Possibilities of a Democratic Social-Welfare Ordering: The Need for a ‘Theory of Justice’
As said earlier, the attempt to derive a social order from individual preferences based on a set of postulates resulted in Arrow's (1951) ‘impossibility theorem'. This established a vital failure of the first pillar of economic policy.
The practical ‘shortcut' devised by Frisch of referring the SWF to policymakers seemed not to be acceptable in theoretical terms. In fact, this solution was based on the assumption that the policymakers (and thus their preferences) reflect citizens' values. The process of selecting politicians would be delegated to the political arena - hopefully in a way that ensures democracy. However, choice of democratic institutions and the form of voting that can ensure them then are also expressions of values. This choice is possibly the most preferred outcome for most citizens, but it remains a choice between values having a subjective nature. This cannot be introduced from the outside if the issue is to be the derivation of social values from individuals' preferences with no external interference, i.e. if values are not given explicit citizenship in the process of searching for a social order.It was Leibenstein (1962: 316) who first stated that ‘the choice rules suggested by contemporary welfare economics are really ethical rules' not only because some of the terms employed - for example, ‘efficiency' and ‘the optimal allocation of resources' - are value loaded but also because procedural rules for (possibly) reaching a social outcome (e.g. the voting rule or the choice principle) protect some individuals or better ensure democratic choices. We can then draw the conclusion that it is a theory of justice that must tell us which is the preferred principle of fairness, according to which the social-choice paradox must be solved.
Various attempts have been made to get around the ‘impossibility' result along this route by eliminating one or another of the axioms that Arrow (1951) thought necessary for constructing a complete social ordering.
Special attention has been directed to assumptions regarding the information on individual satisfactions (measurability and comparability). The Pareto principle is founded on the specific ‘informational' postulates of ordinal measurability and noncomparability. It has been shown that the other axioms introduced by Arrow also rule out cardinal measurability and/or interpersonal comparability of utilities (see, among others, Mueller 1989: chap. 20). Some have argued that these are the key hypotheses - originating from specific values - and that removing them would allow us to construct a complete social ordering.[31]Abandoning simple ordinal measurability does not enable us to avoid the impossibility result, however (Sen 1970a). A complete, non-dictatorial social ordering can be obtained if, in addition to cardinal measurability, some degree of interpersonal comparability is admitted. In particular, if we allow full interpersonal comparability in addition to cardinal measurability of individual utilities, it is possible to construct a simple or generalised utilitarian SWF (see below). ‘The additional information availability allows sufficient discrimination to escape impossibilities’ (Sen 1999b: 357).
Given the necessity of using some sort of interpersonal comparison to define complete social rankings, we must specify the principle that will enable us to evaluate all possible alternative situations, especially in cases where opting for one or another state means improving one person’s situation while worsening that of another. This implies a problem of distributive justice, which consists in defining criteria to judge the desirability of alternative ways of (1) assigning the participants in the productive process (more generally, the members of a society) the benefits of that activity and (2) apportioning the related costs.
Obviously, such criteria depend on each individual’s view of the world. It is therefore no surprise that economists attempting to construct social orderings have often taken their inspiration from principles of social or political philosophy.
This was so e.g. for Pigou (1920), who adopted certain principles of utilitarianism. In some cases, economists have studied these disciplines directly, as e.g. for Harsanyi (1976) and Sen (1970a). In a number of cases, principles of distributive justice have been developed by social philosophers, such as Nozick (1974) and Rawls (1971).When the need for criteria to evaluate distribution became clear to economists, efforts to construct a social ordering found many points in common with contemporary social or political philosophy, which carries on the work begun by modern moral philosophy with partly new concepts and methodologies. In evaluating distribution (and any other aspect of social life), the big issue is whether to accept or not the view that (1) each individual is the best judge of his or her preferences and (2) the value of possible social states is determined by the individual’s perception of these states (and the satisfaction obtained). According to this view, which constitutes the postulate of ‘ethical individualism’, this should be the only allowable basis for assessing social states, and anything not considered relevant by the individual must not be assessed by society as a whole. This principle (‘welfarism’, as Sen terms it; see e.g. Sen 1987), again, is the object of choice and can be debated.[32]
There are criteria of distributive justice that incorporate the postulate of ethical individualism and can be expressed in the form of a SWF, such as the utilitarian and those suggested by Bernoulli-Nash (see Nash 1950), Rawls (1971), Bergson (1938) and Samuelson (1947) (see Acocella 1994 more extensively). Other criteria do not accept ethical individualism but are still of much interest (e.g. Nozick's and Sen's theories of justice).
In particular, Sen has not only demonstrated that the principle of Pareto optimality can be in contrast with democracy (Sen 1970b), but he has also indicated remedies (Sen 1970a). By introducing the innovative concepts of ‘functionings' and ‘capabilities', Sen (1980, 1985) succeeded in merging consideration of material aspects and the results achieved by individuals (including distribution), on the one hand, and rights and liberties, on the other.
He began by observing that both the quantity of a good (food, for example) and the utility generated by its use are inadequate indicators of the welfare of an individual or a community. The total quantity of food available to a community is inadequate as a guide because some people may not have access to it. Nor is utility of individuals any more successful as a criterion because it is a psychological indicator that is incapable of (fully) revealing certain effects. For example, the effects of malnutrition - intended as either a shortfall or excess of food - are not always perceived by the individual; one may take pleasure in a single crust of bread, but the lack of certain fundamental nutrients may scar one's body for life.There are in fact non-psychological aspects of goods that are central to evaluating their advantages to people and
Advocates of ethical individualism hold that if the preferences of individuals are replaced by those of one person (e.g. a philosopher or social reformer) - perhaps not even a member of the community - paternalistic or authoritarian consequences can follow. Such a risk undoubtedly exists, but it can be practically eliminated by setting out which features of the social state - such as certain rights and liberties - must be preserved, by granting them constitutional status. society. Goods have characteristics that people may use to perform certain functionings, and it is the achievement of these functionings (being well nourished, healthy and able to move; having self-respect; being respected; being able to take part in the life and progress of the community) that indicates the benefit enjoyed by people, allowing them to exercise positive liberty. Moreover, it is not only the effective performance of certain functionings that is important but also the very possibility (the ‘capability’) of performing them even if they are not actually performed. For example, it is important to have the right to move without legal or material obstacles, even if a person should decide not to move.
Similarly, freedom of speech is a capability, in that it allows everyone to express their views: it does not ‘require that a person should be continuously speaking but that he should be able to speak if he were to so choose’. Obviously, for basic ‘capabilities’ such as feeding oneself adequately, enjoying good health and acquiring an education, the element of choice may be unimportant, since people will normally take advantage of them when possible. In this case, the capability of performing certain functions corresponds to their effective performance, and the distinction between capabilities and functionings largely disappears; in other cases, however, the distinction remains important (Sen 1982: 29-31; see also Sen 1992: chap. III).Other economists have moved in the same direction. Desai (1991) has found an application in the study of poverty and human development indicators constructed by the World Bank; in particular, purchasing power, education and health are the proxies of ‘capabilities’ that are used to construct indicators of human development. One specific aspect of these indicators, education, on which we will insist in Chapter 7 as an essential element for democracy, deserves more discussion. Achievements of educational functioning are of specific relevance, as they indicate that a wider range of educational choices and perspectives has been available to people. We should look at achievements in terms of education rather than the amount of money put into it or even years of schooling to really assess enhanced functionings and capabilities. This would involve evaluation of the content of schooling, whether it enhances literacy and the capability to process information (Unterhalter, Vaughan and Walker 2007).
Sen's conception of justice is also shared in particular by Martha Nussbaum, who holds that each person's capabilities should be considered as an end, not a means, for the life of people, and a minimum level of his or her capabilities that express his or her quality of life should be ensured by the state (e.g.
Nussbaum 2011). Nussbaum (1999) lists ten ‘central human functional capabilities' that must be respected by a just society: (1) life of a normal, natural duration, (2) bodily health and integrity, including adequate nourishment and shelter, (3) bodily integrity regarding, for example, freedom of movement and security against assault, (4) freedom to exercise one's senses, imagination, and thought as one pleases, which includes freedom of expression, (5) freedom to form emotional attachments to persons and things, which includes freedom of association, (6) the development and exercise of practical reason, the capacity to form one's own conception of the good and to try to plan one's own life, which includes the protection of freedom of conscience, (7) freedom of affiliation on equal terms with others, which involves provisions of non-discrimination, (8) concern for and possible relationships with animals, plants and the world of nature, (9) the freedom to play, to seek amusement and to enjoy recreational activities and (10) some control over one's own political environment, including the right to vote, and one's material environment, including the rights to seek meaningful work and to hold property. All these capabilities are essential to our functioning as flourishing human beings and should be ensured for all citizens of a just society. Anand, Hunter and Smith (2005) find strong evidence that the Nussbaum’s list of capabilities influences well-being.Starting from acceptance of many of the ideas indicated, other later investigations have also added to the ‘possibilities’ of constructing social-welfare orderings or socialordering functions that had been indicated by Sen (see e.g. Arrow, Sen and Suzumura 1997, 2002, 2011). More specifically, they are now addressed by an ancillary discipline that has emerged in more recent decades from previous studies,
i. e. ‘implementation theory’ (see e.g. Maskin and Sjostrom 2002; Fleurbaey and Maniquet 2011). This takes the incentive problem to the forefront of analysis and heavily relies on game theory, thus to some extent dealing also with some of the minor problems raised by asymmetric information. Designing a mechanism that provides incentives is equivalent to solving a problem of devising the proper institutions in a way to guarantee equilibrium outcomes consistent with some choice rule, or to ‘implement’ it. These advances make the first part of the core of economic policy, i.e. the logic of economic policy, well founded and well connected to the second part.
3.2