View sample Philosophical Aspects of Welfare Research Paper. Browse other research paper examples and check the list of research paper topics for more inspiration. If you need a religion research paper written according to all the academic standards, you can always turn to our experienced writers for help. This is how your paper can get an A! Feel free to contact our custom writing service for professional assistance. We offer high-quality assignments for reasonable rates.
‘Welfare’ is not an easy term to deﬁne with precision, even restricting attention to human beings. In the abstract, it refers broadly to the happiness, wellbeing, or good of an individual, society, political association, and so on. More concretely, it also refers to particular ingredients and sources of happiness or good. Life itself is an essential element of any human being’s welfare, for example, and the means of basic subsistence, including food and shelter to some minimal extent, can be viewed as vital sources of human welfare whose provision should be guaranteed by government, as in the term ‘welfare state.’ The sources and ingredients of human happiness are so various, however, that ‘welfare’ in the philosophical sense must not be limited to basic needs or the means of satisfying them. In this research paper, four diﬀerent interpretations of welfare are discussed critically. ‘Welfarist’ moral philosophies are also contrasted with ‘nonwelfarist’ moral philosophies. The ﬁrst step is to dispel some confusion surrounding the term ‘utility’ and its relationship to welfare.
1. Welfare And Utility
Modern economists, decision theorists, and game theorists typically distinguish between welfare and utility. Utility in their terminology refers to the value of a mathematical function—strictly speaking, any function in a family of informationally-equivalent functions—that represents the preferences of a person or of society such that, for any pair of alternatives in the feasible set of outcomes, a larger value is assigned to the preferred alternative. The relevant family of informationally-equivalent utility functions is determined by the given assumptions of measurability and interpersonal comparability of utility. In this regard, measurability must not be confused with comparability across persons. A measurability assumption (such as ordinal measure, cardinal measure, or ratioscale measure) speciﬁes a particular degree of precision with which any person’s preference ranking can be represented by utility numbers—the more precise, the smaller the family of informationally-equivalent utility functions. Such an assumption is logically independent of any comparability hypothesis, according to which the utility representations of diﬀerent persons’ preferences are mutually dependent. Given comparability, the utility function which can be used to represent a person’s preferences depends on the utility function selected from the family available to represent another person’s preferences. In the absence of comparability, however, there is no constraint of this sort on which members of the relevant families can be used to represent the diﬀerent persons’ preferences.
To illustrate, suppose that utility numbers are ordinal and noncomparable. Given the ordinal measure, a person’s utility function is unique only up to positive monotonic transformations. Such transformations preserve the order of relative priority of the utility numbers assigned to the alternatives. Given noncomparability, it is permissible to transform any person’s utility function without altering the utilities of other people. Thus, a person’s ordinal noncomparable utility ranking generally provides no more information than that contained in his own preference ordering—a complete, reﬂexive, and transitive relation—over the alternatives.
If utility numbers are assumed to be cardinal and noncomparable, however, then a person’s utility function is unique up to positive linear (strictly speaking, aﬃne) transformations: the cardinal noncomparable family of informationally-equivalent functions is more restricted than its ordinal counterpart, which admits positive nonlinear transformations. A cardinal noncomparable utility ranking represents preference information that is richer than his preference ordering, namely, the relative intensity with which he prefers one option to another. If his revealed attitudes toward risk are independently assumed to measure how much subjective importance he assigns to his diﬀerent options, for instance, the Von Neumann–Morgenstern method of cardinalization is appropriate and expected utility functions are valid cardinal noncomparable representations of his preference intensities (see, e.g., Binmore 1994, pp. 259–78, 304–15).
To take a ﬁnal illustration, if utility numbers are interpersonally comparable, then one person’s utility function cannot be subjected to any transformation unless every other person’s utility function is similarly transformed. A cardinal comparable utility ranking provides information not only about the person’s preference intensities, for example, but also about how intense his preferences are in comparison to those of other people. To construct such a utility ranking, however, there must be a method of making interpersonal comparisons apart from a method of cardinalization (Sen 1979). For instance, a person might extend sympathy to the other members of his society by adopting their personal preference intensities while imagining himself in their respective positions at any pair of alternatives. He must also be able to render the diﬀerent persons’ intensities dimensionally compatible in order to compare them. By reducing them to a common unit, he could form a suitable extended preference reporting that when he places himself in their respective positions, he believes that one person i’s preference for x over y is of greater intensity than another person j’s preference for y over x, for any two persons and any pair of alternatives. His personal preferences, a subrelation of his extended ones, might continue to be decisive for his actions. But if he can extend sympathy to others in this way, his preferences can be represented by a cardinal comparable utility function. At the same time, diﬀerent people may disagree with respect to interpersonal comparisons, and thus form conﬂicting extended preferences. A fortiori, diﬀerent people may have conﬂicting personal preferences and conﬂicting cardinal comparable utility functions.
In general, a utility function encapsulates information about measurability and interpersonal comparability of the underlying preferences. As a device for representing this preference information (including orderings, intensities, and interpersonal comparisons), utility in this technical sense is quite distinct from any substantive idea of welfare. Indeed, utility in this sense can conﬂict with welfare if the preferences being represented are not valid welfare judgments: larger utility values can be assigned to alternatives which, though preferred by the agent, are worse in terms of his welfare or better than alternatives he rejects. Some have argued that the term utility should thus be conﬁned to the technical meaning so as to avoid needless confusion resulting from the identiﬁcation of utility with ‘good’ (e.g., Broome 1991a, 1991b). Perhaps so. But there are reasons to proceed with caution in this. First, utility continues to be associated with welfare—‘being useful or beneﬁcial’—in our everyday language. Even within economics, this everyday association still surfaces occasionally. More confusion may be caused by trying to banish it than by living with it.
Second, the identiﬁcation of utility with welfare has a long history within both economics and philosophy (Sen 1987, 1991). Bentham, J. S. Mill, and other classical utilitarian philosophers held that utility means happiness or enjoyment in the sense of pleasure and absence of pain, for example, and that actions ought to be evaluated in terms of their consequences for general happiness or social welfare, maximization of which is the sole ultimate standard of good. For these thinkers, any action or other thing is ‘useful’ or ‘has utility’ if and only if it has a tendency to promote pleasure net of pain. They deny (what is commonly asserted) that something may give much pleasure yet be useless, or that something may be useful even though it tends to promote more pain than pleasure. Earlier thinkers such as Hume and Helvetius also may have held this hedonistic idea of utility if not the distinctive utilitarian doctrine that general utility maximization is the sole ultimate test of good conduct.
As Mill (1965, pp. 45–6) elaborated the hedonistic idea, three kinds of utilities can be produced by human actions, namely: ‘utilities ﬁxed and embodied in outward objects’ (such pleasures are created by ‘investing external material things with properties which render these serviceable to human beings’); ‘utilities ﬁxed and embodied in human beings’ (again, such enjoyments are made possible by ‘conferring on human beings qualities which render them serviceable to themselves and others’); and, ﬁnally, ‘utilities not ﬁxed or embodied in any object, but consisting of a mere service rendered; a pleasure given, an inconvenience or a pain averted, during a longer or shorter time, but without leaving a permanent acquisition in the improved qualities of any person or thing; the labor being employed in producing an utility directly, not (as in the two former cases) in ﬁtting some other thing to aﬀord an utility.’ Useful material objects and educated human beings are means to pleasure, of course. After repeated experience of their consequences, however, these more or less permanent sources of pleasure can become inseparably associated with the idea of happiness itself, and thus be seen as concrete ingredients of the good (Mill 1969, pp. 234–9).
Perhaps because utilitarians were so inﬂuential in the development of classical economics, the hedonistic interpretation of utility was also widely accepted by economists prior to the ‘ordinalist revolution’ of the 1930s and 1940s. Classical economists were preoccupied with the production and distribution of material wealth, however, and so tended to focus their attention on the kind of utilities embodied in material objects susceptible to accumulation. As used by Mill (1965, pp. 48–54), for example, the term ‘productive labor’ is conﬁned to those types of actions that produce utilities embodied in material objects. Actions are classed as ‘unproductive labor’ (however useful to society) to the extent that they produce the other two kinds of utilities. Moreover, ‘productive consumption’ (i.e., investment) includes consumption by productive laborers of things which are necessary to maintain and increase their own health, strength, capacities of work, or population. But it does not include any consumption by unproductive laborers (those who do not produce material wealth) or consumption of luxuries by the productive themselves (however pleasing these forms of consumption are to the community). In short, although they did not deny that other kinds of utilities existed, economists in the classical tradition (from Smith and Ricardo to Marshall and Pigou) tended to identify utility with material welfare and to be especially concerned about the needs of those members of the community who produced the kind of utilities embodied in material wealth.
In the post-Second World War period, utilitarian ethics have declined in inﬂuence and hedonistic conceptions of utility have largely been abandoned by economists and philosophers. Whereas philosophers (including utilitarians) have typically gone on to identify utility with some nonhedonistic idea of welfare, however, economists have generally adopted the technical idea of utility in its purely ordinalist form— utility is merely a convenient mathematical tool for representing interpersonally noncomparable preference orderings. It is not obvious that philosophers should follow economists in viewing utility as a merely technical term, distinct from welfare. (For a controversial argument that this shift in the meaning of utility within economics—the product of the ‘ordinalist revolution’—does not imply ‘progress’ because it has involved ‘losses in understanding human welfare,’ in particular, ‘material welfare,’ including any person’s objective interpersonally comparable material needs; see Cooter and Rappoport 1984.) But perhaps they should, without necessarily accepting the narrow restriction to pure ordinalism. Even so, it would be incorrect to conclude that because technical utility values are distinct from, and can conﬂict with, substantive welfare values, there is some inescapable contradiction between utility in the technical sense and welfare. This leads to a third reason for caution in insisting on a divorce between utility and welfare, to wit, the technical idea of utility may remain a formal aspect of any adequate conception of welfare if judgments about welfare are a species of preferences that can be represented by utility functions. In that case, technical utility, as a formal component of the idea of good, could be seen as a vestige of earlier utility concepts that identiﬁed utility with welfare.
Whether or not judgments about welfare can be represented by utility functions, the question remains: what are such judgments about?
2. Conceptions Of Welfare
At least four diﬀerent conceptions of welfare continue to be debated in the literature.
2.1 Welfare As Pleasure
Although currently out of favor, hedonistic conceptions have not vanished entirely. Hedonism is typically dismissed as crude and naive, however, both as a psychology and as a theory of welfare. Expectation of pleasure is not the sole motivational force in human nature, it is said, and human good involves myriad things other than pleasure. No rational and moral person would be happy, for example, if his life consisted solely of a stream of supreme pleasures guaranteed by artiﬁcial means.
Perhaps such arguments are decisive against some versions of hedonism. But, as already discussed, the classical hedonism of Mill, Bain, and others does not deny that many things are means to pleasure which can also become valued as ends-in-themselves, in which case they are seen as concrete ingredients of happiness. Classical hedonism is not a purely subjective doctrine. It insists that some kinds of pleasures are embodied in material objects and human qualities, and can only be experienced by people who possess those objects and qualities, which must ﬁrst be produced. Particularly intense intellectual and moral pleasures can only be experienced by agents who possess the relevant intellectual and moral capacities, for example, and those capacities must be developed in part by the agents themselves.
2.2 Welfare As Preferences
Perhaps currently most inﬂuential, preference-based approaches typically conceive personal welfare in terms of personal preferences and social welfare in terms of social or ethical preferences that are constructed in some way from the set of personal preferences. Nevertheless, there is an ambiguity in the key notion of ‘preference-satisfaction’: it may mean the fulﬁllment of preference by a suitable shift in the way the world is; or it may mean the satiation of preference by a suitable shift in the subject’s conception of the way the world is. Proponents of preference-based approaches typically insist on the latter ‘satiation’ interpretation because of their implicit subjectivism. Fulﬁllment without satiation (e.g., posthumous fulﬁllment of preferences) does not count. From this subjectivist perspective, hedonism can be viewed as a special case, in which a claim is made that expectation of pleasure is the sole ultimate motivation of preference. But defenders of preference-based conceptions usually reject that claim as implausible.
Two versions of the preference-based approach should be distinguished. One deﬁnes ‘preferences’ as actual or hypothetical ‘choice behavior,’ which is in principle observable under certain conditions (see, e.g., Harsanyi 1992, Binmore 1994, 1998). A person’s preference for x over y is ‘revealed’ by his choice of x when y is available, independently of his desires or other motivations for the choice. In contrast, the other version deﬁnes preferences as ‘desires’ rather than ‘choices’ (see, e.g., Hare 1981, Griﬃn 1986). A preference for x over y implies greater desiresatisfaction (though not necessarily pleasure) at x than at y. Nevertheless, the agent could—and perhaps should—still make a counterpreferential choice of y over x in this case. Of course, the notion of ‘counterpreferential choice’ makes no sense under the choice-based approach. But that approach is silent about the motivations people have for making the choices it views as comprising personal or social welfare. It needs to be supplemented by a theory of motivation, not only to explain why people do make those choices but also to justify why those choices ought to be seen as conducive to welfare. For its part, the desire-based approach does not lack a theory of motivation. But its theory may well be inadequate. It is arguable that people are not always motivated to choose the options that they expect will bring them the greatest desire-satisfaction, nor should they be. Perhaps they should be motivated to sacriﬁce their greatest desires because of strategic considerations, for example, or moral commitments (Sen 1977).
However preferences are interpreted, the main appeal of preference-based conceptions seems to reside in their implicit subjectivism: the welfare of any individual or society is judged from the subjective viewpoint of that particular individual or society. The contents of welfare judgments may vary signiﬁcantly across diﬀerent persons or social groups. Nevertheless, the appeal of any purely subjective interpretation of welfare is quite limited. People may be deluded, for example, or illinformed, or resigned to their customary positions as slaves or ‘untouchables’ in highly inegalitarian societies. Their personal preferences may thus be poor indicators of personal welfare. Moreover, people may be unable to imagine themselves in others’ shoes, or they may clash over interpersonal comparisons, or display undue bias against others’ interests. As a result, they may fail to construct ethical preferences that can serve as social welfare judgments.
Defenders of preference conceptions typically respond to these diﬃculties by suitably laundering preferences. Conditions are speciﬁed that a person’s choices or desires must satisfy before they can be accepted as conducive to his own welfare or that of his society. Thus, for example, Harsanyi (1992) suggests that a person’s choices are conducive to his own welfare only if they are informed and conform to the expected utility axioms, whereas his choices are conducive to social welfare only if, in addition, they are ethical. Ethical choices must conform to the equiprobability model of moral choice. The person must forget his own position (deﬁned broadly to include family history and genetic makeup) and imagine himself occupying the position of any person in society with equal probability. He must also extend sympathy to each person j by making the same ‘self-oriented’ choices j would make while in j’s shoes. He can then infer interpersonal comparisons of cardinal utility which—on the independent assumption that all persons share a fundamentally similar human nature (including attitudes toward risk used to measure the relative importance of diﬀerent choices), which is revealed to anybody who ﬁghts against his biases by adopting the ‘impersonal’ perspective of the equiprobability model—are identical to those inferred by anybody else who goes through this procedure. Finally, he must aggregate impartially over the cardinal comparable utilities he experiences at each person’s position to construct social welfare judgments, which turn out to be (average) utilitarian in form. (For diﬃculties surrounding the interpretation of Harsanyi’s theorems as derivations of utilitarianism, see, e.g., Sen 1986, pp. 1122–4.)
Even if such hypothetical preferences have appeal as conceptions of welfare, the subjectivism which is usually thought to be the main strength of preferencebased approaches is lost unless agents actually aﬃrm the hypothetical preferences under the relevant conditions. To preserve an element of subjectivism, individuals and societies must willingly adopt those preferences before they can be said to be conducive to the welfare of those particular individuals or societies. If actual adoption of the preferences is a necessary condition of good, however, attention is redirected to theories of evolution and education which explain how people can be motivated to develop the intellectual and moral capacities required to make the relevant choices or form the relevant desires. It is not clear whether a credible theory of this sort can be supplied to anchor the desire-based variants. But whether or not human capacities can be developed such that human beings invariably obtain the greatest desire-satisfaction from making the choices that comprise personal or social welfare, defenders of choice-based versions of the preference conception require a credible theory of motivation.
2.3 Welfare As Resources
In resource-based conceptions, personal welfare is deﬁned in terms of an objective list of resources broadly deﬁned, whereas social welfare consists implicitly of the eﬀective production and fair distribution of the relevant resources. Rawls (1971, 1993) can be read as oﬀering such a conception, for example. He refers to ‘primary goods’ as ‘things which every rational man is presumed to want’ and ‘things citizens need as free and equal persons,’ and provides a (perhaps incomplete) list that includes ‘basic rights and liberties,’ ‘freedom of movement and free choice of occupation against a background of diverse opportunities,’ ‘powers and prerogatives of oﬃces and positions of responsibility in the political and economic institutions of the basic structure,’ ‘income and wealth,’ and ‘the social bases of self-respect’ (1971, pp. 60–5, 1993, pp. 178–90). Dworkin (1981, 1990) spells out another approach of this sort.
There is little doubt that human welfare depends on the possession of such valuable things as those listed by Rawls or Dworkin. Data on the extent to which a person possesses these items may provide helpful of diﬀerent aspects of his good, even if the indicators cannot be combined into a single index of personal welfare. Similarly, data on the social distributions of these things may provide indicators of diﬀerent dimensions of social welfare, with the same caveat.
Nevertheless, resource-based conceptions face at least two diﬃculties. First, it seems that the relevant resources are conducive to welfare only because they are objects of suitably laundered preferences. As Rawls suggests, for example, primary goods are things which every person would prefer to have if he were rational and reasonable. Furthermore, since a person’s holdings of one resource (say, political responsibilities) can interfere with his holdings of another (say, wealth), it seems that his preferences must settle which balance of resources is proper for his welfare. Similarly, it seems that social or moral preferences are needed to settle which of many competing social distributions of resources are permissible if not essential to social welfare. If this is right, resources may well be essential ingredients of welfare but they properly ﬁgure as objects of preferences.
The second diﬃculty is that resource-based conceptions do not take account of interpersonal variation in human capacities to convert resources into the achievement of ends (Sen 1990). This problem leads to another way of conceiving welfare.
2.4 Welfare As Capabilities
Sen (1985a, 1985b, 1990) argues that human well-being should be conceived in terms of ‘capabilities’—or ‘positive freedoms’—to achieve valuable human ‘functionings’ given by some objective list. He points out that diﬀerent people may have unequal capabilities to convert resources into achievements. Physically handicapped people may need more income than the able-bodied need to achieve a given level of mobility, for example. People who are genetically vulnerable to particular diseases may need more medicines than the less vulnerable need to achieve a given standard of health. Women may need more food than men need to achieve a given standard of nutrition. Racial minorities may need more educational and job opportunities than white males need to achieve a given level of career success if racist attitudes are prevalent. The elderly may need more wealth than the young need to achieve a given standard of living. These are essentially empirical questions.
Such examples suggest that capability-based conceptions can remedy a serious deﬁciency in resource-based approaches to welfare. But, as with resources, it seems arguable that certain capabilities are properly viewed as essential components of welfare because they are objects of suitably laundered preferences. Some capabilities are in eﬀect things that every rational or moral person is presumed to want or choose. To the extent that anybody lacks these benchmark freedoms, reasonable people may agree that some form of special compensation ought to be provided so that the person can function ‘normally,’ for example, income subsidies, extraordinary medical assistance, aﬃrmative action in hiring decisions, and so on. Furthermore, beyond the benchmark capabilities which everyone is presumed to need, it seems that a person’s preferences must determine which balance of extra capabilities is proper for his welfare since capabilities to achieve some functionings can presumably interfere with capabilities to achieve others. Again, since one person’s capabilities might interfere with another’s, social or moral preferences are needed to settle which of many competing social distributions of capabilities (including benchmark capabilities) is permissible if not essential to social welfare.
If this is right, it seems that some version of the preference-based conception of welfare may have appeal as a way to combine insights from the various approaches discussed. The upshot is that preferences must be suitably laundered not only to satisfy certain formal conditions of rationality and morality but also certain substantive conditions. In particular, preferences can be welfare judgments only if they give due weight to certain valuable resources, making due allowance for interpersonal variation in capabilities. Moreover, a credible theory of motivation is required to explain how individuals can come to hold the relevant preferences, even if hedonism is rejected.
If some version of the preference approach has such potentially broad appeal, welfarist moral philosophies might adequately capture the phenomenon of human welfare. This will be the case if the relevant preferences can invariably be represented by technical utility functions, and if utility information is all that matters for the construction of social or ethical preferences.
3. Welfarist Philosophies
Welfarist philosophies are distinguished by their exclusive reliance on utility information. Perhaps the most well known examples of such philosophies are the standard versions of utilitarianism. But there are many diﬀerent forms of welfarist philosophies. They can be clariﬁed by means of the concept of a social welfare functional (SWFL) (Sen 1986, pp. 1111–28). For a given set of alternatives, any SWFL maps a given set of personal utility functions (one for each person in society) into a social or ethical preference which can itself be represented by a technical utility function. Since each person’s utility function is an element of some family of informationally-identical utility functions as determined by the given assumptions of measurability and comparability, however, the SWFL is required to satisfy an invariance requirement: it must generate the same social or ethical preferences if any given set of personal utility functions is subjected to transformations permitted by the measurability and comparability assumptions.
Welfarism is a condition of strong neutrality imposed on the form of a SWFL. It requires that if each person’s utility values with respect to one pair of options x and y are equal to his utility values with respect to another pair w and z—subject to transformations permitted by the given measurability and comparability assumptions, then the SWFL must generate the same social or ethical preference over the one pair as over the other. The only information that matters for a welfarist SWFL is utility information. Nonutility information has no inﬂuence.
Arrow’s (1963) well-known ‘impossibility theorem’ shows that any welfarist SWFL must generally be dictatorial if personal utility functions are ordinal and noncomparable. But a dictatorial procedure cannot generate social or ethical preferences that give positive weight to each person’s utilities: the dictator sympathizes with nobody but himself in determining social preferences. (For variations and extensions of Arrow’s theorem, including its extension to the case of cardinal noncomparable utilities, see Sen 1986, pp. 1073–106, 1114–15.) It follows that welfarism can have appeal as a philosophy of human welfare only if utility information is suﬃciently rich to support the interpersonal comparisons required to construct ethical preferences. As Harsanyi’s equiprobability model of moral choice illustrates, one person can be assumed to replicate the operation of a welfarist SWFL over a set of comparable personal utility functions, without that person becoming an Arrovian dictator.
Although various forms of welfarist philosophies making use of comparable utilities have been investigated in the literature, doubts remain that any adequate philosophy of human welfare can be put into a welfarist form.
4. Beyond Welfarism
Even if some version of the preference-based conception turns out to have broad appeal, a nonwelfarist philosophy—one that relies to some extent on nonutility information—may be required to capture that conception of human welfare. Myriad forms of nonwelfarist philosophies are conceivable. Such a philosophy may take the form of a SWFL that does not satisfy the relevant strong neutrality condition. The SWFL might make use of nonutility information about the alternatives—some alternatives might belong to an individual’s ‘protected sphere’ of choice, for example, or some might involve ‘primary goods’—to generate ethical preferences that can be represented by utility functions. Or it might rely on nonutility information to distinguish between diﬀerent kinds of utility functions representing diﬀerent kinds of preferences, some of which are held to be inherently more valuable than others (Riley 1988, 2001). The relevance of any particular nonutility information is determined by the conception of welfare under consideration. In the latter case, for example, the information that diﬀerent kinds of utility functions ought to be identiﬁed and arranged in a lexical hierarchy is supplied by the idea of welfare itself, which is taken to have this sort of multidimensional structure.
Nonwelfarist philosophies may also eschew the form of an SWFL. Harsanyi’s (1992) version of rule utilitarianism takes the form of a two-stage game, for example, in which a cooperative game is played at the ﬁrst stage to select an optimal moral code and a noncooperative game is played at the second stage where players are restricted to strategies permitted by the optimal code. This game structure gives a special status to whatever outcome is assumed to prevail in the absence of agreement at the ﬁrst stage. Also, the rules of any selected code will govern right and wrong actions independently of any strong neutrality condition. Alternatively, a nonwelfarist philosophy might take a value pluralistic form, in which welfare is only one of plural irreducible values in ethical evaluations. Sen (1985a, 1985b) proposes a dualistic structure in which the value of ‘well-being’ cannot be reduced to that of ‘agency,’ for instance. By implication, incompleteness or even inconsistency might arise in all-things-considered ethical preferences if well-being comes into conﬂict with agency.
- Arrow K J 1963 Social Choice and Individual Values, 2nd edn. Wiley, New York
- Barry B 1973 A Liberal Theory of Justice. Clarendon Press, Oxford, UK
- Binmore K 1994 Playing Fair. MIT Press, Cambridge, MA
- Binmore K 1998 Just Playing. MIT Press, Cambridge, MA
- Broome J 1991a Utility. Economics and Philosophy 7: 1–12
- Broome J 1991b Utility—ideas and terminology—reply. Economics and Philosophy 7: 285–7
- Cooter R, Rappoport P 1984 Were the ordinalists wrong about welfare economics? Journal of Economic Literature 22: 507–30
- Dworkin R 1981 What is equality? Part 2: Equality of resources. Philosophy and Public Aﬀairs 10: 283–345
- Dworkin R 1990 Foundations of liberal equality. In: McMurrin S (ed.) The Tanner Lectures on Human Values. University of Utah Press, Salt Lake City, UT, Vol XI
- Griﬃn J 1986 Well-being. Clarendon Press, Oxford, UK
- Hare R M 1981 Moral Thinking. Clarendon Press, Oxford, UK
- Harsanyi J C 1992 Game and decision theoretic models in ethics. In: Aumann R J, Hart S (eds.) Handbook of Game Theory. Elsevier, Amsterdam, Vol 1
- Mill J S 1965 Principles of political economy. In: Robson J M (ed.) Collected Works of J.S. Mill. University of Toronto Press and Routledge & Kegan Paul, Toronto and London, Vols II–III pp. 48–54
- Mill J S 1969 Utilitarianism. In: Robson J M (ed.) Collected Works of J. S. Mill. University of Toronto Press and Routledge & Kegan Paul, Toronto and London, Vol X pp. 234–9
- Rawls J 1971 A Theory of Justice. Havard University Press, Cambridge, MA
- Rawls J 1993 Political Liberalism. Columbia University Press, New York
- Riley J 1988 Liberal Utilitarianism. Cambridge University Press, Cambridge, UK
- Riley J 2001 Maximizing Security: A Liberal Utilitarian Theory of Justice and Rights. Clarendon Press, Oxford, UK
- Sen A K 1977 Rational fools: a critique of the behavioral foundations of economic theory. Philosophy and Public Aﬀairs 6: 317–44
- Sen A K 1979 Interpersonal comparisons of welfare. In: Boskin M J (ed.) Economics and Human Welfare. Academic Press, New York
- Sen A K 1985a Commodities and Capabilities. Elsevier, Amsterdam
- Sen A K 1985b Well-being, agency and freedom. J. Philos. 82: 169–221
- Sen A K 1986 Social choice theory. In: Arrow K J, Intriligator M D (eds.) Handbook of Mathematical Economics. North-Holland, Amsterdam, Vol III pp. 1073–1128
- Sen A K 1987 On Ethics and Economics. Basil Blackwell, Oxford, UK
- Sen A K 1990 Justice: means versus freedoms. Philosophy and Public Aﬀairs 19: 111–21
- Sen A K 1991 Utility: ideas and terminology. Economics and Philosophy 7: 277–83