelling non-paternalistic justifications for sufficiency-preserving policies are available.27 This, however, is questionable.28 Moreover, it might not be necessary to avoid paternalism, or a certain version of this. Indeed, as I will argue below, paternalistic reasons are especially compelling in the justification of internalizing measures catering to kind of sufficiency at issue in the debate about stakes. III. Accounts of stakes and their limitations (A) Sufficientarian egalitarianism Consider first ‘sufficientarian egalitarianism’ held out by Williams as a ‘natural way to avoid the excesses of the post-libertarian [i.e., luck egalitarian] view’.29 The view is essentially a psychic harm argument of the kind Feinberg develops in his treatment of the case of protective helmets. 30 In Casal’s 26 For classic statements of liberal anti-paternalism, see John Stuart Mill, On Liberty (1859/1974, Penguin Books); Feinberg, Harm to Self; Arneson ‘Mill versus Paternalism’, Ethics, 90 (1980), 470-489. For an alternative account of the badness of paternalism (although also revolving around the value of autonomy) see Shiffrin ‘Paternalism, Unconscionability Doctrine, and Accommodation’, Philosophy & Public Affairs, 29 (2000), 205-250. 27 See, esp., Bou-Habib ‘Compulsory Insurance ’. 28 Søren Flinch Midtgaard ‘Non-Renounceable Rights, Paternalism and Autonomy’, Utilitas, Available on CJO 2015 doi:10.1017/S0953820815000163. 29 Williams, ‘Liberty’, p. 501; Casal ‘Sufficiency’, pp. 321-323. 30 Joel Feinberg, Harm to Self (New York and Oxford: Oxford University Press), pp. 134-142. Williams’ view is not, as Viki Pedersen has pointed out to me, strictly speaking or necessarily a psychic harm argument. However, he refers to Casal’ view (which is essentially a psychic harm argument) as at least a version of the sufficientarian egalitarian view, and the 10 rendering the fact that the costs experienced by bystanders witnessing insufficiency are significant is our central reason for fleshing out the stakes of people’s choices in such a way that certain situations of absolute deprivations are avoided. The importance of avoiding such costs may, for example, justify compelling people to take out insurance. 31 The idea that the psychic costs that witnessing imprudent people’s misery or predicament in situations of insufficiency involve may warrant constraints on the conduct of the imprudent aimed at preventing her from ending up in insufficient situations is proposed by Feinberg as an attractive alternative to competing non-paternalistic liberal reconciliatory arguments. The mentioned competing proposals, according to Feinberg, labour under the handicap of ‘coyness’ or of being evasive of what seems to be our central concern in these cases revolving around the risked grievous injury of the biker or his death. The psychic harm argument, he thinks, does better in this regard by sharing ‘in the assumption of the paternalist that injury or death to the biker is the central point, not some relatively remote and indirect public harms …’32 At the same time it has the significant attraction of avoiding the arguably repugnant paternalistic rationale. The argument is a “harm to others” argument.33 However, the view fails, in fact, to avoid the problem of coyness. Contrary to what Feinberg claims, the assumptions of the paternalist theorist and the anti-paternalistic “psychic harm” theorist are as far apart as they can be: whilst the former is concerned with the ‘injury or death’ of the biker and is latter view in itself constitutes no alternative to the views considered below (rather it accounts for the form of such views). 31 Casal ‘Sufficiency’, pp. 321-323. 32 Feinberg, Harm to Self, p. 141. 33 For another argument of this kind see Paul Bou-Habib, ‘Compulsory Insurance ‘. For critical discussion see Midtgaard ‘Non-Renounceable Rights’. 11 so for the straightforward reason that this would be a very bad thing indeed for this person (and perhaps to others too, but this is secondary to the paternalist), the latter is not. He or she is concerned with this only derivatively, that is, because of the psychic harm the biker’s accident may bring to others, especially other motorists. This anti-paternalistic argument seems to me unhelpful, leaving luck egalitarians with a relative impotent or lame reply to Anderson’s objection. The key concern is indeed, as Feinberg rightly points out, the injury or death of the biker, but Feinberg’s and others’ psychic harm argument fails to foreground this concern; and the non-paternalistic reason it foregrounds is of doubtful validity. One important problem is that its reliance on psychic harm may commit liberals to other policies which they would otherwise be loath to accept, including policies restricting people’s liberty to engage in homosexual relations on the ground that some experience grievous psychic harm relating to witnessing such relations or from the mere knowledge that such relations exist.34 The non-paternalistic account of stakes under examination is especially marred by its failure to supply a compelling justification of preserving sufficiency. It avoids the problem of exploitation, though, in that its policy proposals are of the mentioned internalizing kind. But perhaps exploitation or cost displacement should not be avoided at all times. This, at least, is part of the next proposal to be canvassed, namely Stemplowska’s proposal. (B) Stemplowska’s proposal The gist of Stemplowska’s answer to the question of stakes is that while recognizing the importance of avoiding cost displacement or exploitation, we should also pay attention to a countervailing consideration to the effect that sometimes important interests are being served by people being able to 34 Cf. Russ Shafer-Landau ‘Liberalism and Paternalism’ Legal Theory, 11 (2005), 169-191, at p. 179. 12 act imprudently without disadvantages being attached to doing so, although the prudent will have to pay to make this possible. Stemplowska compares a scheme regulating payment for car damage stating that ‘every car will be repaired for no charge (the resources coming from everyone’s equal contribution to the scheme)’ with a scheme stating ‘that only those cars will be repaired for no charge (i.e. from everyone’s equal contribution to the scheme) that have been well maintained by their owners’.35 To choose between the two schemes Stemplowska proposes the following general approach: [S]ecuring holdings or situations (through the availability of assistance) should take priority for as long as the interests that are served by not needing to avoid a given conduct (lest it leads to a disadvantage) outweigh the interests that would be served if such security did not have to be provided. Thus for example, if the interests that are served by being able to ignore car maintenance without losing access to free (at the point of use) car repair are more important than the interest people have in retaining the resources that would be needed to offer free car repair, even if the car damage stems from a lack of car maintenance, then free accident repair should be offered as a matter of policy.36 A key attraction of Stemplowska’s approach is that it achieves sufficiency or security without paternalism―it does not even constrain people’s liberty in its attempt to ensure sufficiency. It achieves sufficiency, to be sure, by in some cases adopting a strategy of externalization, hence partly landing itself with the problem of exploitation. However, the strategy of externalization it adopts may appear to be of a limited and reasonable nature (reasonable partly in the light of its facilitation of the avoidance of the problems of insufficiency and paternalism). Hence Stemplowska’s approach may be seen to satisfy the desideratum of non-exploitation to a reasonably extent.37 35 Stemplowska ‘Making’, p. 248. 36 Stemplowska ‘Making’, p. 249. 37 My understanding of the lemmas involved and how different views may satisfy them, or satisfy them to a reasonable extent, has been greatly assisted by Viki Pedersen’s Master thesis, Responsibility, Sufficiency, and Freedom, Department of Political Science, Aarhus University, March 2015. 13 Despite these apparent virtues, the proposal faces, I believe, two important problems. First, it does not, in fact, establish that there are plausible cases in which externalization (that is, asking the prudent to bear some of the costs of the conduct of the imprudent, the conduct in question being reasonably avoidable to the imprudent) is appropriate. The cases it offers as plausible cases for externalization are in fact cases in which the compensation for the differential effects of bad brute luck is in place (that is, compensating people for the adverse consequences of acts which are not reasonably avoidable to them). Second, it, implausibly, implies that sometimes we should both refrain from externalizing and from paternalistic interventions―that is, that there are cases in which we should countenance situations of insufficiency. Consider first the kind of cases in which it might seem appropriate to externalize. I begin by considering a variation on Stemplowska’s competing schemes of regulating payment for car damage. I should stress that Stemplowska does not claim that the scheme that secures holdings or situations―that is, the scheme that accommodates imprudent person’s behaviour by externalizing measures―is preferable to the scheme that protects the interests of the prudent in not providing such security or accommodation. The case is offered as an illustration of the relevant kind of considerations that go into our assessment of the interests of the imprudent vis-à-vis the interest of the prudent. Still, considering the case, or a certain version of it, reveals, I believe, the limited appeal of catering to the interests of the imprudent in the way Stemplowska suggests that we sometimes should. The case is this. We go on an Off Road Trip together on our motorcycles. We bring a truck with essential spare parts for our motorcycles and a mechanic. In assessing schemes of the kind Stemplowska mentions, it seems obvious to me that the fair choice is the cost internalizing one, that is, 14 the one making access to spare parts and maintenance conditional on having taken properly care of one’s motorcycle. (I presume here that denying assistance to negligent bikers does not land them in situations of insufficiency, that is, does not, for example, imply that they will be stranded in the desert left to die of thirst.) I can certainly see my interest as a participant in the trip in being able to, for example, go on a joy ride after having reached our destination for the day “riding my bike hard and putting it away wet,” while losing no sleep fearing that my conduct will jeopardize my access to maintenance free of charge drawing on the resources of the truck and the mechanic to which we have all contributed equally.38 But, I do not think that this interest of mine is to count at all. The interest in question seems to me to be a clearly anti-social or self-indulgent one, and it is unclear to me when, if ever, we should cater to such interests. After all, accommodating the interests of the irresponsible joy riders would imply that more responsible bikers, having taken proper care of their bikes, and restrained themselves by not giving in to the temptation to go on joy rides and/or to opt for some extra hours of sleep in preference to lubricating the chain and cleaning the air filter on their bikes, are asked to support the self-indulgent behaviour of the joy riders. In the worst case, prudent riders may even lose out on their aim to reach the planned destination because the stock of spare parts runs dry as a consequence of the imprudent conduct of the joy riders.39 The resources for accommodating the negligent conduct of the imprudent do not, as Stemplowska points out in another context, ‘come out of thin air’.40 38 Cf. Brian Barry on the point that we may have interests in not refraining from anti-social behaviour or in acting unfairly. Barry ‘John Rawls and the Search for Stability’, Ethics, 105 (1995), 874-915. 39 Cf. Rawls objection to pandering to people with expensive preference in his ‘Social unity and primary goods’, Utilitarianism and Beyond, ed. A. Sen and B. Williams (Cambridge: Cambridge University Press, 1982), pp. 159-185, at pp. 168-169. For further discussion of the injustice of compensating people with expensive tastes, see Rasmus Sommer 15 The most plausible cases offered by Stemplowska as cases where externalization is appropriate are instead cases where we should compensate people for differential effects of bad brute luck. I have in mind cases where it is, strictly speaking, within the power of the agent to avoid exposing herself to the risk of incurring a disadvantage, but where it is not reasonable to expect her to do so. In Peter Vallentyne’s vivid example, a person may avoid exposing herself to the risk of being struck by lightning by always lying on the ground, but the risk in question is not reasonable avoidable to her.41 Similarly, we may avoid the risk of accidents in traffic by staying indoors, but always doing so is not reasonable. In Stemplowska’s related case, a person may avoid exposing herself to the risk of catching a disease by refraining from appearing in public, but it would not be reasonable to expect her to do so.42 Claiming that we should compensate in such cases amounts to an invocation of the widely share luck Hansen and Søren Flinch Midtgaard ‘Sinking Cohen’s Flagship―Or Why People with Expensive Tastes Should not be Compensated’, Journal of Applied Philosophy, 28 (2011), 341-354. 40 Stemplowska ‘Making’, p. 252. 41 Vallentyne ‘Brute Luck, Option Luck, and Equality of Initial Opportunities’, p. 533. 42 See Stemplowska’s example in ’Reconciling’, pp. 127-128. It is confusing in this context that Stemplowska refers to the risk as being ‘reasonably avoidable’ to the agent (when it is, in fact, precisely not reasonably avoidable). However, the nature of the case seems to me to be similar to cases mentioned by others to emphasize that some disadvantages may be avoidable, but not reasonably so (see, G.A. Cohen, ‘Expensive Taste Rides Again’, Dworkin and His Critics, ed. J. Burley, Oxford, Blackwell Publishing, 2004, 3-29; Shlomi Segall, Health, Luck, and Justice, Princeton and Oxford, Princeton University Press, 2010, 19-24; Vallentyne ‘Brute Luck’). Moreover, it is misleading, it seems to me, in the context of such cases, to speak, as Stemplowska does, about it being costly or involving unacceptable sacrifices for people to ‘act prudently’. It would be imprudent for people to avoid the risk in question which is why its eventual materialization should be conceived as an instance of compensable bad brute luck (cf. note 16*). 16 egalitarian commitment to the elimination of the unequalizing effects of brute luck. It does not add to our understanding of the question of stakes.43 With regard to the first aspect of Stemplowka’s proposal, I conclude that she fails to establish that there are justified deviations from the desideratum of non-exploitation. What she does it to reaffirm a shared luck egalitarian commitment to neutralizing the differential effects of brute luck. I turn now to considering the other side of Stemplowska’s proposal, to wit, the idea that sometimes the interest of the prudent in not having to provide security of holdings are weightier than the interests of the imprudent in enjoying such security. For example, Stemplowska submits that when we assess the relative weight of the interests of people in being able to pursue high risk activities, such as bungee jumping, uninsured whilst still being assisted if the need arises, on the one hand, and the interests of others in not having to provide such security, on the other, the latter interests are weightier.44 Considering the case of ‘a mountain climber who insists on going uninsured and unprepared on numerous mountain expeditions’ she rhetorically asks: ‘Does he or she really have a claim of egalitarian justice to be saved free of charge each time he or she needs it?’ and ‘[s]hould the climber’s claim on resources win each time against the claim of others in virtue of the fact that if he or she does not get the resources he or she will be left to die?’45 Moreover, it is impermissible, because unacceptably paternalistic, Stemplowska thinks, to compel people to take out insurance.46 By consequence, some imprudent people may end up in situations of insufficiency, and we have no reasons (or at least no reasons pertaining to choice-sensitive 43 Cf. Olsaretti ‘Responsibility’, p. 177 n. 14. 44 Stemplowska, ’Making’, p. 250. 45 Stemplowska, ’Making’, p. 252. Italics in original. 46 Stemplowska, ’Making’, p. 249. 17 egalitarian justice) to alleviate their suffering or indeed to save them from dying. Allowing instances of insufficiency of this kind is not, according to Stemplowska, implausible, and the reply to Anderson is that focusing clearly on the cases in which the luck egalitarian does indeed countenance situations of insufficiency reveals that this is highly plausible and not an unwelcome implication of the luck egalitarian’s view. This part of the Stemplowska’s proposal is, I believe, impaled by the sufficiency horn of the trilemma. The intuitions driving Stemplowska’s objection to securing holdings or situations for uninsured climbers is the implausibility and injustice of allowing his or her imprudent conduct, conduct that is reasonably avoidable to him or her, to eat into or to trespass on the equal resource shares of others. As suggested by my discussion of the case for externalization in some cases, I share these intuitions. However, conjoining resistance against exploitation with a traditional liberal repugnance against paternalism has the, I believe, unpalatable consequence of allowing some people to bring disaster on themselves. Accordingly, we should, I propose, curb our standard liberal repugnance against paternalism. Restricting the liberty of people for their own good or to prevent self-regarding harm, is plausible particularly in cases where insufficiency threatens; and as a desirable side-effect of this measure for securing sufficiency it protects the interests of others in not having to finance the costly and perhaps morally problematic conduct of others. I turn now to spelling out this argument in defence of a paternalistic account of stakes. IV. A paternalistic account of stakes Legal paternalism, in Feinberg’s description, considers it ‘always a good and relevant (though not necessarily decisive) reason in support of a criminal prohibition that it will prevent harm (physical, 18 psychological, or economic) to the actor himself.’47 The case for interventions is especially compelling when the self-regarding harm in question is far-reaching, grave, and irreparable. The reason we have to prevent the harm in question should always be weighed against the reason we have for protecting an agent’s voluntary self-regarding choice.48 As pointed out by Peter de Marneffe, a paternalistic intervention, by interfering with a person’s liberty or self-direction, always curtails his autonomy to some extent or in some respect.49 However, the autonomous choice blocked may constitute a great threat to the person’s autonomy, all things considered. Hence the intervention may preserve and enhance ‘for the individual his ability to rationally consider and carry out his own decisions’.50 When the threat to the latter posed by a voluntary self-regarding choice is far-reaching, grave, and irreparable then, as Richard Arneson forcefully puts it, a hard coercive shove away from the bad, in the form, for example, of a criminal prohibition outlawing the choice, may be warranted.51 I conceive of autonomy along the lines of Joseph Raz’ influential conception.52 A person being autonomous, according to this, means, roughly, that he or she is part author of his or her own life, that he or she arranges it in accordance with his or her own will against the background of an adequate range of options (and revises it if he or she comes to see the need for this) being in possession of the requisite capacities for doing so. The value of autonomy understood in this way may be accounted for in different ways. In Dworkin’s seminal paper he makes it clear that autonomy should be conceived in 47 Feinberg, Harm to Self, p. 4. Cf. p. 12. 48 Shafer-Landau ’Liberalism’, p. 187. 49 Peter de Marneffe ‘Avoiding Paternalism’, Philosophy & Public Affairs, 34, 68-94, at p. 81. 50 Dworkin ‘Paternalism’, Monist, 1972, 64-84, at p. 83. 51 Arneson ’Feinberg’, p. 270. 52 See, Raz, The Morality of Freedom, esp. chap. 14 19 non-consequentialist terms, that is, as distinct from ‘the happiness or welfare, in some broad sense, of the individual’.53 Bou-Habib makes a similar proposal in his discussion of how one can plausibly account for the right to receive aid when in an emergency as being non-waivable.54 However, I believe that the value of autonomy is most plausibly accounted for in broadly consequentialist terms, that is, as being related essentially to the well-being of the person whose autonomy it is.55 Relevant examples of activities that constitute threats to autonomy of the sort indicated above include those involving or carrying a high risk of involving grave and irrevocable damage to vital capacities for an autonomous life such as a well-functioning brain, and activities carrying the risk of landing people in situations in which they are deprived of the requisite set of options for autonomous pursuits (with no recourse). Professional boxing and football are examples of the former kind.56 So is riding on a motorcycle without wearing a crash helmet. I will elaborate on this in a moment when I contrast the paternalistic analysis of this case with sufficientarian egalitarianism’ analysis of the same case. Activities of the latter kind include uninsured mountaineering where the default arrangement is that stranded uninsured climbers are not assisted. In this way, proceeding without insurance, may land a 53 Dworkin, ‘Paternalism’, p. 83. 54 It is an interesting question whether the non-consequentialist rendering of the value of autonomy proposed by Dworkin and Bou-Habib is in fact a form of moral paternalism (as opposed to standard welfare paternalism), but it is immaterial to the paper on hand. For the distinction see Dworkin ‘Paternalism’, Stanford Encyclopedia of Philosophy, first published Wed Nov 6, substantive revision Tue Jun 1, 2010, 1-8, at p 5. 55 Cf. Midtgaard ‘Non-Renounceable Rights’. 56 Nicholas Dixon ‘Boxing, Paternalism, and Legal Moralism’, Social Theory and Practice, 27 (2001), 323-344; Mark Fainaru- Wada and Steve Fainaru, League of Denial: The NFL, Concussions, and the Battle for Truth (New York: Crown Archetype, 2013). 20 person in circumstances similar to Raz’ ‘The Man in the Pit’.57 This man is unable to climb out of the pit or to summon help and he is left with options of a trivial nature inadequate for an autonomous life. As these examples make clear, sufficiency concerns of the kind that is centrally at stake in the debate between friends and enemies of luck egalitarianism―i.e. concerns related to serious injuries and emergencies where people need the assistance of others―fall squarely within the paternalistic account on offer. Return now to the case of protective helmets legislation at the centre of my criticism of the psychic harm argument relied upon by sufficientarian egalitarianism. The paternalistic analysis of this case along the lines sketched above is, in contrast to Feinberg’s own analysis, centrally, and in as noncoy way as one may wish for, concerned with the ‘death or injury’ of the motorcyclists or with the threat to his well-being constituted by his bareheaded riding. Specifically, it is concerned with the brain damage that may be the regrettable result of his imprudent behaviour; a result with detrimental effects on his autonomy and thus his well-being. The importance of avoiding such dreadful consequences for imprudent persons, often young people with an otherwise long life of potentially great value to themselves and others ahead of them, looms large among our reasons for imposing restrictions on their activities―indeed this is the consideration that, among others, stands out in a clear and compelling way. This consideration is arguably what warrants imposing protective helmet legislation. Whilst protective helmets legislation certainly restricts Bert’s liberty in non-trivial ways―for example, he cannot any longer ride for the wind in his hair which may be of some value to him―the restriction would on balance seem justified, given the protection offered by the helmet against Bert experiencing a serious loss of autonomy. 57 Raz, Morality of Freedom, pp. 373-374. 21 We may get an indication of the salience of the paternalistic reason for imposing helmet restrictions by abstracting from the psychic harm potentially involved in the case that Bert crashes and is injured. We may assume, for example, that Bert’s accident takes place on a deserted road during night and that no one is ever confronted with his accident (and that he has no relatives or friends worrying sick about what has happened to him); or we may presume that fellow motorists for some reason experience no psychic harm when confronted with the accidents of imprudent bareheaded motorcyclists (they “had it coming to them” they may think and pass on unaffected). Still, it seems to me that even in such circumstances there would be no discontinuity in our reasons for upholding the protective helmet requirement with goes to show that the salient reason at play pertains to the selfregarding harm that may befall imprudent motorcyclists.58 The upshot so far is that the paternalistic account of stakes is uniquely sensitive to what is arguably our fundamental concern with respect to situations of sufficiency―significant losses of wellbeing. Hence complementing the luck egalitarian view with such an account of stakes allows it an effective answer to charges regarding its apparent complacency with the suffering of the unlucky imprudent. My considerations so far stress the salience of the harm to the motorcyclists and downplay the harm to bystanders. Many would in one sense accept this point, even regard it as being uncontroversial. Surely, they may say, harm of the former kind is in some sense weightier in the broad sense of involving a more serious curtailment of the interests of the motorcyclists in comparison with the curtailment of interests suffered by the bystander from witnessing the accident. The harm to the bystander, it seems plausible to say, is dominated by the harm to the motorcyclists in that apart from 58 This point draws on Arneson’s hunting case in ‘Joel Feinberg’, pp. 271-274. 22 the physical injury experienced by the motorcyclists psychic harm will befall him too and although the nature of this is arguably different from the one experienced by the bystander it is plausible not of less significance or disutility. However, in another, perhaps more crucial, sense, they may say, the harm befalling the motorcyclist is less weighty or perhaps categorically different from the psychic harm that may befall others by virtue of his accident. The psychic harm befalling the bystander is one he incurs inadvertently: he could not reasonably have avoided it. The harm to the motorcyclist, on the other hand, is one he could reasonably have avoided (he could have worn a protective helmet). Scanlon’s ‘Curious vs. Walker’-case brings out the importance of the question of whether or not a given burden were reasonably avoidable to an agent in our assessment of the degree to which it is problematic that she bears that burden. 59 The officials of a city undertake a project of removing hazardous waste from a residential area to a safe spot. Walker is inadvertently exposed to chemicals released into the air (the warnings issued by the officials fail to reach him, and he goes for his daily walk when the waste is removed). Curious is exposed as well, but is so by his own making or in a way that is reasonably avoidable to him (having heard the warnings he still seeks out the site and climbs over the fence to get a closer look). Many regard the avoidance of the kind of harm experienced by Walker as of much greater urgency than is the avoidance of the harm experienced by Curious.60 If this view is true, it calls in question the priority I gave to the injuries, perhaps death, suffered by the biker in that it, as dreadful as it may be, is at the same time a harm that he could reasonably have avoided. To what extent it does 59 Scanlon, Owe , p. 257; Alex Voorhoeve ‘Scanlon on Substantive Responsibility’, Journal of Political Philosophy, 16 (2008), 184-200. 60 Cf. Voorhoeve ‘Scanlon’, p. 185. 23 so, depends, however, on the strength of the priority involved to the prevention of the kind of harm that people could not avoid by choosing appropriately over the prevention of the kind of harm that people can avoid in this way. On one view on the relative importance of preventing the two types of harms, it is of no importance morally to prevent harms that are reasonably avoidable to people. They have no reasons to complain if harms of this kind befall them. The view is related to what Scanlon calls the ‘Forfeiture View’ according to which ‘a person who could have chosen to avoid a certain outcome, but who knowingly passed up his choice, cannot complain of the result: volenti non fit iniuria’.61 It features centrally among Feinberg’s reasons for holding that we never have a reason to prevent voluntarily selfregarding harm. The latter does not count as a genuine form of harm at all (genuine harm, according to Feinberg, concerns only, roughly, wrongful setback of interests, where harm consented to, or voluntarily incurred, is not wrongful).62 This view would undercut the plausibility of my attempt to highlight the death or injury of the motorcyclist as the salient harm at issue. The view seems, however, just as the blanket dismissal of a reason to prevent voluntary self-regarding harm (however profound and irrespective of it being preventable by limited interferences with autonomy rights) it supports, implausibly extreme. A more plausible view is arguably suggested by Alex Voorhoeve in the course of his discussion of Scanlon’s view on responsibility: ‘[T]hough we are entitled to give priority to eliminating a danger of suffering a harm that one cannot avoid by choosing well over eliminating a danger of coming to harm by choosing badly, this priority is not absolute. As the magnitude of the danger of coming to harm through choosing badly increases vis-à-vis the magnitude of the danger of suffering harm that one cannot avoid by choosing appropriately, there is a point at which the claim to be rid of the danger of ending up badly through one’s 61 Scanlon, Owe, p. 259. 62 Feinberg ‘Legal Paternalism’, pp. 106-9; Harm to Self, pp. 10-11; Harm to Others, pp. 34-5, 115-7. 24 choices becomes stronger than the claim to be rid of the danger that one will suffer a harm one cannot avoid by choosing appropriately’. 63 Voorhoove applies this principle to a case in which a choice has to be made between two policies. Simplifying somewhat, we have to choose between the policy “Low Emissions” under which Walker is inadvertently exposed to a minor harm (an unpleasant cough for a week, with no lasting effects), while Curious suffers no harm, and the policy “Inform Everyone” under which Curious advertently suffers grievous harms (severe, permanent health problem), while Walker suffers no harm. In this case, Voorhoeve submits, we should opt for the former policy. The magnitude of the harm that Curious would incur through choosing badly under Inform Everyone vis-à-vis the magnitude of the harm that is reasonably unavoidable to Walker under Low Emissions is plausibly such that the point at which the prevention of the former harm become more important than the prevention of the latter harm has been reached. The situations discussed by Voorhoeve involve, as we have seen, choosing between the prevention of various kinds of harm. In this respect they differ from the question pertaining to the case of crash helmets and the relative salience of our reasons to prevent harm inadvertently or advertently incurred. The policy in question, to wit, making the use of crash helmet mandatory, prevents both kinds of harm. However, the principle invoked by Voorhoeve seems to be applicable also in an assessment of the relative salience of the reasons involved. The challenge to my paternalistic analysis of the case was that the harm incurred by the bystander is in the relevant sense the salient reason by reference to which it is justifiable to compel the biker to wear a helmet. However, it seems to me that Voorhoeve’s arguably plausible principle 63 Voorhoeve ‘Scanlon ’, p. 199. 25 challenges this assessment. Although our reasons to prevent a harm that one cannot reasonable avoid incurring are weightier or more salient than is preventing a harm that is reasonable avoidable to one, as the magnitude of harm that one may incur by choosing badly increases vis-à-vis the magnitude of the harm that one cannot avoid by choosing appropriately, at some point our reasons to prevent the harm that is reasonable avoidable to the agent emerge as the salient ones. And it seems to me that the point has been reached in this case. The biker, recall, risks injuries of a kind that affect his capacities for autonomous life profoundly, whereas the psychic harm risked by the bystander, although real, is clearly less serious. (It bears mentioning that the endorsement of Voorhoeve’s principle implies a slight revision of the luck egalitarian view. It implies namely that sometimes it is more important to prevent inequalities or disadvantages due to option luck than it is to prevent inequalities due to the differential effects of brute luck. This does not strike me as implausible, and it is consistent with the stakes focus on capping certain disadvantages pertaining to option luck. However, the plausibility of the indicated revision is obviously worthy further discussion although I cannot provide it here.) This allows me then to fend off the challenge to my paternalistic account of stakes to the effect that what it takes to be the salient reasons supporting policies aimed at securing sufficiency are not in fact so. Accordingly, we may reaffirm the significant virtue of the paternalistic account of stakes highlighted above, to wit, its compelling reply to Anderson’s concern about situations of insufficiency. Add to this that the policies or strategies associated with paternalism of a kind that is focused on the prevention of self-regarding harm64 are internalizing one’s, that is, roughly, policies that secures sufficiency by restricting the liberty of risk takers in various way, not by countenancing cost displacement or exploitation. 64 Cf. Note 10. 26 However, we need to address a concern that may call in question precisely the latter virtue claimed for the paternalistic account. It is one thing to put in place various kinds of internalizing measures in order to preserve sufficiency while avoiding exploitation. Another is to what extent people comply with these measures. People may find ways around internalizing measures. By doing so they may come to stand in need of our assistance, and providing it will be a public charge. It seems them that we can either secure sufficiency or avoid exploitation. We cannot, it seem, meet the two desiderata simultaneously. Consider, for example, Stemplowska’s case of a mountain climber that manages to set off uninsured and where ‘the costs of any rescue operation are so high that the mountain climber would not be able to cover them except through purchasing priori insurance’ (i.e., rescuing him would necessarily involve a drain on public fund in that any policy requiring him to pay for the rescue after the event will fall short of covering the costs involved).65 Should we, in such cases, at the cost of countenancing exploitation, ensure sufficiency? We should begin by noticing that there are viable paths of reducing the problem―paths in line with the general drift of the paternalistic account. That is, we may adopt supplementary internalizing measures. We may, for example, put in place a scheme that requires uninsured risk takers to contribute to covering the costs of rescue operations. This ex hypothesis (in the case of the mountain climber) is insufficient to cover the cost of the relevant rescue operations. However, the mountain climber having to pay a large amount to contribute towards her own rescue may serve as deterrence from embarking on expeditions uninsured. Surely this policy will not deter everyone. To some individuals the importance of undertaking a dangerous endeavour, involving a costly rescue operation if it misfires, may be worth the risk of having to pay monthly contributions for the rest of their lives, even of a magnitude that 65 Stemplowska, ’Making’, p. 252. 27 barely allows them to avoid destitution. Still it may deter a significant number of individuals from undertaking risky expeditions uninsured. Moreover, note another characteristic of the mountain climber emphasized by Stemplowska: she ‘insists’ on going uninsured ‘on numerous mountain expeditions’ which raises the question whether she should be assisted ‘each time’ she needs it.66 In brief, she has a history of going on dangerous expeditions uninsured or is highly disposed to do so. If this is so, it makes relevant further supplementary internalizing measures. Especially, society may justifiably put the mountain climber under some kind of restraining order or add her name to a list of person that rangers may justifiably turn away from embarking on mountain expeditions. It may do so to preserve sufficiency or to avoid situations of insufficiency while at the same time catering to non-exploitation. In these and other ways the problems raised by insurance evaders may be reduced considerably. Still, some may, of course, find ways of circumventing these further measures too. In this case, and assuming that there would not be many such climbers, I believe that the Voorhoeve’s principle endorsed above recommends that we should, in the interest of preserving sufficiency, accept a minor infringement of the principle of non-exploitation. The magnitude of the unavoidable harm imposed on the ordinary citizen by having to contribute to a scheme which assist insurance evaders in need is limited vis-à-vis the magnitude of the avoidable harm befalling the insurance-evading mountain climber in need of assistance. Assisting insurance-evading mountain climbers may for the indicated reasons be a plausible policy. However, in some cases a conflict may exists between assisting insurance evaders and assisting others who inadvertently come to stand in need of assistance. Take Arneson’s illuminating case of, 66 Stemplowska, ’Making’, p. 252. 28 roughly, a choice between saving a groups of negligent hikers and a group of non-negligent hikers, where their choices or how they have exercised their responsibility is the only potentially morally relevant factor distinguishing them.67 Here the magnitude of the harm is on a par, and it seems clear to me that we should give priority to rescuing the latter. Hence my favoured account delivers what I take to be the intuitively plausible view with regard to this case (that responsibility makes a difference). As the injury or predicament of the non-negligent group declines, moving away from a situation of genuine insufficiency, while the predicament of the imprudent remains stable it becomes more plausible that we should give priority to the foolish groups of hikers. Imagine for example that we face a choice between rescuing a non-negligent group of hikers to spare them the discomfort of having to stay the night in the open and rescuing a foolish hiker who will lose much blood and eventually die if left unassisted, it seems to me that the Voorhoeve principle, plausibly, implies that we should rescue the latter. The upshot is that although non-compliance generates problems for the paternalistic account the latter appears capable of handling them, and to do so in ways consistent with its guiding principles. V. Conclusion The question of stakes―in essence, the question about the kind of disadvantages it is plausible to attach to features of persons for which it is plausible to hold them responsibility―is in many ways a fertile ground for the further development of luck egalitarian theories. It requires us to complement the luck egalitarian preoccupation with responsibility in an important respect, and plausible ways of doing so may serve to rebut apparently forceful objections levelled against the view, such as Anderson’s 67 Arneson ’Luck Egalitarianism and Prioritarianism’, Ethics, 110 (2000), 339-349, at p. 348. 29 objection concerning luck egalitarianism’s apparent acceptance of sever cases of absolute deprivation. However, leading answers to the question of stakes fail to solve the trilemma facing such account. That is, they are incapably of meeting the desiderata of sufficiency, non-exploitation and anti-paternalism simultaneously. Whilst both sufficientarian egalitarianism and Stemplowska’s proposal appear to meet the desideratum of non-paternalism, the former flouts the crucial sufficiency desideratum. It does so by relying on a problematic psychic harm argument. This argument shows considerable coyness in trying to account for the importance of securing sufficiency. Each of the two prongs of Stempolwska’s proposal face is subject to crucial objections too. First, it fails to show that there are in fact cases in which it is plausible to deviate from the doctrine of non-exploitation. Second, it, implausibly, in some cases denies both externalization policies and paternalistic measures. This brings it in conflict with the sufficiency desideratum. The paternalistic account, I have argued, suggests a more plausible way in which to handle the trilemma. First, it recommends that we should in certain cases deviate from a stringent principle of nonpaternalism. Paternalism is sometimes justified and particularly so in cases in which the good at stake or the harm the may be prevented is profound and the interferences of autonomy required to do so are relatively limited. Cases of insufficiency fall within this set of cases. Moreover, preventing insufficiency on this ground is uniquely sensitive to the concerns raised by Anderson’s objection. It is so by revolving around what is arguably our salient reason for being concerned to prevent such situations from arising, namely the detrimental effects that such situations may have on the lives of those experiencing it―how such events detract in significant ways from the well-being of people involved in them. Furthermore, because the paternalistic account suggests internalizing policies for preserving sufficiency, it satisfies the desideratum of non-exploitation; and while this virtue of the 30 proposal is called in question in light of the problem of non-compliance or insurance evasion, the paternalistic account is capable of handling it in a plausible way. Some readers might worry that my attempt to cap the kind of absolute disadvantages that luck egalitarian should countenance partly detracts from the attractiveness of the view.68 The attraction supposedly is along the lines of G.A. Cohen’s suggestion to the effect that Dworkin and other luck egalitarians have ‘performed for egalitarianism the considerable service of incorporating within it the most powerful idea in the arsenal of the anti-egalitarian right: the idea of choice and responsibility’.69 However, I believe that such worries can relatively easily be put to rest by drawing attention to the modest sufficiency threshold relied upon in this paper. The restricted nature of the latter may indeed raise renewed concerns about the pligt of imprudent choosers making bad choices although not some that land them in situations of insufficiency. Addressing such worrries will have to await another occasion. 68 69 Jessica Cohen, Currency, p. 32. 31