To this end, if it is assumed, as some economists do, that utility functions are the same for all people, then the maximisation of total utility, for a given population size, will give equal distribution. However, if utility functions are not the same across individuals (which is a more sensible assumption to make), then the instruction to maximise total utility will result in extremely unequal distribution levels for individuals. This situation is not significantly changed if the maximisation of utility through the market is backed by decent government-backed safety nets. This is because it does not follow from it that “maximizing total utility would justify a decent minimum for everyone” [Allen Buchanan (1985); p. 59; italics in the original]. Furthermore, utility as a subjective mental state measure does not adequately register the pleasure/pain experiences of different individuals. Thus, a disabled person, or one who is resigned to his/her fallen state (e.g., a slave), may not register the same pleasure/pain as a person enjoying good health and the blessings of freedom. But, on the utilitarian scale, the latter will get more than the former; and some of the former may be altogether excluded from the benefits flowing from public policy. Thus, far from suggesting any progressive egalitarianism, utilitarianism will actually recommend a highly regressive social policy!
Secondly, utilitarianism suffers from a deep dilemma. it works only if interpersonal comparisons of the utilities of different people affected by a change is allowed; but in this case all the unsavoury distributional consequences noted above will follow. But, if interpersonal comparisons of individuals are not allowed, which is what neo-classical economists have invariably come to believe since the 1930s, then utilitarianism will lose all operational significance.12 It is for this reason that it has been replaced by the Pareto-optimality criterion, which does not make this assumption. In this sense, Pareto-optimality can be regarded as a second-best indicator of human well-being (the first-best indicator being the inoperational utilitarianism). But, as noted above, Pareto-optimality and utilitarianism, are entirely focused on efficiency and not at all on distributive justice; nor do they represent a mutually advantageous arrangement – the former, because it places restrictions on how the sum-total of utility is distributed as long as the net amount of benefits are as high as possible; and the latter, because it does not recognise any conflict of interest between parties to the exchange. Little wonder, then, that utilitarianism has all but been discarded in modern scientific discourse. The irony is that the inter-personal comparisons of utilities are again back in intellectual fashion; and yet utilitarianism remains discredited. The reason is that utilities are not the proper metric to register adequately and correctly any increase/decrease in human happiness, or well-being.
The (positive) collective-choice theories reviewed so far assume no trade-offs between rival interests; public choices, therefore, help everyone. We now turn to those (normative) collective-choice theories which postulate a more realistic social setting where ideas and interests and passions clash, sometimes violently. Since the social choices made inevitably generate changes which benefit some and hurt others, there must be a mechanism to redress at least some, if not all, the consequent inequities of the human condition. The problem of social justice is, then, to create a just (democratic) society which puts “all citizens in a position to manage their own affairs and to take part in social cooperation on a footing of mutual respect under appropriately equal conditions” [Rawls (1999); p. xv].13 This conception of a just society, which distinguishes itself by procedural impartiality as well, differs radically from that presented by a much narrowly defined utilitarian consequentialism – and is light years away from Nozickian non-consequentialism. Rather than compress all aspects of human well-being into just one object of value, i.e., the metric of utility, it is more realistic to recognise it to be pluralistic in character. This is because the criteria used to evaluate social states are essentially diverse. To this end, normative collective-choice theories (with some notable exceptions) move away from the subjective mental-state experiences to more objectively ‘seen’ evaluators of human happiness – e.g., primary goods, resources, functionings – all of which assign centrality to human freedom and admit the ‘possibility’ of improving social conditions by egalitarian public policy.
i) The Many Faces of Social Justice / Welfare
The analysis presented in the preceding section highlights the importance of the ‘right’ moral values – ones that guide individuals to undertake socially desirable activities, and those which enhance social justice.
a) Social Welfare
Social justice, or distributive equity, is generally understood to mean maximising social welfare in an individualistic economy. But this usage is not free from ambiguity because social welfare does not always imply distributive equity. Positive public-choice theories regard Pareto-optimality as an index of social good or welfare; which, however, is a misnomer. Even some normative public-choice theories toy with the idea of maximising social welfare along the Pareto-optimality frontier (i.e., choosing an optimal point on the frontier) with the help of a distortion-free optimal lump-sum taxes and transfer mechanism. This intellectual feat was achieved by Bergson (1938) and Samuelson (1947) who proved that maximising social welfare is equivalent to maximising a Bergson–Samuelson social welfare function.14 But these efforts have been much ado about nothing! The reason is that the basic components of the Bergson–Samuelson social welfare function do not allow it to address the problem of distributive justice. These are: (a) the distributionally-neutral Pareto-optimality criterion, with its unifocal concentration on efficiency-related issues; (b) a neutral lump-sum transfer of resources, which is no more than a hypocritical neo-classical trick of looking egalitarian without actually having to distribute anything; and (c) the impossibility of making interpersonal comparisons of individual utilities, which are no more than mental-state comparisons of pleasures or desires.15 The way to cure the neo-classicals’ trained incapacity to handle distributional issues is either to discard utilitarianism, or reject the impossibility of making interpersonal comparisons of individual utilities, or both. Harsanyi (1955) rejects the neo-classical’s consensus about the impossibility of interpersonal comparisons of utility, but retains the utilitarian calculus; while Rawls (1971; 1999) and Sen (1999a) reject both.16 And, chameleonlike, ‘social welfare’ changes its ‘colour’ and meaning in each of these formulations.
b) Harsanyi’s Just Social Welfare Function
Take interpersonal comparisons of utilities to be a factual proposition, and assume that if x is preferred to y by the individuals in a society, it is also preferred by the society as a whole. Then a remarkable result follows – namely, that social welfare (W) is a sum of the individual utilities.17 Harsanyi’s originality is to combine self-interested individualism (i.e., no role is assigned to the state even by implication) with the moral quality of caring for others. The deus ex machina is to assume that: (a) individual preferences can be divided into one’s self-interested personal preferences and his/her moral preferences which are an indicator of his/her concern for other persons in society; (b) that every individual has an equal probability of being any other individual; and (c) that each individual has the capacity of being in someone else’s shoes (determined by mental comparisons alone). Here (b) and (c) provide the ethical foundations of Harsanyi’s (1955) formulation.18 Once these assumptions are made, the impartiality of social decisions is ensured. Here, social decisions are made, not by the state, but by Adam Smith’s “impartial spectator”, who possesses the entirely agreeable qualities (a) to (c), and whose preferences are representable by a social welfare function.
The basic
