Justice for Foxes n Knight, C. (2015) Justice for foxes. Law and Philosophy, 34(6), pp. 633-659. There may be differences between this version and the published version. You are advised to consult the publisher’s version if you wish to cite from it. http://eprints.gla.ac.uk/107363/ Deposited on: 16 June 2015 Enlighten – Research publications by members of the University of Glasgow http://eprints.gla.ac.uk 1 CARL KNIGHT JUSTICE FOR FOXES* ABSTRACT. Ronald Dworkin maintains that value is unitary, in the sense that different values do not conflict. This article resists this ‘hedgehog’ view with reference to the values of equality and utility. These appear to yield conflicting prescriptions in cases where one possible distribution gives different individuals the same amount of advantage, and the other contains an unequal distribution of a greater overall amount of advantage. Hedgehogs might respond to such a case in two ways. First, they might claim that equality and utility are not truly in conflict. However, this claim seems implausible on our ordinary concepts of these values, and Dworkin does not provide grounds for revising our concepts. Second, they might deny that one of these values—utility—is a genuine value. However, one of the two aspects of Dworkin’s fundamental principle of equal concern appears to be supportive of utility, and he offers no good arguments for his preferred strategy of accommodating this aspect of equal concern within the value of equality. Furthermore, the alternative ‘fox’ view, which recognizes equality and utility as conflicting values, each accommodating one aspect of equal concern, has the advantage that it stays truer to familiar moral concepts. Moral, political and legal decision-making is strongly responsive to value judgments. Where an agent recognizes some action as furthering a value, they have a reason to perform that action. In considering which course of action to follow, an agent does then * This article is in press, as of June 2015, at Law and Philosophy (http://www.springer.com/law/journal/10982). 2 immediately face two questions. First, which things count as values? And second, can values conflict? Fundamental disagreements about public policy are often grounded in contrasting answers to these questions (and the further question of how to resolve value conflicts if they do arise). Addressing these conceptual questions may then have great practical importance, as a large literature in moral, political, and legal philosophy recognizes. In his last major work, Justice for Hedgehogs, Ronald Dworkin decisively answered the second question in the negative.1 His main claim is that value is unitary, in the sense that its different components complement each other rather than conflict. He endorses what he calls “full value holism—the hedgehog’s faith that all true values form an interlocking network, that each of our convictions about what is good or right or beautiful plays some role in supporting each of our other convictions in each of these domains of value.”2 This holism requires that “there are no genuine conflicts in value.”3 Dworkin’s primary conceptual resource in support of the thesis of the unity of value is his account of values as being ‘interpretative concepts’. On this account, the meaning of any given value can only be established by examining other values. As Dworkin puts it, a defense of some particular conception of a political value like equality or liberty must draw on values beyond itself. … We cannot defend a conception of any of 1 Ronald Dworkin, Justice for Hedgehogs (Cambridge, MA: Harvard University Press, 2011). 2 Dworkin, Justice for Hedgehogs, p. 120. 3 Dworkin, Justice for Hedgehogs, p. 119. 3 them without showing how our conception fits with and into appealing conceptions of the others. That fact provides an important part of the case for the unity of value.4 Presumably Dworkin does not here mean to say that conceptions of values must draw on other values in the specific sense that they must not conflict with them. Were that the case, his interpretative concept view of values could not “be an important part of the case for the unity of value,” as it would include the unity of value among its premises. He must rather mean that we take into account other values when developing our conception of a particular value, and that it is only once we do this that it becomes apparent that values do not conflict. I broadly accept this interpretative concept approach to value. But as William Edmundson has commented, “one might worry, Is there any reason to assume that the best interpretation of something will render that thing as a unified whole, especially if the ‘package of purposes’ attributed to it is rich and diverse?”5 I argue that, once we consider how to describe each of two major distributive justice values, in light of the role of the other value, it becomes apparent that we can readily conceive of them as conflicting. That is, among values, there are (at least) two that conflict, given the role each plays in relation to the other. 4 Dworkin, Justice for Hedgehogs, p. 7; see also Ronald Dworkin, Law’s Empire (Harvard: Harvard University Press, 1986), chs 1-2. 5 William A. Edmundson, “Book Review: Ronald Dworkin, Justice for Hedgehogs,” Law and Philosophy, 31 (2012), 759-767, p. 763. 4 One of these values is equality. It is concerned with reducing differences in advantage levels between the well off and the badly off. I make no assumptions about the kind of advantage with which equality is concerned; candidates include welfare, resources, and capabilities.6 I also set aside choice and responsibility. Those who find arguments for equality convincing only where individuals’ responsibility has been equivalently exercised can simply assume that it has been so exercised in my examples.7 The other value is concerned with social maximization, which may be interpreted as achieved by the highest total amount of advantage, by the highest average of advantage, or in some other way.8 Dworkin refers to this value when mentioning the “familiar idea in political theory that a just society will make some compromise between efficiency and distribution.”9 Elsewhere in this passage Dworkin is explicit that he is describing a conflict between utilitarianism and egalitarianism, and efficiency is often used to refer to non-advantage maximizing concepts (such as Pareto optimality), so I will refer to the value in question as utility. I will, however, interpret that value quite broadly. 6 The typically greater marginal utility of benefits for the worse off compared to the better off will tend to make real world inequalities more pronounced when measured in welfare rather than resources. This has no bearing on the examples I use, which should be interpreted as referring only to whichever measure of advantage the reader favours. 7 It should therefore not be complained that my value of equality would reward those who deliberately develop expensive tastes. See Richard J. Arneson, “Equality and Equal Opportunity for Welfare,” Philosophical Studies, 56 (1989), 77-93; G. A. Cohen, “On the Currency of Egalitarian Justice,” Ethics, 99 (1989), 906-44. Dworkin inspired these ‘luck egalitarians’, but refuses to describe himself as a luck egalitarian; see “Equality, Luck and Hierarchy,” Philosophy and Public Affairs, 31 (2003), 190-8, p. 191. 8 See Derek Parfit, Reasons and Persons (Oxford: Oxford University, 1984), part IV. 9 Ronald Dworkin, Sovereign Virtue (Cambridge, MA: Harvard University Press, 2000), p. 54. 5 Though my value of utility is characterized by its maximization of advantage, advantage need not be construed in welfarist terms, as the most familiar utilitarian theories interpret it. It might be welfare, or alternatively resources, capabilities, or whatever is thought to be important.10 I also do not, and need not, resolve a tricky internal interpretative issue for utilitarians: the difference between average and total utilitarianism is only felt where population size varies, and it is stable in my examples. In order to defeat the hedgehog’s thesis, and substantiate the rival ‘fox’ thesis that values, even on their best interpretations, can come into conflict with each other, we need only show that a rather weak kind of value conflict is possible. I have in mind conflicts of pro tanto reasons for action.11 One value provides us with a pro tanto reason for acting in one way, the other value provides us with a pro tanto reason for acting in a second way, and we cannot act in both ways.12 That such a conflict between equality and utility is possible seems obvious. Consider this example: Equality versus utility: two distributions are available, one of which secures two units of advantage for each half of the population, the other of which secures one 10 Hence those impressed by arguments to the effect that utilitarianism as usually conceived relies on some dubious conception of welfare (see, for instance, Ronald Dworkin, Taking Rights Seriously (London: Duckworth, 1978), pp. 233-8; Sovereign Virtue, ch. 1) cannot object to my value of utility on that score. 11 See Susan Hurley, Natural Reasons (Oxford: Oxford University Press, 1989), ch. 7. 12 One plausible basis for these conflicting pro tanto reasons are conflicting judgments of value, which are central to Joseph Raz’s definition of value conflict; see his The Practice of Value (Oxford: Oxford University Press, 2003), p. 44; cf. Michael Stocker, Plural and Conflicting Values (Oxford: Oxford University Press, 1990), pp. 86-7 6 unit of advantage for a worse-off half of the population and five units of advantage for the better-off half. Assuming (as I do) that the population size is stable and that the advantages of the better off are not traceable to their choices, equality is uncontroversially best furthered by the first distribution, because it is equal.13 Utility is uncontroversially best furthered by the second distribution, because it contains more advantage for society (and a higher average advantage level). As equality gives us a pro tanto reason for selecting one distribution, utility gives us a pro tanto reason for selecting the other, and we cannot have both, these values are in conflict. Furthermore, when we are conceptualizing equality, the fact that its recommendation conflicts with utility’s gives us no obvious reason to revise our conceptualization. Similarly, when we are conceptualizing utility, the fact that its recommendation conflicts with equality’s gives us no obvious reason to revise our conceptualization. As Dworkin acknowledges, legal and political philosophers routinely treat these two interpretative concepts as conflicting. That this pro tanto conception of conflict is weak is brought out by comparison with two alternative conceptions. First, my sense does not imply that anyone who endorses conflicting values must face indeterminacy. Dworkin understands indeterminacy as holding “not that some particular act is neither forbidden nor required, 13 The first distribution is also better for the worst off in absolute terms. Thus, any readers who might reject the value of equality, understood as a view concerned with relative advantage levels, in favour of a focus on the absolute advantage levels of the worst off, as Rawls’ difference principle suggests, can nevertheless agree that there is a conflict in values in this and similar examples presented later in the article. 7 but that there is no correct answer to the question whether it is forbidden or whether it is required.”14 My claim is not that, where equality and utility conflict, there is necessarily no correct answer as to what we should do. I make only the weaker claim that equality and utility give rise to mutually incompatible pro tanto reasons. As these are pro tanto reasons, there may be no problem with resolving the conflict determinately.15 A position intermediate between straightforward resolution and indeterminacy is what Dworkin calls “uncertainty,” which applies where considerations on each side seem finely balanced and one takes “no view of the matter.”16 I do not take a stand on which of these three positions to take. Whether or not value conflicts can be resolved, their existence would make trouble for the hedgehog. 14 Dworkin, Justice for Hedgehogs, pp. 88-9. Indeterminacy seems to be implied by what Robert Talisse refers to as metaphysical pluralism; see his Pluralism and Liberal Politics (Abingdon: Routledge, 2012). 15 There is a diversity of views about how this might be done. Some, such as James Griffin, see value pluralism as consistent with commensurability; see “Incommensurability: What’s the Problem?” in Ruth Chang (ed), Incommensurability, Incomparability, and Practical Reason (Cambridge, MA: Harvard University Press, 1997). Others, like William Galston, argue that we can choose between even incommensurable values on reasonable grounds; see his Liberal Pluralism (Cambridge: Cambridge University Press, 2002), p. 35; see also Stocker, Plural and Conflicting Values, ch. 7. George Crowder further distinguishes between particularist and universalist approaches to choosing between conflicting values; see Liberalism and Value Pluralism (London: Continuum), ch. 3. For discussion of trade-offs among reasons for action, see Gerald Gaus, The Order of Public Reason (Cambridge: Cambridge University Press, 2011), ch. 3. 16 Dworkin, Justice for Hedgehogs, p. 91. Dworkin’s ‘uncertainty’ roughly corresponds to Talisse’s preferred epistemic pluralism. 8 Second, though I say equality and utility conflict, that does not mean that, whatever we do, we do some wrong. Isaiah Berlin, who brought the fox/hedgehog terminology to prominence, combines his foxy position with the view that the conflict of value leaves (in Dworkin’s words) “an irredeemable moral stain” and that we must sometimes choose “not whether to wrong a group, but which group to wrong.”17 But though Dworkin only discusses foxy views which have this ‘tragic’ character, it is evidently quite detachable from the recognition of value conflict itself. A non-tragic alternative suggests that, if we bring one of the distributions in equality versus utility about, after due consideration of the reasons presented by equality, utility, and any other relevant values, we can quite plausibly hold that we do no wrong, though we inevitably do not promote one value as it conflicts with another.18 Our obligations extend only to the possible. I again take no stance here, and only mention this non-tragic view to show that, if the tragic view is indeed mistaken, we need not rush to accept Dworkin’s value holism. But is there such a thing as value conflict? Examples such as equality versus utility cannot be batted away by appeal to the fact that values are interpretative, as the conflict they present is familiar and fits with our wider conceptual scheme. Dworkin must then appeal to conceptual resources other than, or in addition to, his main idea that values are interpretative concepts. He might take one of two strategies in order to preserve the thesis of the unity of value. First, he might deny that the conflict is genuine, by arguing 17 Ronald Dworkin, Justice in Robes (Cambridge, MA: Harvard University Press, 2006), p. 109; see also Isaiah Berlin, Liberty (Oxford: Oxford University Press, 2002). 18 This view seems to arise quite naturally from Dworkin’s powerful objection to Berlin that he would require that there is something wrong in preventing murder, as it is a restriction on Berlin’s idea of liberty; see Dworkin, Justice in Robes, pp. 115-6. 9 (for it cannot, on pain of circularity, be assumed) that values, qua interpretative concepts, respond to other values in such a way that value conflict is impossible. Second, he might deny that one or other of the putatively conflicting values really is a value. The second of these strategies appears to be favoured by Dworkin, as he does not appear to believe that utility is a value. Given this, it is unsurprising that Dworkin does not attempt to show that equality and utility are compatible, as the first strategy suggests; nevertheless, his comments regarding other values suggest the broad lines of a Dworkinian defence of the compatibility of equality and utility. I treat the first strategy, which denies the conflict between equality and utility, in the first part of this article, and the second strategy, which denies the value of utility, in the second part. I argue that, as neither strategy succeeds, equality and utility conflict. In the final part I complement these mostly defensive arguments with the positive claim that the fox’s conceptual scheme has an advantage over the hedgehog’s. I. IS THERE REALLY A CONFLICT? Richard Fallon suggests that, where your colleague solicits comments on a book manuscript, and you find it to be bad, you face a conflict between the values of honesty and kindness.19 This has at least a superficial similarity with my claimed conflict between equality and utility. It is therefore worth seeing whether Dworkin’s response to Fallon succeeds. 19 Richard Fallon, “Is Moral Reasoning Conceptual Interpretation?” Boston University Law Review, 90 (2010), 535-550, p. 539. 10 Dworkin initially draws attention to the aforementioned distinction between uncertainty and indeterminacy.20 It could easily be the case that, say, honesty was the best policy in this case, but that it was just hard to know as much. As Dworkin acknowledges, this point does nothing to resolve the apparent conflict, so we can quickly set it aside. Though Dworkin writes that “[i]t is … natural to say in a case like Fallon’s that we are torn between kindness and honesty” he holds that “[w]e might disagree, however, as to why it seems natural.”21 In support of this suggestion he suggests that our concepts of honesty and kindness are, as interpretative concepts, works-in-progress. They can both be satisfied in most cases, but in others, like this case, they are in tension with each other, and so we understandably report a conflict. But this shows only that the concepts we have been working with are not the finished article. We arrive at further refinement by asking questions such as “whether it is really cruel to tell an author the truth” and “whether it is really dishonest to tell him what it is in his interests to hear and no one’s interests to suppress.”22 Dworkin needs one more step of argument to establish the possibility of values not truly conflicting: he must offer reasons for thinking that our process of refining our values has the result that values do not conflict.23 But the suggestions Dworkin makes 20 Dworkin, Justice for Hedgehogs, pp. 118-19. 21 Dworkin, Justice for Hedgehogs, p. 119. 22 Dworkin, Justice for Hedgehogs, p. 119. 23 A referee comments that “Dworkin might well hold that we, qua finite human beings with limited cognitive capacities, can never reach a level of insight in which all value conflicts are dissolved. However, counterfactually fully rational and well-informed agents might arrive at such a conclusion.” Gerald Gaus 11 here are problematic. He first asserts that “[w]e reinterpret our concepts to resolve our dilemma: the direction of our thought is towards unity, not fragmentation.”24 On the one hand, it may be that progress in thinking tends towards unity in the sense of coherence or consistency, but there is nothing incoherent or inconsistent in the thought that honesty and kindness might conflict. So a tendency towards unity in that weak sense cannot be used to support a movement away from conflicting concepts of honesty and kindness. On the other hand, it is not at all obvious that progress in thinking tends towards unity in Dworkin’s favoured ‘no inter-value conflict’ sense. Indeed, that is roughly what the argument under consideration is supposed to demonstrate, so it cannot serve as a premise. The reference to a dilemma may be intended to support this strong claim, but the dilemma is not such a threat that we should rework our moral concepts to avoid it. Situations in which we feel we can be honest or kind, but not both, are familiar parts of moral life.25 A non-conflicting conceptual scheme may, then, be unacceptably at odds similarly argues that “given the complexity of our belief systems, it is, other things equal, always plausible to conjecture that any given disagreement is an instance of inconclusive reasoning.” See his Justificatory Liberalism (Oxford: Oxford University Press, 1996), pp. 155-6. That may be so, but the assertion that rational beings would accept the unity of value is prima facie no more plausible than the assertion that fully rational beings would accept value conflict. Dworkin would still need to offer reasons for thinking that fully rational beings would tend towards a unity rather than plurality of values. These reasons would take much the same shape as the reasons I consider later in this paragraph. 24 Dworkin, Justice for Hedgehogs, p. 119. 25 Of course, the role of such intuitions in ethics is controversial; see, for instance, Peter Singer, “Ethics and Intuitions,” The Journal of Ethics, 9 (2005), 331-52; Folke Tersman, “The Reliability of Moral Intuitions,” Australasian Journal of Philosophy, 86 (2008), 389-405. But Dworkin himself aims to build theories that 12 with our considered judgments.26 We might be suspicious that it overlooked something that mattered morally. The other way in which Dworkin tries to bolster the view that refinement removes conflict is through critique of the rival view that “[m]oral conflict is real, and any theory that denies this is false to moral reality.” This rival holds that the conflict between honesty and kindness “is not an illusion produced by incomplete moral interpretation; it is a matter of plain fact.”27 Dworkin objects that it is unclear what this plain fact could consist in, given that there are no moral particles to fix the content of value. But this objection relies on lumbering the rival view with unnecessary metaethical baggage. The relevant rival view need not take a stance on such issues and, as I have suggested, might even accept Dworkin’s position that moral concepts are interpretative concepts, a matter for substantive moral philosophy which takes into account the role of other concepts.28 It just says that moral concepts are properly construed as in principle open to conflict. Against this view Dworkin claims that the concepts of honesty and kindness do not “have a precise and conflicting content just in virtue of linguistic practice.”29 He are compatible with ordinary people’s understanding of ethics; see, for instance, Sovereign Virtue, pp. 289- 90. 26 William Galston, The Practice of Liberal Pluralism (Cambridge: Cambridge University Press, 2005), p. 17. 27 Dworkin, Justice for Hedgehogs, p. 120. 28 The “radical pluralism” of G. A. Cohen, which is “consistent with many views of the status of our fundamental convictions, including the view that they are noncognitive expressions of emotion or attitude,” is one form of the relevant view. See his Rescuing Justice and Equality (Cambridge, MA: Harvard University Press, 2008), p. 4 n. 5. 29 Dworkin, Justice for Hedgehogs, p. 120. 13 contends that “[m]any people do believe that it would be an act of kindness to tell your colleague the truth. Or that it would not be dishonest, in this circumstance, to trim.”30 The evidence Dworkin cites is insufficient to establish even that refinement of our conflicts leads away from conflict. First, and to repeat, the task before Dworkin is not to show that there are some cases in which an apparent value conflict dissolves under closer inspection. His claim is that values never conflict, so any case of genuine value conflict defeats his position. Second, even supposing that Dworkin’s claims about people’s beliefs were true, they would not really support even the possibility of honesty and kindness being in harmony. It is not enough just to observe what “[m]any people … believe” about values, as they might be mistaken about them. As Dworkin himself emphasizes, the correct interpretation of such arguments is a matter of moral argument, not mere head- counting.31 Finally, and bearing in mind the previous two points, Dworkin does not consider really difficult (and hence, most relevant) counterexamples to his position, such as equality versus utility. In the kindness and honesty case there is vagueness about whether the colleague is really harmed by the truth, such that some people may hold that telling the truth is simply ‘being cruel to be kind’. But in equality versus utility, there is no such vagueness—no one could claim that the equal outcome promoted utility or that the unequal outcome promoted equality. 30 Dworkin, Justice for Hedgehogs, p. 120. 31 See, for example, Dworkin, Justice for Hedgehogs, pp. 170-1. 14 What might one who is minded to deny this conflict say? I will now briefly revisit the points Dworkin makes regarding honesty and kindness, to see if they fare any better as a way of resolving the equality and utility conflict. The starting point of a Dworkinian attempt to reconcile equality and utility would be to treat our versions of these concepts as incomplete, and in need of refinement. We refine them by asking questions such as “is it really contrary to efficiency to favour four equally distributed units of welfare to six unequally distributed units?” or “is it really contrary to equality to prefer six unequally distributed units of welfare to four equally distributed units?” The problem for the Dworkinian is that such questions seem, if anything, more obviously to give rise to answers that show value conflict than did the honesty/kindness questions. At least assuming, as any attempt to reconcile the values of equality and utility assumes, that equality and utility are separate values,32 it seems clear that equality will not favour (unchosen) inequality over an equal distribution, and utility will not favour less advantage rather than more. What about the suggestions that I earlier said Dworkin offered for the refinement of values as a process that removes conflict? His initial assertion that “the direction of our thought is towards unity, not fragmentation” does, as before, either fail to challenge value conflict, assume rather than argue for its falsity, or argue for its falsity on the ground that we are so unsettled by the thought of two values pulling in different directions that we must revise one to fit with the other. This argument is no more plausible for equality and 32 The thought here is that, if one of these values were not a value, we might try to broaden the other value in an otherwise counterintuitive way in order to accommodate considerations traditionally associated with the other value. But as we are here assuming that both values actually are values, there is no motivation for a counterintuitive broadening of meaning. That assumption is relaxed in the next section. 15 utility than it was for honesty and kindness. It is routine to think that a distribution of benefits and burdens is good in one regard, but bad in another. If one country is rich, but unequal, and another country equal, but poor, it is natural to think that there is something good and something bad about both countries. If we reframed our concepts so that to say that a country was rich was to imply that it was equal, and vice versa, we would have lost clarity in our thinking. And, though it is harder to judge advantage levels empirically, the same conceptual point would follow from an attempt to say that an equal distribution was necessarily a distribution high in overall advantage, and vice versa. It is not value conflict but the absence of value conflict that is unsettling. We certainly have no reason to invent it, as Dworkin urges. There is, finally, a quite different argument specific to the distributive sphere that Dworkinians might offer for the unity of value.33 Suppose we all face a particular risk— the risk of congenital disease, for example. This risk, when it materializes for some of us but not others, will cause an unchosen unequal distribution between us. Dworkin has long argued that we should correct for that unchosen inequality only to the extent that we would have insured against it under the equal background conditions of equal risk and an equal ability to purchase insurance.34 Under such equal background conditions, we may well decide that we would not want to purchase insurance that completely restores us to equality after risks materialize for us if that scheme would impose extreme costs on us in the form of insurance premiums. We may well prefer an insurance scheme that would compensate us to a lesser extent than what would be necessary to restore us to equality 33 The remainder of this paragraph is attributable, with only one minor change, to an anonymous referee. 34 Dworkin, Sovereign Virtue, ch. 2. 16 with others. If the hypothetical insurance model is a plausible guide to what the value of equality requires, then, contrary to what I have argued, an efficient but unchosen inequality can be compatible with the value of equality. A major problem with this argument is that, in at least some cases, the hypothetical insurance market cannot be a plausible account of what both equality and utility require.35 Suppose that there is a one-in-a-hundred possibility of acquiring a congenital disease that requires expenditure of £100,000 per annum to secure basic mobility. This expenditure will not be retained or reinvested by society, though it does enable the income of the person with the disease to be increased from £1,000 per annum to £6,000 per annum as they will be able to work part time. Average contributions of £1010 per year from those ninety-nine in a hundred who do not acquire the disability would then be required to provide the disease expenditure. Average pre-contribution income is £10,000 per annum, usually enough for a welfare level of 10 units, and post- contribution income would be £8,990 per annum, which is sufficient only for a welfare level of 9 units. As it happens, the person who develops the disease will have a low (5 unit) level of welfare even if they are mobile, as they will still be very ill, though that is five times their welfare level if they are not mobile. To keep matters simple, I assume that the options come down either to insuring at the level described above, or not insuring at all. Any expenditure below the full £100,000 has no benefit for the person who develops the disease as, say, providing half of the cost of the yearly procedure necessary for 35 Another problem is that the argument is too limited in scope for the present task. It claims to show that equality and utility might be rendered consistent were equality of resources put into practice. But unity of value requires that values never conflict, even under unfavourable economic conditions. See Galston, The Practice of Liberal Pluralism, p. 174. 17 mobility is no better than providing none of it—unless the full cost is met, the procedure will not go ahead. I can of course assume this, as any example of values conflicting disproves the Dworkinian thesis of the unity of value. The question is, would people antecedently accept insurance? If they do, it is arguable that equality is satisfied, as inequality in resources and welfare is reduced. But it seems intuitively (and independently of my earlier description of utility) implausible to say that utility has been satisfied. After all, society has foregone £100,000 of income (divided between 99 of its members) for the sake of £5,000 of income (concentrated on one of its members). It has similarly foregone 99 units of welfare (divided between 99 of its members) for the sake of 4 units of welfare (concentrated in one of its members). Whether or not insurance should be taken out, it is clear that doing so does not maximize the amount of advantage in the society. What, then, if insurance is not taken out? In that case, utility is probably satisfied—welfare and resources have been promoted. But it is then intuitively (and independently of my earlier description of equality) implausible to say that equality has been satisfied. Society could have accepted the insurance increased the worst off’s income from £1,000 per annum to £6,000 per annum, and their welfare level from 5 to 1, both massive improvements to their lives which would also decrease (relative) inequality. The insurance would also have been recommended by one of Dworkin’s favoured measures of equality, the envy test.36 It can be assumed that the person with the disease would envy other people’s bundles of personal (e.g. money) and impersonal (e.g. 36 Dworkin, Sovereign Virtue, ch. 2. 18 marketable talent) resources.37 The envy test therefore recommends that resources are transferred to the person with the disease. But instead of improving the position of the worst off, and reducing inequality, and respecting the envy test, people have chosen to increase the benefits of the already advantaged. Whether or not insurance should be taken out, it is clear that not doing so is contrary to equality on any ordinary understanding. Thus, there is no policy that satisfies both equality and utility. Now Dworkinians might respond that I have, in the foregoing analysis, relied on ‘ordinary’ understandings of equality and utility. The point, they may claim, is that we should revise these to fit with the results of hypothetical insurance markets. For instance, it might be claimed that equality should be measured by applying the envy test only ex ante to the distribution of diseases.38 But no reason to undertake such a revision of our ordinary view of equality is presented by considering hypothetical insurance markets. As I have already mentioned, there is nothing unsettling about equality and utility being in conflict, and that remains the case when we consider hypothetical insurance markets. Nothing feels out of place in my analysis of the choice to insure being the choice between an equal or advantage-maximizing distribution. It is the Dworkinian analysis of this result, which yields the conclusion that we can satisfy both equality and utility, that is problematic. It makes things easier than they can possibly be. Here is one last argument in support of that view. Consider this example: 37 For the distinction see Dworkin, Sovereign Virtue, pp. 322-3. 38 See Ronald Dworkin, “Sovereign Virtue Revisited,” Ethics, 113 (2002), 106-43; “Equality, Luck and Hierarchy.” 19 Insuredland and Uninsuredland. At t1 there are two identical countries facing the insurance decision described above. At t2 Insuredland insures and Uninsuredland does not. Neither of the possible unity of value analyses of Insuredland and Uninsuredland is plausible. It might be claimed that Insuredland fully satisfies the demands of equality and utility.39 But if Insuredland fully satisfied the demands of utility, then Uninsuredland can be no better as regards utility. And it is absurd to claim that it is no better from the perspective of utility to favour £5,000 over £100,000, and 4 units of welfare over 99. Alternatively, it might be claimed that Uninsuredland fully satisfies the demands of equality and utility. But if Uninsuredland fully satisfied the demands of utility, then Insuredland can be no better as regards equality. And it is absurd to claim that it is no better from the perspective of equality to favour £6,000 of income for the worst off to £1,000 of income for the worst off, 5 units of welfare for the worst off to 1 unit of welfare for the worst off, and a distribution rejected by the envy test to one recommended by the envy test. The inescapable conclusion is that Insuredland better realizes equality, and that Uninsuredland better realizes utility. Hypothetical insurance markets cannot salvage the Dworkinian claim that equality and utility do not conflict. II. IS UTILITY REALLY A VALUE? 39 To fully satisfy a value is to realize its ends as far as is presently possible. Unity of value requires that all values can be simultaneously fully satisfied. Merely partially satisfying one value, when there is some alternative that would fully satisfy it at the cost of decreased satisfaction of the other value, would expose the values as conflicting. 20 The above findings regarding equality and utility might be of little concern to Dworkin. His position appears to be that utility is not really a value. Thus Dworkin might favour a version of the second strategy for defending the unity of value that I mentioned in the introduction. Specifically, he might deny that there is a value conflict between equality and utility by saying that, though they conflict, utility is not a value.40 To see how he defends that view, I must first briefly set out his approach to appraising values. In Dworkin’s view, a legitimate government must endorse two fundamental principles: “First, it must show equal concern for every person over whom it claims dominion. Second, it must respect fully the responsibility and right of each person to decide for himself how to make something valuable of his life.”41 He even describes this as an “unchallengeable proposition.”42 I will not challenge the proposition here, so in that regard my critique of Dworkin is internal. Applied to distributive justice, this proposition requires the “simultaneous equation” of “find[ing] a solution that respects both the reigning principles of equal concern and personal responsibility.”43 Even when satisfying equal concern, we must satisfy personal responsibility, and vice versa. On what basis, then, might Dworkin deny that utility is a value? One line of argument considers a form of utilitarianism that is framed in putatively egalitarian 40 The other version of the second strategy is obviously closed to Dworkin, as he is committed to equality as a value; see the discussion below. 41 Dworkin, Justice for Hedgehogs, p. 2. 42 Dworkin, Justice for Hedgehogs, p. 422. 43 Dworkin, Justice for Hedgehogs, p. 3. 21 terms.44 On this view “we treat each person as an equal by valuing his pleasure (or happiness or welfare or success) equally in choosing policies that will increase the aggregate of pleasure (or one of those other commodities) in the community as a whole.”45 According to Dworkin, this view “offers an unpersuasive interpretation of equal concern.”46 Here is his argument for that conclusion in its entirety: Parents would not show equal concern for all their children if they spent their entire available budget educating only those who were likely to earn heavily in the market. That would not treat the success of each child’s life as equally important. Concern for a large group of persons is not the same as concern for its members one by one. Yes, an aggregation strategy values happiness or welfare or some other interpretation of utility, no matter in which person it resides. But that is concern for a commodity, not for a person.47 The analogy with assigning a domestic educational budget seems to entirely miss the mark. Happiness is obviously not equivalent to mere money. It is not typically a means to an end. You would clearly lack equal concern if you invested in your children’s education with the sole goal of maximizing their earning potential—when this might mean that the one potential investment banker gets all the investment and the potential 44 See also Will Kymlicka, Contemporary Political Philosophy, 2nd ed. (Oxford: Oxford University Press, 2002), ch. 2. 45 Dworkin, Justice for Hedgehogs, p. 354. 46 Dworkin, Justice for Hedgehogs, p. 354. 47 Dworkin, Justice for Hedgehogs, p. 354. 22 teachers get nothing. But this says nothing about utilitarians, as they do not endorse this strange objective. The claim that “[c]oncern for a large group of persons is not the same as concern for its members one by one” might be true, but seems beside the point as the utilitarian does not claim only to show concern for the group. She claims also to show concern for each individual by, in Dworkin’s own words, “treat[ing] each person as an equal by valuing his pleasure (or happiness or welfare or success) equally.” It is true that the concern for the group and the concern for the individual are two sides of the same coin for the utilitarian, but that seems also to be the case for Dworkinian egalitarianism: its idea of showing equal concern to the group will just be respect for individual a, plus respect for b, plus respect for c, etc. Dworkin is an individualist, so his objection here cannot be that the group has a moral life of its own that the utilitarian overlooks. He may mean to gesture at a separateness of persons objection to utilitarianism, but that objection is typically directed at what Rawls calls “[t]he most natural way … of arriving at utilitarianism,” which “is to adopt for society as a whole the principle of rational choice for one man.”48 As Dworkin has made no reference to individual rationality in describing utilitarianism, and characterizes it as attempting to treat each person as an equal, it is hard to see how this objection might be relevant. The final part of Dworkin’s objection is more troubling for utilitarians and, by extension, those who endorse utility as a value. They might fairly complain that utility is not really a commodity. But while Dworkin’s choice of words here could have been better, his objection does not rely upon making utility into something that matters only 48 John Rawls, A Theory of Justice, revised ed. (Oxford: Oxford University Press, 1999), pp. 23-4. 23 instrumentally, like corn or oil. The utilitarian claim is essentially that equal concern consists in refusing to discriminate between persons when it comes to allocating utility. But the thing that is being maximized is utility itself, “no matter in which person it resides,” as Dworkin puts it. And at least sometimes, the effects of this maximization, even if it is motivated by a particular idea of what equal concern consists in, seems substantively unequal. Dworkin has mentioned a clear example of this in earlier work. Utilitarians can argue with great plausibility that racial discrimination would usually fail to maximize happiness, but they cannot plausibly claim that utilitarianism displays opposition to discrimination in principle, and in all conceivable cases.49 We might grant that utilitarianism shows concern for persons—after all, utilitarianism counts the happiness of every individual, and that people are happy is a legitimate concern—but there is a case for saying that it does not show equal concern for persons. The same follows, mutatis mutandis, for utility. Consider again the distributions in equality versus utility (2, 2; 1, 5). If the second, unequal distribution is selected, the worse off persons at least have an argument for saying that they have not been shown equal concern. Though no truly canonical definition of equal concern is provided, in other contexts Dworkin treats it in a way that suggests it might be inconsistent with the second distribution. It is described as showing “equal concern for fate,”50 which suggests a focus on a concern with how individuals end up, in contrast to the second, ‘full respect’ principle’s focus on the means taken to get there. In that vein, Dworkin imagines the first principle being invoked in critique of 49 Dworkin, Law’s Empire, pp. 290-1. 50 Dworkin, Justice for Hedgehogs, p. 2, see also p. 352. 24 laissez-faire policies, with a member of the poor pointing out that “[t]here are other, more regulatory and redistributive, sets of laws that would put me in a better position. How can government claim that this system shows equal concern for me?”51 Given the need to also take into account the second principle, the full complaint of “people with little market talent or bad luck” is that laissez-faire policy “does not show equal concern, because a different economic arrangement is available that also satisfies the requirements of individual responsibility and that shows more appropriate concern for them.”52 They might plausibly say the same if the second distribution is favoured in equality versus utility on utility grounds. I think that there is something to the worse off’s complaint here, but it is not the whole story about equal concern. There is a second aspect to equal concern. Consider what those who would be better off in the second distribution might say were the first, equal distribution chosen. In what way does the selection of the first distribution show equal concern for their fate? It gives them two units, when they would have had five under the other available distribution. The justification for this is that it is necessary to give the other group an extra benefit, creating outcome equality. But the size of that benefit is only one unit, compared to the loss of three units of benefit that the first distribution imposes on one group. At least where the worse off group under a distribution containing much more advantage is not badly off in absolute terms, the better off can fairly ask why three units of advantage for them counts for less than one unit of 51 Dworkin, Justice for Hedgehogs, p. 3. 52 Dworkin, Justice for Hedgehogs, p. 363. 25 advantage for the others. They can claim, with some plausibility, that they have been discriminated against when it comes to assigning benefits. These considerations suggest the following claim: The second aspect of equal concern: equal concern is not satisfied where small benefits for some individuals (the worse off) are brought about at the cost of losing big benefits for others (the better off). Equal concern, and even equal concern for fate, are ambiguous enough that they can be interpreted as praising the selection of less advantage more evenly distributed in one regard, as the first aspect of equal concern requires, and as criticising that selection in another regard, as the second aspect requires. The fox can accommodate both aspects of equal concern. His response to the objection that utilitarianism might conceivably have results, such as racial discrimination, which evidently do not show equal concern, is that it is more precise to say that it reflects only one aspect of equal concern. It is for that kind of reason that the best account does not endorse full-blown utilitarianism. My kind of fox, at least, endorses both equality and utility as conflicting values, each reflecting one aspect of equal concern. Though they conflict, in the sense that sometimes one recommends one course of action, the other another course of action, they are complementary in the sense that they act as constraints on each others’ excesses. Our concern with equality will ensure that there is no possibility of racial discrimination, just as our concern with utility ensures that the well- being of the would-be better off is not disregarded. 26 The hedgehog may be appalled by this turn of events. The suggestion that equality and utility are values was bad enough, but now it is being claimed that this is because equal concern—a basis on which values are to be assessed—itself has two sides to it. What counterarguments might Dworkin advance here? A first, and obvious, strategy, given his general hostility to utilitarianism, would be to show that the second aspect of equal concern, which underpins the utility strand of the foxy position outlined above, is false. This approach says that the only valid complaint under the heading of equal concern comes where the worst off (or representatives) say “[t]here are other, more regulatory and redistributive, sets of laws that would put me in a better position. How can government claim that this system shows equal concern for me?” This view of equal concern as just a matter of maximizing the position of the worst off (maximining) is implausible both in general and, especially, as a defence of Dworkin’s position. Even if the reader is still drawn to maximining when considering equality versus utility (2, 2; 1; 5), it seems barely credible when considering ‘super equality versus utility’ cases (for instance, 2, 2; 1.9999, 10). Here maximin would be willing to forego a massive benefit for one group, for the sake of a barely noticeable benefit for another.53 Furthermore, Dworkin himself is committed to rejecting this monistic interpretation of equal concern. This is both in the obvious sense that he does not give maximin an explicit expression in his final principles, as Rawls famously does, 53 Were the ‘benefit’ merely instrumentally valuable, this might not seem unreasonable, as the benefit might then be subject to diminishing marginal returns, with the result that the transfer to the benefit of the worst off looks like better value. I should emphasize, then, that I use ‘benefit’ in a technical sense, as a placeholder for the reader’s favoured currency of justice—something which is presumably not merely instrumentally valuable and therefore not subject to diminishing marginal returns. 27 and the less obvious sense that Dworkin himself invokes a sense of equal concern which is relevantly similar to the second sense described above. In a discussion contrasting Rawls’ theory and his own, Dworkin argues that “a government that allows a much greater loss to fall on one citizen, in order to avert a much smaller loss for a second, would not be treating the former as an equal.”54 Similarly, in developing his own equality of resources, Dworkin asks whether this equality should be achieved ex post—that is, after the natural allocation of talents and disabilities, and the market allocation of resources and opportunities, have taken place. He objects that “the ex post approach, even so far as it is possible, is a very poor understanding of equal concern”.55 This is because “[a]ny community that undertook to spend all it could to improve the position of its blind or crippled members, for example, until further expenditure would not even marginally benefit them, would have nothing left to spend on anything else, and the lives of all other citizens would be miserable in consequence.”56 Clearly, Dworkin thinks that something other than how the worst off fare matters, as he objects to some expenditure that would benefit them. And the nature of his objection is very similar to that suggested by the second aspect of equal concern that I suggested above: a sole concern with the interests of the worst off will lead to disproportionate losses for the better off. His view seems, like mine, to say that equal concern is upset, in one regard, where small benefits for some individuals (the worse off) are brought about in favour of big benefits for others (the better off). 54 Dworkin, Sovereign Virtue, p. 115. 55 Dworkin, Justice for Hedgehogs, p. 358. 56 Dworkin, Justice for Hedgehogs, p. 359. 28 Now it is obviously a further jump from there, and one which I need not make, to claim that there is a significantly utilitarian branch to Dworkin’s thinking here.57 Although Dworkin may accept the general proposition of the second aspect of equal concern, he may say that, within his overall theory, some other aspect of equal concern— the first aspect, or something similarly non-aggregative—has much more weight. On this view, it is only where very disproportionately large losses for the better off follow from benefitting the worst off that the overall theory disapproves. This, and no more, is certainly what is suggested by his rather bluntly stated example of the “blind or crippled.” The utilitarian, by contrast, thinks that the second aspect of equal concern is the only aspect. She always objects when small benefits for some individuals are brought about in favour of big benefits for others, even if the difference in the size of the two sets of benefits is small. But while Dworkin may invoke the second aspect of equal concern more cautiously than the utilitarian, that does not change the fact that he endorses the second aspect to some extent. And as he endorses the second aspect, it is not plausible that he can resist my foxy defence of equality and utility on the basis that this aspect is false. It might be thought that he could resist it by claiming that only his particular restricted form of the second aspect of equal concern is plausible, but this approach faces two problems. First, Dworkin presents no arguments for it, and there is no obvious way of developing it. Second, as I do not propose utilitarianism, but only an equality-utility hybrid, I need myself only endorse the second aspect of equal concern in a restricted form. Having 57 For Dworkin’s rebuttals of a similar suggestion, see Sovereign Virtue, pp. 110, 116, 349-50. 29 granted that the bigness of benefits sometimes outweighs their distribution, Dworkin grants too much for the first counterargument that we have been considering. I move on, then, to consider a second counterargument that Dworkin might—and in this case, actually does—use to resist the idea that utility is a value as equality is. In earlier work he writes that a utilitarianism justified on grounds of equal concern “cannot coherently offer to compromise highest average welfare over a given population in the name of wealth equality or simple welfare equality.”58 This is on the basis that “[t]reating people as equals, on the utilitarian’s conception of equality, demands the highest aggregate welfare, and one cannot coherently treat people other than as equals in the name of some deeper conception of equality.”59 The wording of the final clause here is a little tricky. We should just set aside the notion that the non-utilitarian interpretation of equal concern is somehow ‘deeper’, as that is not part of the view I am defending—I treat equality and utility, and their respective supporting aspects of equal concern, as equally fundamental. I believe that the core of Dworkin’s objection is that the ‘partial utilitarian’ is, qua utilitarian, committed to maximizing welfare on equal concern grounds, but also, qua non-utilitarian, committed to not maximizing welfare on equal concern grounds, which he thinks is incoherent. But that is only an incoherent thought as long as we treat the two conflicting thoughts as absolute imperatives. If we instead say just that we have a pro tanto reason of equal concern to act in one way, and another pro tanto reason of equal concern to act in another, there is no incoherence: we just acknowledge that some non-conclusive but 58 Ronald Dworkin, A Matter of Principle (Cambridge, MA: Harvard University Press, 1985), p. 274. 59 Dworkin, A Matter of Principle, p. 274. 30 relevant considerations support one action and others support another, and try to balance them against each other to come up with a practical solution. To reiterate, that such a process may be difficult is, both logically and by Dworkin’s own lights, no evidence that there is indeterminacy about what it is best to do. Dworkin might reply that the incoherence here is not a general problem of conflict, but specifically to do with the fact that this is an alleged conflict within equal concern, which is impossible. But such a reply would both assume what needs to be shown (that conflict within values is impossible) and be inconsistent with the dualism of Dworkin’s own interpretation of equal concern. This brings us on to a third counterargument. Dworkin claims that “much energy has been wasted by the defeating assumption that liberty, equality and democracy are criterial concepts that can be explicated through some neutral analysis that makes no assumption about their value or importance.”60 It is this conception of equality that “encourages the dismissive view that equality means flat equality—everyone having the same wealth throughout their lives—because no other definition is plausible if we take equality to be criterial.”61 Now clearly I have been assuming a fairly strict conception of equality. Although I have allowed for the possibility of departures from equality on account of responsibility considerations, I have been committed to the notion that equality would not allow inequality for the sake of larger benefits for the better off. So Dworkin might claim that it is only by making equality a very uncompromising doctrine, which does not value benefits unless they fall to the worse off, that I have been able to 60 Dworkin, Justice for Hedgehogs, p. 345. 61 Dworkin, Justice for Hedgehogs, p. 346. 31 show that equality needs a confederate such as utility. His ‘interpretative’ notion of equality, by contrast, builds in the full set of equal concern considerations. I can grant Dworkin this particular argument. If it succeeds, it only establishes Dworkin’s broader notion of equality as an alternative to an equality-utility combination. And establishing an alternative to a theory is not itself an argument against that theory. How might Dworkin provide the further argument to show that his alternative is better? One way is by appealing to Occam’s razor. Dworkin has one value, and I have two, so it might look like we should break any tie in favour of his theory, as it is simpler. But this is far from the truth. Though I have two values, they each have a relatively simple internal structure, while Dworkin’s is much more complex. Not only does he incorporate the same two kinds of considerations—corresponding to the two aspects of equal concern—that my overall theory appeals to in his one value, but his full theory has such complexities as hypothetical insurance markets, auctions, and the envy test, which do not have analogues in my theory.62 Although I doubt that it is a weighty consideration, parsimony, if anything, tells in favour of a combination of equality and utility. Another approach is to appeal to the fact that, on Dworkin’s view, equality is a value that should always be promoted, while in my view, equality’s prescriptions, though they are taken into account, will sometimes be overridden in practice. (This is certainly so in the ‘super equality versus utility’ cases that I mentioned earlier.) As equality is one of the great values of contemporary philosophy, and more importantly of modern societies, it might be thought a significant advantage of Dworkin’s theory that its allegiance to it is unconditional. 62 See especially Dworkin, Sovereign Virtue, ch. 2. 32 But it is unclear why it would be a good thing that some notion that has wide acceptance can be rendered so that it never admits of exceptions. For the theorist, the really important questions when deciding which values to support, and how much support they should have, is how far true they are, and how much good they do. In trying to show that equality is worthy of unconditional support, a liberal such as Dworkin cannot plausibly cite the status it happens to have in support of its truth and goodness. I cannot, then, detect any good argument for Dworkin’s view to be preferable to my position on account of the former’s apparent denial of utility as a value. III. PROCRUSTES I have defended the fox’s thesis that true values may conflict, with special reference to distributive justice. My specific claims are (1) that equality and utility may conflict with each other, and (2) that they both genuinely are values. The first part of the article considered the possibility of denying (1), mostly using the materials Dworkin provides when attempting to resolve another value conflict. The second part of the article considered Dworkin’s apparent denial of (2), specifically taking the form of the claim that utility is not a value. I found that the resistance to both of my claims can be overcome, and so the fox escapes unscathed. It is worth emphasizing that Dworkin defends a strong version of value holism. Weaker versions of hedgehoggery can accept my central claims. One such view accepts that there might be conflicts between values, but holds that these are resolved at a higher level of abstraction. For instance, a consequentialist might concede that the likes of honesty, kindness, and equality are genuine values that can conflict, while claiming that 33 there is a master value of Moorean good that conclusively resolves these conflicts.63 This view accepts my claims (1) and (2). The fox’s arguments against such a view would have to take a different form to those advanced here. Though Dworkin’s denial of value conflicts is a strong one, it retains some intuitive plausibility by retaining considerations typically associated with rival values. Most relevantly, while my view accommodates the first and second aspects of equal concern under the banners of equality and utility, Dworkin encompasses them under one expanded banner of equality. I will conclude by providing a brief sketch of a further argument that suggests that this position lacks a key advantage held by the fox.64 When we are confronted with a decision such as that of equality versus utility, we should be able to feel that we are pulled in two directions. There is something good about people on the whole faring better, and there is also something good about distributions being more equal. To deny that is to fail to recognize something that matters fundamentally. Dworkin acknowledges as much by appealing to both aspects of equal concern. So when we can have more advantage in our distribution, or more equality, but not both, we should feel at least a little conflicted, and we should be able to specify the source of that as a conflict between things that matter to us. The fox can have all of this: he may recognize equality and utility as values, he may feel some disappointment 63 See Donald Regan, “Value, Comparability, and Choice,” in Chang (ed), Incommensurability, Incomparability, and Practical Reason. 64 A similar argument is powerfully presented in Bernard Williams, “Liberalism and Loss,” in Mark Lilla, Ronald Dworkin, and Robert Silvers (eds), The Legacy of Isaiah Berlin (New York: New York Review of Books, 2001), pp. 100-2. 34 (though as I said, not blame) when he cannot have much of both, and he may know that this is because, on this occasion, his values pull in opposite directions. It is less clear how the hedgehog can tell a similar story. He does, of course, think that there is something good about distributions being more equal, and he also thinks there is, in some cases at least, something good about people on the whole faring better. But any explanation of what is happening in a situation where we can have one of these things, but not the other, must be strange. There is not a conflict between two values here, so what is there? Minow and Singer suggest that “[w]hat is at risk is sheer honesty about competing views and contrasting tugs on our perceptions and hearts.”65 This may slightly overstate the point. There can be a persisting conflict on the hedgehog’s view. But it must be within the value of equality. That is problematic in the first place because Dworkin seems to claim that such a conflict is incoherent. But even granting that there is such a conflict, the framework within which Dworkin must place it serves only to muddy the waters of moral discourse. For instance, if he wanted to give a principled explanation to congenitally disabled people of why it is that he is not going to give them a small benefit, because it comes at great cost to the rest of the community, what would he say? Dworkin might prefer to provide an explanation in terms of mechanisms such as the hypothetical insurance market, but these are not plausibly themselves values: they are just ways of thinking about the implications of values. On Dworkin’s view, such an explanation boils down to this: we are selecting a more unequal distribution rather than a more equal one 65 Martha Minow and Joseph William Singer, “In Favor of Foxes: Pluralism as Fact and Aid to the Pursuit of Justice,” Boston University Law Review, 90 (2010), 903-920, p. 907. 35 for the sake of equality.66 This adds insult to injury—the involuntarily worst off are not only denied resources, but in the name of the value on which they thought they could rely. Imagine a disabled person trying to formulate a response. Were the policy enacted by a fox, she could have said “you hold that equality has the weighting x and utility has the weighting y, but I propose that the justified weighting of utility is only z.” This is a clear and, as Dworkin says, familiar picture. But to the hedgehog, she has to say something like “you hold that equality type 1 has the weighting x and equality type 2 has the weighting y, but I propose that the justified weighting of equality type 2 is only z.” We should not rule out such revisionary treatment of concepts, but for them to be attractive we must first be shown that something was wrong with the more familiar conflict between equality and utility. We have not been shown that. These points are reinforced by revisiting Insuredland and Uninsuredland. On Dworkin’s view (as opposed to the Dworkinian view I considered earlier in relation to this case, which recognized utility as a value), one or other of Insuredland and Uninsuredland fully satisfies equality, and utility does not come into the picture. Suppose that it is Insuredland that satisfies Dworkin’s equality. In that case, a member of society who does not develop the disease, and who opposes the insurance premiums, cannot claim, if she is to address questions of value, that utility has been unduly overlooked. She must rather claim that equality type 2 has been unduly overlooked. But this seems 66 Were the disabilities in the example the result of choice, Dworkin might have been able to appeal to the second fundamental principle—that of full respect for responsibility. But it is specified that the disabilities are congenital and as such involuntary. 36 misleading as, intuitively, her claim does not seem to concern equality at all. She wants those who are better off already to be even better off. It seems much more intuitively sound to describe her objection as one from utility. If the person with the disease responds to her that the premiums should continue to be paid as equality should take precedence, that seems like a coherent response, not a category mistake as Dworkin has it. Dworkin anticipates the worry “that [y]ou might think me Procrustes, stretching and lopping conceptions of the great political virtues so that they neatly fit each other. I would then be achieving unity on the cheap: a meaningless victory.”67 He assures us that he “mean[s] to submit each of the political conceptions I describe to the test of conviction. I will not rely on any assumption that a theory is sound just because it fits with other theories we also find agreeable.”68 This assurance notwithstanding, I suggest that, as regards equality, Dworkin’s thought has this procrustean quality, which is attributable to his apparent denial of utility’s value. My combination of equality and utility may, then, be superior in one regard, and as we saw earlier, inferior in none. ACKNOWLEDGMENTS An earlier version of this article was discussed at a meeting of the theory group at the Department of Politics, University of Glasgow. I would like to thank the participants on that occasion, as well as the editors of Law and Philosophy and the anonymous referees, 67 Dworkin, Justice for Hedgehogs, p. 5. 68 Dworkin, Justice for Hedgehogs, p. 5. 37 who provided valuable comments. I am also grateful to the British Academy, who supported research for this article. Politics University of Glasgow, Glasgow, United Kingdom E-mail: carl.knight@glasgow.ac.uk Politics University of Johannesburg, Auckland Park, South Africa