Hostname: page-component-7c8c6479df-xxrs7 Total loading time: 0 Render date: 2024-03-28T17:01:58.669Z Has data issue: false hasContentIssue false

THE INTERPRETATION OF MAXIMIZING UTILITARIANISM

Published online by Cambridge University Press:  24 November 2008

Jonathan Riley
Affiliation:
Philosophy and Political Economy, Tulane University

Abstract

Utilitarians and their critics commonly assume that maximizing utilitarianism necessarily aggregates over cardinal comparable personal utility rankings that are homogeneous in quality independently of their sources or objects, whether utility is conceived in terms of pleasure or preference satisfaction. Although familiar versions of utilitarianism, crude or sophisticated, do make such rich homogeneous utility information part of the very meaning of the doctrine, utilitarian philosophy loses credibility as a result. A more credible version of maximizing utilitarianism along John Stuart Mill's lines relies instead on plural kinds of ordinal non-comparable personal utilities that are heterogeneous in quality. One kind of utility is qualitatively superior to another if and only if the higher kind is more valuable than the lower kind irrespective of quantity. Given that the kind of utility associated with the moral sentiment of justice is qualitatively superior to any competing kinds, Mill's pluralistic ordinal utilitarianism gives absolute priority to a social code of equal rights and duties over competing considerations. Some key practical implications of this extraordinary utilitarianism are discussed with reference to various examples of the sort discussed in the literature.

Type
Research Article
Copyright
Copyright © Social Philosophy and Policy Foundation 2008

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

1 Mill, John Stuart, Utilitarianism, in Robson, John M., ed., Collected Works of John Stuart Mill (hereafter CW), 33 vols. (London and Toronto: Routledge and University of Toronto Press, 1963–1991), X, 257Google Scholar.

2 Ibid., 258.

3 Mill's condition that “proper allowance” must be made for different kinds of utilities is typically ignored by utilitarians and their critics because of a received opinion that such qualitative distinctions are logically incompatible with utilitarianism. I shall argue later in the essay that this received view is mistaken. Nonetheless, it is worth noting at this point that the received opinion is not restricted to hedonistic conceptions of utility. Its status as conventional wisdom may be due in large part to its immediate acceptance by influential utilitarian contemporaries of Mill's such as Alexander Bain, Henry Sidgwick, W. Stanley Jevons, and Francis Y. Edgeworth, who (like Mill) also happened to be ethical hedonists. But the view that Millian qualitative distinctions are illegitimate has persisted even as hedonistic conceptions of utility have been replaced by nonhedonistic conceptions in recent sophisticated versions of utilitarianism such as those proposed by Richard Hare and John Harsanyi. Anti-utilitarians have not had much incentive to take Mill's side or challenge a view that has always been endorsed by leading utilitarian philosophers as well as by neoclassical economists.

4 For a given population, maximizing the sum total is equivalent to maximizing the average pleasure per capita. But these two criteria diverge and neither is very appealing by itself if population is allowed to vary. The problems that arise under variable population must be faced by any plausible version of consequentialism, but I shall not address them here. For further discussion, see Blackorby, Charles, Bossert, Walter, and Donaldson, David, Population Issues in Social Choice Theory, Welfare Economics, and Ethics (Cambridge: Cambridge University Press, 2005)CrossRefGoogle Scholar.

5 Hare, Richard M., Moral Thinking: Its Levels, Method, and Point (Oxford: Clarendon Press, 1981)CrossRefGoogle Scholar; , Hare, Sorting Out Ethics (Oxford: Clarendon Press, 1997)Google Scholar; and , Hare, Objective Prescriptions and Other Essays (Oxford: Clarendon Press, 1999)Google Scholar.

6 Dworkin, Ronald, Taking Rights Seriously (London: Duckworth, 1977), 234Google Scholar.

7 Dworkin rejects utilitarianism as unfair because he thinks that the self-oriented and external elements in any person's preference are sometimes “inextricably tied together” (ibid., 236). Prejudiced people may hold very intense preferences to avoid blacks or homosexuals, for example. In that case, “the personal preferences upon which a utilitarian argument must fix will be saturated with that prejudice,” because it is impossible to isolate and ignore the underlying external preferences (ibid., 237). Harsanyi admits that the self-oriented and external elements may be inseparable, but he nevertheless suggests that utilitarianism can be suitably reconstituted so as to ignore unfair choices. His view seems to be that some personal choices can be classified as morally impermissible on the basis of their content and that a rational utilitarianism can be consistently required not to count such impermissible choices, including choices that exhibit undue bias for or against others' interests. See Harsanyi, John C., “Game and Decision Theoretic Models in Ethics,” in Aumann, Robert J. and Hart, Sergiu, eds., Handbook of Game Theory, vol. 1 (Amsterdam: North Holland, 1992), 703–6Google Scholar. Hare is reluctant to ignore external preferences as he interprets them, although he struggles with the issue. See, for instance, Hare, Objective Prescriptions, 126–31, 158.

8 Harsanyi, John C., “Morality and the Theory of Rational Behavior,” Social Research 44 (1977): 623–56Google Scholar; reprinted in Sen, Amartya K. and Williams, Bernard A. O., eds., Utilitarianism and Beyond (Cambridge: Cambridge University Press, 1982), 3962CrossRefGoogle Scholar. See also Harsanyi, “Game and Decision Theoretic Models in Ethics,” 669–707. Note that Hare's universalizability criterion is a more demanding condition for moral choice than Harsanyi's equiprobability criterion, because Hare assumes no veil of ignorance.

9 Hare, Moral Thinking, 117; emphasis in the original.

10 Ibid., 93.

11 See, e.g., Harsanyi, , “Normative Validity and Meaning of Von Neumann and Morgenstern Utilities,” in Skyrms, Brian, ed., Studies in Logic and the Foundations of Game Theory (Dordrecht: Kluwer, 1992)Google Scholar. For further discussion of Harsanyi's rational choice version of utilitarianism with examples of how the method of cardinalization works, see Binmore, Ken, Playing Fair (Cambridge, MA: MIT Press, 1994), 259315Google Scholar.

12 Hare, Moral Thinking, 46.

13 Sidgwick, Henry's The Methods of Ethics (London: Macmillan, 1874)Google Scholar came out when the so-called marginalist revolution was beginning to transform classical economic theory into its neoclassical successor. Among other things, the revolution introduced the use of mathematical calculus to provide more precise models of hedonistic utilitarianism, which was the dominant ethical doctrine within normative economics, as Sidgwick urged. Perhaps the most influential work in this regard was Edgeworth, Francis Y.'s Mathematical Psychics: An Essay on the Application of Mathematics to the Moral Sciences (London: Kegan Paul, 1881)Google Scholar, which makes the assumption that the analyst can speak meaningfully of interpersonally comparable natural units of pleasure. For further discussion of the evolving interpretations of utilitarianism within economic theory, see Riley, Jonathan, “Utilitarianism and Economic Theory,” in Durlauf, Steven and Blume, Lawrence, eds., New Palgrave Dictionary of Economics, 2d ed., 4 vols. (London: Palgrave Macmillan, 2008)Google Scholar.

14 Modern economists have proposed a more flexible, ordinal version of utilitarianism that abandons cardinal measure but retains interpersonal comparisons of utility based on the extended sympathy method. Utility represents preference orderings revealed by consistent choice behavior, and it is assumed that different persons' levels of preference-satisfaction can be meaningfully compared through sympathy. No commitment is made to hedonism. See, e.g., Arrow, Kenneth J., “Some Ordinalist-Utilitarian Notes on Rawls's Theory of Justice,” Journal of Philosophy 70 (1973): 245–63Google Scholar; Arrow, , “Extended Sympathy and the Possibility of Social Choice,” American Economic Review 67 (1977): 219–25Google Scholar; and Gibbard, Allan F., “Ordinal Utilitarianism,” in Feiwel, George R., ed., Arrow and the Foundations of the Theory of Economic Policy (London: Macmillan, 1987), 135–53CrossRefGoogle Scholar. But this ordinal variant is seen by some as a nonstandard version of the doctrine, which is perhaps better classified as something other than utilitarianism. See Sen, Amartya K., “Utilitarianism and Welfarism,” Journal of Philosophy 76 (1979): 463–89CrossRefGoogle Scholar.

15 For evidence of Bentham's doubts about cardinal measurement and interpersonal comparison of utility in the sense of pleasure including freedom from pain, see Mitchell, Wesley C., “Bentham's Felicific Calculus,” Political Science Quarterly 33 (1918): 161–83CrossRefGoogle Scholar.

16 Mill, Utilitarianism, 211; emphasis in the original.

18 Moore, G. E., Principia Ethica (Cambridge: Cambridge University Press, 1959)Google Scholar. Unlike Sidgwick and the economists, Moore was a sharp critic of ethical hedonism, although he seems to have remained a utilitarian for whom utility or good is “one definite thing, absolutely indefinable, some one thing that is the same in all the various degrees and in all the various kinds of it that there may be” (ibid., 13; emphasis added).

19 Ibid., 78.

20 See, e.g., Blackburn, Simon, Being Good (Oxford: Oxford University Press, 2001), 82Google Scholar. John Dewey is another prominent figure in this tradition of criticizing Millian qualitative superiorities as incoherent. Indeed, like Moore and transcendental idealists such as Kant, Hegel, Thomas Hill Green, and Bernard Bosanquet, he rejects hedonism altogether. Nevertheless, he seems to endorse a nonstandard pragmatist version of rational preference utilitarianism. In an ideal Deweyan society, rational individuals would habitually resolve practical problems by cooperating in terms of a shared but evolving idea of the general welfare in which great importance is assigned to a code of justice that distributes equal rights and correlative duties. By cooperating successfully on these terms, reflective individuals would jointly satisfy their desires (“ends-in-view”) in harmony as they resolved an indefinite series of problematical situations arising over time in their natural environment. Among his many works, see Dewey, John and Tufts, James Hayden, Ethics, in Boydston, Jo Ann, ed., John Dewey: The Later Works: 1925–1953, vol. 7 (Carbondale: Southern Illinois University Press, 1985), 191–99, 240–52Google Scholar.

21 Sen, Amartya K., “Plural Utility,” Proceedings of the Aristotelian Society 81 (1980): 193215CrossRefGoogle Scholar.

22 Edwards, Rem B., Pleasures and Pains: A Theory of Qualitative Hedonism (Ithaca, NY: Cornell University Press, 1979)Google Scholar. As will become obvious, I do not entirely agree with his view of the matter. But his discussion provides an excellent corrective to the conventional prejudice that utility (in this case, hedonistic utility) must be seen as homogeneous in quality independently of its sources or objects.

23 The term ‘infinite’ does not connote a real number or quantity. Rather, it properly refers to a hypothetical procedure of endless or unlimited expansion. True, the procedure might converge on a finite limit, but the procedure itself is endless and, strictly speaking, never actually reaches the limit. This is not incompatible with the invention of ideal objects such as ‘infinite sets’ for mathematical purposes, keeping in mind that nobody could ever observe the complete contents of an infinite set and, indeed, nobody can even conceive what such a completed set would look like in practice. For relevant discussion, see John Stuart Mill, An Examination of Sir William Hamilton's Philosophy, in Mill, CW, IX, 37–49, 82–86, 425–29.

24 See, e.g., Norcross, Alastair, “Comparing Harms: Headaches and Human Lives,” Philosophy and Public Affairs 26 (1997): 135–67CrossRefGoogle Scholar.

25 See, among many others, Foot, Philippa, “The Problem of Abortion and the Doctrine of Double Effect,” Oxford Review 5 (1967): 515Google Scholar; Williams, Bernard, “A Critique of Utilitarianism,” in Smart, J. J. C. and Williams, Bernard, Utilitarianism: For and Against (Cambridge: Cambridge University Press, 1973), 77150Google Scholar; Thomson, Judith Jarvis, “Killing, Letting Die, and the Trolley Problem,” The Monist 59 (1976): 204–17CrossRefGoogle ScholarPubMed; and Kamm, Frances, Intricate Ethics (Oxford: Oxford University Press, 2007)CrossRefGoogle Scholar. Mill discusses the case of Hannibal in a letter to William Thomas Thornton dated April 17, 1863, in Mill, CW, XV, 853–54. See also Mill's letter to E. W. Young dated November 10, 1867, in CW, XVI, 1327–28.

26 Scheffler, Samuel, Human Morality (Oxford: Oxford University Press, 1992), 103Google Scholar. See also Vallentyne, Peter, “Against Maximizing Act-Consequentialism,” in Dreier, James, ed., Contemporary Debates in Moral Theory (Oxford: Basil Blackwell, 2005), 2137Google Scholar.

27 Scheffler apparently does not identify morality with deservingness of punishment for failure to do what we think is right. But I shall adopt Mill's view in Utilitarianism: “We do not call anything wrong, unless we mean to imply that a person ought to be punished in some way or other for doing it; if not by law, by the opinion of his fellow creatures; if not by opinion, by the reproaches of his own conscience” (184).

28 Scheffler notes that his notion of overridingness is stronger than Hare's linguistic thesis that an individual who competently uses moral language to make sincere moral judgments treats moral principles as overriding (Scheffler, Human Morality, 56 n. 4). Hare allows that a rational agent may know how to use moral language yet opt out of morality, whereas the claim of overridingness as Scheffler understands it implies that a rational agent must, in his own interest, give absolute priority to morality over competing considerations. Although Scheffler largely ignores the point, a society may take various steps to promote harmony between prudence and morality by threatening to duly punish by law or opinion any individual who fails to meet his moral obligations. Society need not take for granted that all of its members are rational and impartial agents who are always willing to do what is right even in the absence of external incentives.

29 Scheffler himself defends a conception of morality as pervasive but moderate and perhaps not overriding. I shall come back to his view of morality briefly in Section VII.

30 Critics of utilitarianism recognize that the utilitarian account of right action can be distinguished from the utilitarian account of optimal deliberation, but this does not lead them to endorse utilitarian ethics. See, e.g., Scheffler, Human Morality, 38–39. For criticisms of both Hare's and Harsanyi's versions of utilitarianism, see Sen, Amartya, Collective Choice and Social Welfare (San Francisco, CA: Freeman, 1970), 131–46Google Scholar; Williams, Bernard, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985), 71119Google Scholar; and Seanor, D. and Fotion, N., eds., Hare and Critics (Oxford: Oxford University Press, 1988)Google Scholar.

31 Hare, Moral Thinking, 42–43.

32 Ibid., 145.

33 Hare's claim that Kant might be read as a sophisticated utilitarian is no doubt controversial. Cf. Hare, Sorting Out Ethics, 147–65; and Williams, Ethics and the Limits of Philosophy, 54–70, 82–92, 189–96.

34 Hare, Moral Thinking, 122.

36 Ibid., 49. See also Hare, Objective Prescriptions, 161–63.

37 Hannibal apparently saved his fellow citizens from having to make a choice by fleeing into exile in Asia Minor.

38 Hare, Moral Thinking, 198–205.

39 Ibid., 169–82.

40 Williams, Ethics and the Limits of Philosophy, 86–87; emphasis in the original. As I have already indicated, Hare is reluctant to deal with this problem more directly by endorsing Harsanyi's move of excluding “external” preferences in Dworkin's sense. Utilitarians do seem unable to consistently make such a move, given that pleasures or satisfactions are homogeneous in quality.

41 Harsanyi presents his version of rule-utilitarianism as a two-stage game in which participants mutually cooperate to choose an optimal code at the first stage and then compete with each other by playing their permissible strategies to maximize their own advantage at the second stage. See Harsanyi, “Game and Decision Theoretic Models in Ethics,” 692–94.

42 For further discussion of this point, see Riley, Jonathan, “Rule Utilitarianism and Liberal Priorities,” in Fleurbaey, M., Salles, M., and Weymark, J., eds., Justice, Political Liberalism, and Utilitarianism (Cambridge: Cambridge University Press, 2008)Google Scholar.

43 Diamond, Peter A., “Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility: Comment,” Journal of Political Economy 75 (1967): 765–66CrossRefGoogle Scholar.

44 Binmore, Ken, Natural Justice (Oxford: Oxford University Press, 2005)CrossRefGoogle Scholar. In Binmore's view, cultural evolution supplies standards of interpersonal comparison which participants in that culture typically believe are reasonable. For a critical discussion of Binmore's contractarian theory of justice, see Riley, Jonathan, “Genes, Memes, and Justice,” Analyse & Kritik 28 (2006): 3256CrossRefGoogle Scholar.

45 The considered appeal of various common moral intuitions has led many moral thinkers, including those who assess acts and omissions in terms of their utility consequences for other persons or sentient creatures, to reject even sophisticated versions of maximizing utilitarianism. Brad Hooker, for instance, defends a version of rule-consequentialism that admits some reductions in the sum total when needed to achieve a fair distribution of utilities: extra weight is given to the utilities of the worst-off members of society so that fairness may be attained. In his view, such a rule-consequentialism provides a systematic impartial justification for shared moral convictions that have their own independent credibility. See Hooker, Brad, Ideal Code, Real World (Oxford: Clarendon Press, 2000)Google Scholar. One serious worry, however, among others, is that Hooker's consequentialism continues to rely, in principle, on rich cardinal and interpersonally comparable utility information.

46 Mill, Utilitarianism, 214; emphasis added.

47 In his discussion of the “desire to be in unity” that he considers “a natural basis of sentiment for utilitarian morality,” Mill admits that even in fairly advanced societies “a person cannot indeed feel that entireness of sympathy with all others, which would make any real discordance in the general direction of their conduct in life impossible” (Utilitarianism, 233). He says that “differences of opinion and of mental culture [may] make it impossible for him to share many of their actual feelings—perhaps make him denounce and defy those feelings” (ibid., emphasis added). Even so, he suggests that at least some individuals have some sympathy for “promoting” other people's “own good,” that is, their personal happiness correctly perceived in point of quantity and quality, as opposed to what they may actually desire and take pleasure in. The implication seems to be that sympathy for others' actual feelings is not necessary for an individual to develop the desire for social unity. A fortiori, utilitarianism, as Mill conceives it, does not depend on interpersonal comparisons based upon sympathy.

48 Mill, Utilitarianism, 213.

49 On the link between utilitarianism and majority rule in the context of interpersonally noncomparable ordinal utility information, see Riley, Jonathan, “Utilitarian Ethics and Democratic Government,” Ethics 100 (1990): 335–48CrossRefGoogle Scholar. Majority rule is subject to inconsistencies, unfortunately, as Arrow's impossibility theorem implies. See Arrow, Kenneth J., Social Choice and Individual Values, 2d ed. (New Haven, CT: Yale University Press, 1963)Google Scholar. One way to remove the inconsistencies is to employ a positional rule or scoring function such as Borda count. Under Borda count, each person ranks the m feasible outcomes from best to worst, and m − 1, m − 2,…,1,0 points are assigned to the best, next-best, …, second-worst, and worst outcomes, respectively. The procedure then selects an outcome with the greatest aggregate point total. More generally, to escape from the Borda method's insistence on equal spacing of points awarded for first choice, second choice, and so forth, a generalized scoring rule could be employed whose point assignments are unique only up to a positive monotonic transformation so long as such transformations are applied to every person's ordering. Although it has the formal appearance of an interpersonally comparable ordinal utility function, this generalized scoring function is really just a way to consistently count one person's preference ordering for exactly as much as another's. There is no attempt to justify the scoring function in terms of interpersonal comparisons of actual pleasures or satisfactions. It does not reflect the results of the extended-sympathy method applied to different people's enjoyments as estimated in their respective preference orderings. Despite its possible attractions as a way of achieving the utilitarian aim of impartially counting everyone for one and only one, however, this scoring approach does occasionally fail to choose Condorcet winners when they exist; that is, it may fail to select an outcome that is majority-preferred to every other outcome in a series of binary contests. In any case, I shall henceforth assume that such a generalized scoring procedure is used to remove any inconsistencies otherwise associated with majority voting. (For recent critical surveys of the vast literature relating to scoring rules, including Borda count, see Brams, Steven J. and Fishburn, Peter C., “Voting Procedures,” in Arrow, K. J., Sen, A. K., and Suzumura, K., eds., Handbook of Social Choice and Welfare, vol. 1 [Amsterdam: Elsevier, 2002], 173236CrossRefGoogle Scholar; and Prasanta K. Pattanaik, “Positional Rules of Collective Decision-Making,” in ibid., 361–94.) Needless to say, Bentham and Mill never discuss such a scoring procedure, but it is arguably in the spirit of utilitarianism as they conceive it. For further discussion, see Jonathan Riley, “Classical Ordinal Utilitarianism,” unpublished working paper.

50 Mill, On Liberty, in CW, XVIII, 252–57. See also Riley, Jonathan, Mill on Liberty (London: Routledge, 1998), 6768Google Scholar.

51 Bentham's awkward phrase “the greatest happiness of the greatest number” makes more sense if he is interpreted as referring to possibly distinct notions of personal happiness held by the majority rather than to some independent notion of collective happiness based on rich utility information.

52 Mill, Utilitarianism, 213.

53 On the use of majoritarian voting procedures as epistemic devices for providing maximum-likelihood estimates of some objective good, such as aggregate net pleasure, see Young, H. Peyton, “Condorcet's Theory of Voting,” American Political Science Review 82 (1988): 1231–44CrossRefGoogle Scholar.

54 Mill, Utilitarianism, 211.

55 For an extended discussion of this interpretation of Millian qualitative superiorities as infinite superiorities, see Riley, Jonathan, “Millian Qualitative Superiorities and Utilitarianism, Parts 1 and 2,” Utilitas 20 (2008)CrossRefGoogle Scholar. It is important to keep in mind that infinity is not a real number or quantity: there is no such thing as a completed infinity in nature. Thus, it cannot be maintained that a higher pleasure is equal in value to an actual infinite amount of lower pleasure.

56 Mill, Utilitarianism, 213; emphasis in the original.

57 Ibid., 211.

58 This sketch of Mill's higher pleasures doctrine requires further clarification, but I cannot provide it here. For further discussion of the different kinds of pleasures and of why it may be reasonable to regard the higher kinds as infinitely superior in value to the lower kinds, see Jonathan Riley, “Millian Qualitative Superiorities and Rational Agency,” unpublished working paper.

59 In his excellent discussion of pluralistic or qualitative hedonism, Edwards suggests that Millian qualitative superiorities might be captured by a simple ordinal ranking of the different kinds or qualities of pleasant feelings (Edwards, Pleasures and Pains, 68–72, 111–19). Like Edwards, Wilson, Fred, in Psychological Analysis and the Philosophy of J. S. Mill (Toronto: University of Toronto Press, 1990)Google Scholar, also interprets Mill as groping for an ordinal utility scale that orders different qualities of utility from higher to lower without trying to quantify how much a higher quality differs from a lower one (ibid., 220–23, 253, 275–93).

60 As a technical matter, the fact that a lexical preference ordering cannot be represented by a continuous real-valued mathematical function means that a utility function, in the purely formal sense of ‘utility’ used by modern economists, cannot be used to represent the information contained in the lexical preference. Thus, an unbridgeable gap is created in this case between the economists' purely formal sense of ‘utility’ and philosophical conceptions of utility as preference-satisfaction, whatever the motivations or reasons assumed to underlie the preferences. In particular, an ordinal utility function might be used to represent the information about the estimated quantities of pleasure to be expected from outcomes contained in an ordinary hedonistic preference ordering restricted to a given kind or quality of pleasure. But a utility function cannot be used to represent the information about the different qualities of pleasures to be expected from outcomes contained in the lexical preference ordering.

61 Given that higher intellectual, moral, and aesthetic capacities must be developed to appreciate higher pleasures such as that associated with the moral sentiment of justice, the majority competently acquainted with a higher kind of pleasure is not necessarily the same majority as the majority competently acquainted with a lower kind of pleasure. A fortiori, the majority competent to decide on a qualitative preference ordering may be an educated minority in some social contexts.

62 Mill, Utilitarianism, 250; emphasis added.

63 Ibid., 251.

64 Ibid.; emphasis added.

65 It should be recalled that a scoring function may be used to remove inconsistencies in majority rule. (See note 49 above.) Strictly speaking, majority rule per se does not give equal weights to different persons' preferences. Rather, it impartially counts the different noncomparable preference orderings without attempting to weigh them relative to one another, and it reflects the shared preferences of the greater number. A scoring function can be superimposed to mimic the way in which ballots are distributed to voters, so that one person's preferences may be counted for exactly as much as another's in the stronger sense that each gets equal weight in the political decision. But the ballot typically covers only the voter's top-ranked alternative instead of his entire ordering.

66 Mill was a firm defender of popular government with a universal franchise, although he was also far more inclined than either his father James Mill or Bentham to institute constitutional checks and balances against unrestrained majoritarianism. His notorious device of plural votes based on education should be regarded as a check against abuses by ignorant or tyrannical majorities. But the device as he describes it would not permit the educated minority to elect so many representatives that they could pass legislation opposed by the representatives of the popular majority, keeping in mind that he also advocates a system of proportional representation. In the special case of an ideal society of highly educated citizens, there would be no need for plural votes. For further discussion of his argument that a form of constitutional representative democracy is the best form of government for any civil society, see Riley, Jonathan, “Mill's Neo-Athenian Model of Liberal Democracy,” in Urbinati, Nadia and Zakaras, Alex, eds., J. S. Mill's Political Thought: A Bicentennial Reassessment (Cambridge and New York: Cambridge University Press, 2007), 221–49CrossRefGoogle Scholar.

67 Mill tells us that he regards “utility as the ultimate appeal on all ethical questions; but it must be utility in the largest sense, grounded on the permanent interests of man as a progressive being” (On Liberty, 224). With respect to the interest in security associated with justice, security is a kind of pleasure or satisfaction (including freedom from pain) which can only be had in the context of a code of equal rights. The amount of security experienced by any individual varies across different codes, depending on whether he is regarded as a member of the relevant community to whom equal rights are distributed and, if so, on the contents of the rights distributed to all members. If he is excluded from the group, then he experiences no security. For any given population, general security can be maximized only if equal rights are distributed to all individuals, none being excluded. Mill realizes that the idea of justice as equal rights has been clouded historically by mistaken notions of expediency that call for excluding this or that subset of people from the rights distribution. But this is simply another way of saying that history is a record of injustice, of unfair societies in which general security has not been maximized. See, e.g., Mill, The Subjection of Women, in CW, XXI.

68 For Mill, the lexical priority of a code of justice over competing considerations is justified by the infinite superiority of this higher kind of utility over competing kinds. In contrast, Kant and John Rawls in their different ways trace the absolute priority of the code to a self-justifying principle of practical rationality or procedure of public reasoning that transcends and constrains any person's idea of his utility or good. See, e.g., Rawls, John, Political Liberalism, expanded edition (New York: Columbia University Press, 2002)Google Scholar.

69 Mill, Utilitarianism, 251.

70 Mill, On Liberty, 224–25.

71 I am ignoring the relatively rare occasions on which the scoring procedure may fail to select a Condorcet winner when it exists.

72 Unlike the sophisticated utilitarianisms discussed earlier, Millian utilitarianism supposes that any prudent and impartial individual must be willing to universalize his prescriptions of the rights which he thinks are best for promoting his own security. Thus, he is willing to assert a general rule that distributes equal rights and correlative duties which all rational agents might adopt with a view to promoting their own security. In addition, he must be willing to comply with the rule which emerges as the outcome of the impartial political process, even if the rule favored by the majority is not his personal favorite. In contrast, both Hare and Harsanyi require, in effect, that the individual must be willing to universalize his prescriptions so as to maximize an independent notion of general welfare which is fully understood only by a fictitious impersonal observer in possession of the requisite cardinal and interpersonally comparable utility information of homogeneous quality.

73 Instead of giving extra weight to the utilities of the worst-off people when calculating overall good, as in Brad Hooker's rule consequentialism, Mill's utilitarianism as interpreted protects the worst-off by impartially distributing an equal claim to subsistence or to some minimum level of income guaranteed by the general taxpayer. Security from abject poverty is certainly something that most rational and impartial agents can recognize as a part of justice, at least in societies where a threshold level of collective wealth has been achieved. In addition, the Millian doctrine prescribes moral duties of charity, which are imperfect in the sense that they do not correlate to anybody's rights, and thus, unlike perfect duties of justice, admit some personal discretion as to the occasions of their performance. Supererogatory acts of giving to others are also viewed as praiseworthy, although they are not morally required. This tripartite classification of acts of redistribution into just, charitable, and supererogatory is seen as a dynamic social construction that evolves as the members of society develop their intellectual, moral, and practical capacities. As society advances, competent majorities may choose to reclassify acts formerly seen as acts of charity into acts of justice, for instance, so that a right to some level of poor relief comes to be accepted as morally required. Or acts of sacrifice formerly viewed as supererogatory may come to be seen as morally obligatory on at least some occasions. For further discussion, see Riley, Jonathan, “Mill's Extraordinary Utilitarian Theory of Morality,” Politics, Philosophy, and Economics 8 (2009)Google Scholar. Hooker's distinct approach to these matters is discussed in his Ideal Code, Real World, 147–74.

74 Recall that the qualitative difference implies a discontinuity of value in any list of pains or dissatisfactions such that any finite amount of injustice—the kind of suffering that is associated with violations of equal rights—is regarded by most rational people as incomparably worse than any competing finite amount of lower kinds of dissatisfactions or annoyances, however large.

75 Mill, Letter to Thornton, in CW, XV, 854.

76 As it turned out, the relentless Romans eventually did destroy Carthage at the end of the Third Punic War in 146 b.c.

77 Police, firefighters, medical personnel, and other professionals may well be much more likely than the ordinary person to find themselves in tragic situations akin to the runaway trolley case. A separate argument might be made for rules that distribute professional obligations to risk one's life to rescue a greater number of other people from accidental death in these situations. Such obligations would become part of the job description. In any case, it would not be surprising if professionals developed some such convention to govern interactions among themselves.

78 Brad Hooker also makes this point (Ideal Code, Real World, 169).

79 This point about freedom does not require acceptance of Mill's doctrine of a moral right to absolute liberty of so-called purely self-regarding conduct. For a defense of the self-regarding liberty doctrine, see Riley, Mill on Liberty; and Riley, , Mill's Radical Liberalism (London: Routledge, 2009)Google Scholar.

80 As a result, Millian utilitarianism as conceived can consistently endorse both “an Ideal of Purity and an Ideal of Humanity,” in Scheffler's terms (Human Morality, 101–32). For Scheffler's claim about stringency, see Section III above.

81 For further discussion of my interpretation of Mill's utilitarian theory of justice, see Jonathan Riley, “Justice as Higher Pleasure,” to be published in John Stuart Mill: Thought and Influence, ed. Paul J. Kelly and Georgios Varouxakis; and Riley, “Mill's Extraordinary Utilitarian Theory of Morality.” For a Millian response to Sen's Paretian liberal impossibility result, see Riley, , “Liberal Rights in a Pareto-Optimal Code,” Utilitas 18 (March 2006): 6179CrossRefGoogle Scholar.

82 Mill, Utilitarianism, 218.