Jump to ContentJump to Main Navigation
Women, Culture, and DevelopmentA Study of Human Capabilities$

Martha C. Nussbaum and Jonathan Glover

Print publication date: 1995

Print ISBN-13: 9780198289647

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198289642.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 16 June 2021

Gender Inequality and Theories of Justice

Gender Inequality and Theories of Justice

(p.259) Gender Inequality and Theories of Justice1
Women, Culture, and Development

Amartya Sen

Oxford University Press

Abstract and Keywords

Turning to concrete questions of justice for women, Sen introduces the issue of ‘co‐operative conflicts’ and argues that these conflicts are often rooted in traditional conceptions of women's role, which are internalized as ‘natural’ by the women themselves. Sen's contention is that the capabilities approach can handle these conflicts better than Rawlsian liberalism and economic utilitarianism. To Sen, the central problem is to confront the underlying prejudice directly and to outline the need for and scope of reducing inequalities in capabilities without accepting that this project necessarily causes great inefficiency.

Keywords:   capabilities, co‐operative conflicts, gender justice, justice, Rawls, utilitarianism

1 Practice and Theory

Empirical research in recent years has brought out clearly the extent to which women occupy disadvantaged positions in traditional economic and social arrangements. While gender inequalities can be observed in Europe and North America (and in Japan), nevertheless in some fields women's relative deprivation is much more acute in many parts of the ‘Third World’.

Indeed, there are extensive inequalities even in morbidity and mortality in substantial parts of Asia and North Africa. Despite the biological advantages that women have in survival compared with men (the ratio of women to men averages around 1.05 or so in Europe and North America, partly due to biological differences in mortality rates), the number of women falls far short of men in Asia and North Africa, though not in sub‐Saharan Africa. If we took the European and North American ratios as the standard, the total number of ‘missing women’ in Asia and North Africa would be astonishingly large (more than 50 million in China alone). Even if the sub‐Saharan African ratio of females to males is taken as the standard, the number of ‘missing women’ would be more than 44 million in China, 37 million in India, and a total exceeding 100 million world‐wide.2 While looking at female: male ratios in the population is only one way of examining the relative position of women, this approach does give some insight into the acuteness of the problem of gender inequality in matters of life and death. It also throws some indirect light on the history of inequalities in morbidity and of unequal medical care. Direct observation of these other data confirm the intensity of gender inequality in vitally important fields.3

I have begun with a rather stark account of some features of gender inequality. What bearing does a theory of justice have on our understanding and analysis of these dreadfully practical matters? One bearing is obvious enough. In describing some arrangements as ‘unjust’ we invoke—explicitly or by implication—some conception of justice, and it is necessary at some stage to come to grips with the appropriateness of the respective theories of (p.260) justice to pronounce judgement on these matters. An observation of inequality can yield a diagnosis of injustice only through some theory (or theories) of justice.

A second context is a bit more complex but no less important. The tolerance of gender inequality is closely related to notions of legitimacy and correctness. In family behaviour, inequalities between women and men (and between girls and boys), are often accepted as ‘natural’ or ‘appropriate’ (even though they are typically not explicitly discussed). Sometimes the operational decisions relating to these inequalities (e.g. providing more health care or nutritional attention to boys vis‐à‐vis girls) are undertaken and executed through the agency of women themselves. The perceived justness of such inequalities and the absence of any contrary sense of deep injustice play a major part in the operation and survival of these arrangements.4 This is not the only field in which the survival of extraordinary inequality is based on making ‘allies’ out of those who have most to lose from such arrangements. It is, therefore, important to scrutinize the underlying concepts of justice and injustice, and to seek a confrontation between theory and practice.

2 Co‐operative Conflicts

There are many areas of social organization in which all the parties have something to gain from having a workable arrangement, but the gains that are made respectively by different parties differ greatly from one working arrangement to another. There are co‐operative elements in these arrangements, but also elements of conflict in the choice of one arrangement rather than another.

This class of problem can be called ‘co‐operative conflicts’.5 Such problems have been investigated in the literature of economics and game theory in different ways. For example, what J. F. Nash (1950) calls ‘the bargaining problem’ is a case of co‐operative conflict in which each party has well‐defined and well‐understood interests which coincide with their objectives.

Sometimes, simplifying assumptions are made that eliminate crucial aspects of co‐operative conflicts. One example is the assumption (used powerfully by Gary Becker, 1981) that the ‘altruistic’ head of the family acts in the joint interest of all, and everybody else in the family has exactly the same rational perception of the family's joint interest, which they all want to maximize in a rational and systematic way. This avoids the problem of (p.261) conflict in co‐operative conflicts by making everyone pursue the same objectives, as a result of which they have no disharmony of interests, or of objectives. If women (or girls) die in much larger numbers than men (or boys), because of differential medical attention and health care, then this model requires that such differentials are what every member of the family (including the relatively more‐stricken women) rationally promote and their consequences are what they jointly seek.

The existence of conflicts is, however, fully acknowledged in game‐theoretic discussions of ‘the bargaining problem’ inside the family (see, for example, Manser and Brown (1980); Lundberg and Pollak (1994)). Different family members are seen to have partly divergent interests. It is taken for granted that every member of the family acts on the basis of promoting his or her rationally perceived individual interests, and there is no ambiguity about this. This has the effect of abstracting from the role of implicit theories of justice and of appropriateness, and instead of Beckerian ‘collectivism’, we have here thoroughly individualistic perception of interests and choices based on them.

There is an interesting contrast here that is worth a comment. The situation of real conflict between different members of the family is well caught by the game‐theoretic perspective in a way that the Beckerian formulation does not. On the other hand, the socially influenced perception of the absence of conflict between family members may well be closer to Becker's formulation than to the standard game‐theoretic one. What is needed is a combination, which acknowledges the possibility of real conflicts of interests (unlike in Becker's framework) coexisting with a socially conditioned perception of harmony (unlike in the standard game‐theoretic model). Implicit theories of justice and traditional understandings of what is ‘natural’ and ‘proper’ can play a major part in making people with divergent interests feel united around shared perceptions of common objectives. Thus, despite the illumination about conflicts provided by game‐theoretic models, they do tend to ignore some of the more important causal influences—related to perceptions of legitimacy—that give stability to extreme inequalities in traditional societies.6

Theories of justice are important in bringing out the tension between perceptions of justice and what may be required by the demands of fairness or less partial rational assessment. Practical uses of theories of justice can be particularly important in the long run, since social change is facilitated by a clearer understanding of tensions between what happens and what is acceptable. While such an impact may be indirect, and while the connections between ethical analysis (on the one hand) and social perceptions and (p.262) practical politics (on the other) may not be instantaneous, it would be a mistake to ignore the long‐run practical importance of a clearer understanding of issues of justice and injustice.

3 The Claims of Utilitarian Justice

No ethical theory has had as much influence in the modern world as utilitarianism. It has been the dominant mode of moral reasoning over the last two centuries. We can do worse than begin with the question: Why not go for the utilitarian theory of justice as the basis of analysis of gender inequality? The fact that utilitarianism had a radical role in providing effective critiques of many traditional inequities (Bentham's own 1789 practical concerns were much inspired by his outrage at what he saw around him) makes it particularly appropriate to look for a positive lead from that quarter.

Unfortunately, utilitarianism provides a rather limited theory of justice for several distinct reasons. First, utilitarianism is ultimately an efficiency‐oriented approach, concentrating on promoting the maximum sum total of utilities, no matter how unequally that sum total may be distributed. If equity is central to justice, utilitarianism starts off somewhere at the periphery of it.

It is, of course, possible to use utilitarianism to reject many inequalities, since inequalities are often also thoroughly inefficient. But given the lack of a basic concern with equality in the distribution of advantages, the utilitarian concentration on the promotion of utilities is not particularly oriented towards justice.

Secondly, the efficiency that utilitarianism promotes is, of course, specifically concerned only with the generation of utilities. Under different interpretations of utilities variously championed by different utilitarian authors, this amounts to promoting either maximal pleasures, or maximal fulfilment of felt desires, or maximal satisfaction of perceived preferences, or some other achievement in a corresponding mental metric.7 As was discussed in the last section, one of the features of traditional inequalities is the adaptation of desires and preferences to existing inequalities viewed in terms of perceived legitimacy. This plays havoc with the informational basis of utilitarian reasoning since inequalities in achievements and freedoms (e.g., in morbidities, mortalities, extents of undernourishment, freedom to pursue (p.263) well‐being) get concealed and muffled in the space of conditioned perceptions.

There is, in fact, some empirical evidence that the deprived groups such as oppressed women in deeply unequal societies even fail to acknowledge the facts of higher morbidity or mortality (even though these phenomena have an objective standing that goes beyond the psychological perception of these matters).8 Basing the assessment of justice on a measuring rod that bends and twists and adapts as much as utilities do, can be formidably problematic. The difficulties are certainly big enough to discourage us from looking for a utilitarian theory of justice as an ethical arbitrator or as a conceptual frame of reference for analysing the problem of gender inequalities.

4 The Rawlsian Theory of Justice

Compared to the utilitarian approach the Rawlsian theory of ‘justice as fairness’ has many decisive advantages. The Rawlsian theory also has merits in terms of scope and reach over more relativist and less universalist approaches that have sometimes been proposed.9

The Rawlsian approach avoids the peculiar reliance on selected mental characteristics that utilitarianism recommends. It also provides a foundation based on the idea of fairness that links the demands of justice to a more general mode of reasoning.10 The use of ideas of fairness, rationality, reasonableness, objectivity, and reflective equilibrium provides Rawls's theory of justice with a depth of political argumentation that is remarkably effective. More substantively, the concern with equity in addition to efficiency as reflected in Rawls's principles of justice puts equity at the centre of disputes about justice in a way that utilitarianism (peripherally concerned, as it is, with equity) fails to do.11

(p.264) The Difference Principle of Rawls focuses on primary goods as the basis of assessing individual advantages. Primary goods are things that every rational person is presumed to want, such as income and wealth, basic liberties, freedom of movement and choice of occupation, powers and prerogatives of office and positions of responsibility, and the social bases of self‐respect. In this list there is a clear recognition of the importance of a variety of concerns that affect individual well‐being and freedom and which are sometimes neglected in narrower analyses (e.g., in the concentration only on incomes in many welfare‐economic analyses of inequality).

Despite these advantages there are some real problems in using the Rawlsian theory of justice as fairness for the purpose of analysing gender inequality. In fact, these problems are quite serious in many other contexts as well, and constitute, in my judgement, a general deficiency of the perspective of the Rawlsian theory of justice. Perhaps the most immediate problem relates to Rawls's use of the respective holdings of primary goods as the basis of judging individual advantage. The difficulty arises from the fact that primary goods are the means to the freedom to achieve, and cannot be taken as indicators of freedoms themselves.

The gap between freedoms and means to freedoms would not have been of great practical significance if the transformation possibilities of means into actual freedoms were identical for all human beings. Since these transformation possibilities vary greatly from person to person, the judgements of advantage in the space of means to freedom turn out to be quite different from assessments of the extents of freedoms themselves. The source of the problem is the pervasive diversity of human beings which make equality in one space conflict with equality in other spaces.12 The particular issue of inter‐individual variations in converting primary goods into freedoms to achieve fits into a more general problem of divergence between different spaces in which the demands of equity, efficiency, and other principles may be assessed.

One of the features of gender inequality is its association with a biological difference which has to be taken into account in understanding the demands of equity between women and men. To assume that difference away would immediately induce some systematic errors in understanding the correspondence between the space of primary goods and that of freedoms to achieve. For example, with the same income and means to buy food and medicine, a pregnant woman may be at a disadvantage vis‐à‐vis a man of the same age in having the freedom to achieve adequate nutritional well‐being. The differential demands imposed by neo‐natal care of children also have considerable bearing on what a woman at a particular stage of life can or cannot achieve with the same command over primary goods as a man might have at the corresponding stage in his life. These and other differences, (p.265) in which biological factors are important (though not exclusively so), make the programme of judging equity and justice in the space of primary goods deeply defective, since equal holdings of primary goods can go with very unequal substantive freedoms.

In addition to these differences which relate specifically to biological factors, there are other systematic variations in the freedoms that women can enjoy vis‐à‐vis men with the same supply of primary goods. Social conventions and implicit acceptance of ‘natural’ roles have a major influence on what people can or cannot do with their lives. Since the sources of these differences may appear to be ‘external’ to the human beings, it is possible to expect that they can be somehow accounted in when constructing a suitable basket (and index) of primary goods. If this could be adequately done, problems arising from these ‘external sources’ would be accountable within Rawlsian calculus.

However, in many circumstances this may not prove to be possible. Some of the social influences appear in most complex forms and may be hard to formalize into some component of primary goods. The sources of pervasive social discouragement are often hard to trace and harder to separate out.

Perhaps more importantly, as was discussed earlier, some of the constraints that are imposed on what women are free or not free to do may closely relate to women's own perceptions of legitimacy and appropriateness. The presence of this influence plays havoc, as was discussed earlier, with the utility‐based evaluation of justice. That problem has some bearing on the Rawlsian perspective as well. The behavioural constraints related to perceptions of legitimacy and correctness can strongly affect the relationship between primary goods and the freedoms that can be generated with their use. If women are restrained from using the primary goods within their command for generating appropriate capabilities, this disadvantage would not be observed in the space of primary goods. It is not clear how these constraints, many of which are implicit and socially attitudinal, can be incorporated within the framework of the ‘external’ category of primary goods.

I would, therefore, argue that despite major advantages in adopting the Rawlsian theory of justice in analysing gender inequality, there are also serious problems, arising particularly from variations in the correspondence between primary goods and freedoms to achieve. These problems are not specific to gender justice, but they apply with particular force in this case.

There is another problem that may be briefly mentioned here. This relates to the domain of applicability of the Rawlsian theory of justice. In the original presentation (Rawls, 1958; 1971), ‘justice as fairness’ did appear to be a theory with a very wide domain, applicable in many diverse social circumstances, with a universalist outlook. Without formally contradicting anything presented in that earlier version, Rawls's more recent (p.266) presentations (Rawls, 1985; 1987; 1988a; 1988b; 1993) have increasingly stressed some special features of Western liberal democracies as preconditions for applying the principles of justice.

Rawls has emphasized that his ‘political conception’ of justice requires tolerance and acceptance of pluralism. These are certainly attractive features of social organization. If these were parts of the requirement imposed by Rawls's theory, without making it illegitimate to apply other parts of his principles of justice even when these conditions were not entirely met, the domain of his theory would not have been substantially reduced, even though its demands would have been significantly expanded. However, Rawls has sometimes asserted precisely that conditionality—making the requirements take a fairly ‘all or nothing’ form. This has the immediate effect of making it an illegitimate use of his theory to apply his principles of justice in circumstances where the conditions of tolerance are not met.

In the context of many ‘Third World’ countries in which the problems of gender inequality are particularly acute, Rawls's requirements of toleration are not at all well met. If, as a result, it becomes right to conclude (as seems to be suggested by Rawls) that his theory cannot be applied in such societies, then there is not a great deal to be said about gender inequality in those circumstances with the aid of ‘justice as fairness’.

I personally would argue that Rawls over‐restricts the domain of his theory, since it has usefulness beyond these limits.13 The theory comes into its own in the fuller context of toleration that make Rawls's ‘political conception’ more extensively realizable, but the important questions of liberty, equity, and efficiency outlined by Rawls have substantial bearings even in those circumstances in which the demands of toleration are not universally accepted.

5 Freedoms, Capabilities, and Justice

I have argued elsewhere in favour of judging individual advantage directly in terms of the freedom to achieve, rather than in terms of primary goods (as in Rawls, 1971), incomes (as in standard welfare‐economic discussions), resources (as in Dworkin, 1981), and other proposed spaces. The ‘capability perspective’ involves concentration on freedoms to achieve in general and the capabilities to function in particular (especially when assessing freedoms to pursue well‐being).14 Individual achievements in living could be seen in terms of human functionings, consisting of various beings and doings, varying from such elementary matters as being adequately nourished, (p.267) avoiding escapable morbidity, etc., to such complex functionings as taking part in the life of the community, achieving self‐respect, and so on.

An important part of our freedom to achieve consists of our capability to function. In the functioning space an achievement is an n‐tuple of functionings that are realized, whereas a capability set is a collection of such n‐tuples of functioning combinations. The capability set of a person represents the alternative combinations of functioning achievements from which the person can choose one combination. It is, thus, a representation of the freedom that a person enjoys in choosing one mode of living or another.15

When we want to examine a person's freedom to achieve in a more general context (including the achievement of social objectives), we shall have to go beyond the functioning space into the corresponding representations of broader achievements, e.g., promoting her social objectives such as reforming some feature or another of the society in which she lives. By pointing our attention towards freedoms in general, the capability approach is meant to accept the relevance of freedom over this broader space, even though the formal definition of capabilities may not take us beyond human functionings as such.16

A number of questions have been raised about the cogency, scope, and applicability of the capability approach to justice. I have dealt with some of the issues elsewhere (Sen, 1992a; 1992) and will not go into them here.17 There are also interesting issues in the relationship between this approach and the perspective emerging from Aristotelian analysis of capability, virtues, and justice, and these have been illuminatingly discussed by Martha Nussbaum (1988a; 1988b). These issues too I shall not pursue here. Instead I shall try to comment on some particular features of this approach that may be particularly relevant in developing a capability‐based theory of justice in general, and can be usefully applied specifically to analyse gender inequality.

I would argue that any theory of justice (1) identifies a space in which inter‐personal comparisons are made for judging individual advantages, and (2) specifies a ‘combining’ procedure that translates the demands of justice to operations on the chosen space. For example, the utilitarian approach identifies the relevant space as that of individual utilities (defined as pleasures, fulfilment of desires, or some other interpretation), and picks the combining formula of simply adding up the individual utilities to arrive (p.268) at a sum total that is to be maximised. To take another example, Nozick's (1974) ‘entitlement theory’ specifies the space as a set of libertarian rights that individuals can have, and uses as a combining formula an equal holding of these rights. Similarly, the Rawlsian approach demands maximal equal liberty for all in the space of some specified liberties (through the ‘First Principle’) and supplements it by demanding a lexicographic maximin rule in the space of holdings of primary goods (included in the ‘Second Principle’ in the form of the ‘Difference Principle’).

It should be obvious that the specification of the space of functionings and capabilities in particular, and of achievements and freedoms in general, does not amount to a theory of justice. It merely identifies the field in which the ‘combining’ operations have to be defined. The assertiveness of the claim rests on the acceptance of the peculiar relevance of this space in judging individual advantage in formulating a theory of justice.

I have argued elsewhere that a theory of justice must include aggregative considerations as well as distributive ones.18 It will be a mistake to see the space of functionings and capabilities as being exclusively related to specifications of the demands of equality. In assessing the justice of different distributions of individual capabilities and freedoms, it would be appropriate to be concerned both about aggregative considerations and about the extent of inequality in the distribution pattern.

It is not my purpose here to argue for a particular formula for combining the diverse considerations of equality and efficiency, and I am not about to propose a rival specification to the lexicographic maximin rule used by Rawls, or to the simple summation rule used by the utilitarians. There are good grounds for attaching importance both to overall generation of capabilities (this includes aggregative considerations in general and efficiency considerations in particular) as well as to reducing inequalities in the distribution of capabilities. Within that general agreement various formulae can be found that do not coincide with each other but which can be—and have been—defended in a reasonable way in many presentations. I have not gone beyond outlining a space and some general features of a combining formula, and this obviously falls far short of being a complete theory of justice. Such a complete theory is not what I am seeking, and more importantly for the present purpose, it is not especially needed to analyse gender inequality. The class of theories of justice that are consistent with these requirements is adequate for the present purpose.

6 Gender and Justice

Earlier in this paper I have tried to outline the connection between common perceptions of legitimacy and appropriateness (shared even by women (p.269) themselves) in traditional societies and the gender inequalities that are generally accepted in those societies (even by the women themselves). In that context I illustrated the inequalities with some standard indicators of minimal success in living, such as survival rates. This was just one illustration of the kind of variable in terms of which inequalities can be assessed. Being able to survive without premature mortality is, of course, a very basic capability. When a fuller accounting is done, many other capabilities would have obvious relevance, varying from the ability to avoid preventable morbidity, to be well‐nourished, to be comfortable and happy, etc., on the one hand, as well as more complex freedoms to achieve, including social goals and objectives, on the other.

This way of judging individual advantage provides an immediate connection between (1) the basis of the class of theories of justice outlined in the previous section, and (2) the empirical realities in terms of which gender inequality can be effectively discussed. The main advantage in being concerned with this space rather than the space of resources, primary goods, incomes, etc., is that the perspective of freedom to achieve tells us a great deal more about the advantages that the persons actually enjoy to pursue their objectives (as opposed to the means they possess that may differentially privilege different people to promote their aims).

It has been suggested by Rawls (1988b), in a critique of my line of reasoning, that comparing people's capabilities would require the use of one universal set of ‘comprehensive’ objectives shared by all, and that demanding such uniformity would be a mistake. I agree that it would be a mistake to demand such uniformity, but is it really needed?

People do, of course, have different particular aims. Whether at a deep and sophisticated level a shared set of general objectives can be fruitfully assumed is an important question that has been addressed in the Aristotelian perspective by Martha Nussbaum (1988b).19 But no matter what position we take on that particular question, it is important to recognize that inter‐personal comparison of capabilities are not rendered impossible by the absence of an agreed ‘comprehensive doctrine’. By looking at ‘intersections’ between different individual orderings, agreed judgements on capabilities can be made without invoking a single ‘comprehensive’ doctrine shared by all.20 There can be incompletenesses in such orderings but that is a problem that applies to the indexing of primary goods as well.21 The really serious cases of inequities that tend to move us towards agitating for social change would typically be captured by a variety of orderings, even when they would disagree with each other in many subtler issues.

The specification of the relevant space opens the way not only for the assessment of inequalities in those terms but also for understanding the (p.270) demands of efficiency in that context. This is particularly important in understanding gender inequality for two distinct reasons.

First, as was argued earlier, gender relations do involve ‘co‐operative conflicts’. There are benefits for all through co‐operation, but the availability of many different arrangements (yielding different levels of inequality in the generated capabilities) superimpose conflicts on a general background of co‐operative gains. To deny the existence of the efficiency problem would be a great mistake, and cannot serve the cause of gender equality in a practical world. Efficiency issues have to be tackled along with problems of inequality and injustice.

Secondly, gender inequality is made acceptable to women themselves (along with the more powerful male members of the society at large) by playing up the demands of efficiency in particular social arrangements. The relatively inferior role of women and the shockingly neglected treatment of young girls are implicitly ‘justified’ by alleged efficiency considerations. The alternative of chaos and gross inefficiency is frequently presented, explicitly or by implication, in discussions on this subject. That line of argument has to be critically scrutinized and challenged.

To meet that general presumption and prejudice, what is needed is a serious analysis of the feasibility of alternative arrangements that can be less iniquitous but no less efficient. To some extent such an analysis can draw on what has already been achieved in other countries. In the light of specific circumstances, more particular analysis of feasibilities may also be needed.22 The identification of deprivation has to be linked directly to the demands of fair division.

The central issue is to confront the underlying prejudice directly and to outline the need for and scope of reducing inequalities in capabilities without accepting that this must cause great inefficiency. The implicit prejudices call for explicit scrutiny. We have to be clear on the nature of the ‘theory’ underlying the practice of extreme inequality, and be prepared to outline what justice may minimally demand. The advantage of a theory of justice defined in terms of the capability space is to place the debate where it securely belongs.


Bibliography references:

Anand, S., and Ravallion, M. (1993). ‘Human Development in Poor Countries: On the Role of Private Incomes and Public Services’, Journal of Economic Perspectives, 7.

(p.271) Becker, G. S. (1981). A Treatise on the Family. Cambridge, Mass.: Harvard University Press.

Blair, D. H. (1988). ‘The Primary‐Goods Indexation Problem in Rawls' Theory of Justice’, Theory and Decision, 24.

Bentham, J. (1789). An Introduction to the Principles of Morals and Legislation. (London: Payne), republished Oxford: Clarendon Press, 1907.

Boserup, E. (1970). Women's Role in Economic Development. London: Allen and Unwin.

—— (1990). ‘Economic Change and the Roles of Women’, in Tinker (1990).

Chen, L., Huq, E., and D'Souza, S. (1981). ‘Sex Bias in the Family Allocation of Food and Healthcare in Rural Bangladesh’, Population and Development Review, 7.

Chen, M. (1992). ‘A Matter of Survival: Women's Right to Work in India and Bangladesh’, in this volume.

Coale, A. J. (1991). ‘Excess Female Mortality and the Balance of the Sexes in the Population: An Estimate of the Number of “Missing Females” ’, Population and Development Review, 17.

Crocker, D. A. (1991a). ‘Toward Development Ethics’, World Development, 19.

—— (1991b). ‘Functioning and Capability: The Foundations of Sen's Development Ethics’, IDEA Montclair Workshop, mimeographed, Colorado State University.

Drèze, J., and Sen, A. (1989). Hunger and Public Action. Oxford: Clarendon Press.

Dworkin, R. (1981). ‘What is Equality? Part 1: Equality of Welfare’, and ‘What is Equality? Part 2: Equality of Resources’, Philosophy and Public Affairs, 10.

Gibbard, A. (1979). ‘Disparate Goods and Rawls's Difference Principle: A Social Choice Theoretic Treatment’, Theory and Decision, 11.

Griffin, J. (1982). ‘Modern Utilitarianism’, Revue Internationale de Philosophie, 36.

—— (1986). Well‐being. Oxford: Clarendon Press.

Griffin, K., and Knight, J. (eds.) (1989). ‘Human Development in the 1980s and Beyond’, Journal of Development Planning, 19 (special number).

Harriss, B., and Watson, E. (1987). ‘The Sex Ratio in South Asia’, in J. H. Momson and J. Townsend (eds.), Geography of Gender in the Third World. London: Butler and Tanner.

Hart, H. L. A. (1973). ‘Rawls on Liberty and Its Priority’, University of Chicago Law Review, 40.

Kant, I. (1785). Fundamental Principles of Metaphysics of Ethics. English trans. T. K. Abbott. London: Longman, 1907.

Klasen, S. (1994). ‘Missing Women Reconsidered’, World Development, forthcoming.

Kynch, J., and Sen, A. K. (1983). ‘Indian Women: Well‐Being and Survival’, Cambridge Journal of Economics, 7.

Laden, T. (1991). ‘Freedom, Preference and Objectivity: Women and the Capability Approach,’ mimeographed, Harvard University.

Lundberg, S., and Pollak, R. A. (1994). ‘Noncooperative Bargaining Models of Marriage’, American Economic Review, 84.

Manser, M., and Brown, M. (1980). ‘Marriage and Household Decision Making: A Bargaining Analysis’, International Economic Review, 21.

(p.272) Nash, J. F. (1950). ‘The Bargaining Problem’, Econometrica, 18.

Nozick, R. (1974). Anarchy, State and Utopia. Oxford: Blackwell.

Nussbaum, M. C. (1988a). ‘Nature, Function, and Capability; Aristotle on Political Distribution’, Oxford Studies in Ancient Philosophy, suppl. vol.

—— (1988b). ‘Non‐Relative Virtues: An Aristotelian Approach’, Midwest Studies in Philosophy, 13; revised version in Nussbaum and Sen (1993).

—— and Sen, A. K. (eds.) (1993). The Quality of Life. Oxford: Clarendon Press.

Okin, S. M. (1987). ‘Justice and Gender’, Philosophy and Public Affairs, 16.

—— (1989). Justice, Gender and Family. New York: Basic Books.

Plott, C. (1978). ‘Rawls' Theory of Justice: An Impossibility Result’, in H. Gottinger and W. Leinfellner (eds.), Decision Theory and Social Ethics. Dordrecht: Reidel.

Putnam, H. (1987). The Many Faces of Realism. La Salle: Open Court.

—— (1991). ‘Objectivity and the Science–Ethics Distinction’, in Nussbaum and Sen (1993).

Putnam, R. A. (1992). ‘Why Not a Feminist Theory of Justice?’, in this volume.

Rawls, J. (1958). ‘Justice as Fairness’, Philosophical Review, 67.

—— (1971). A Theory of Justice. Cambridge, Mass.: Harvard University Press.

—— (1985). ‘Justice as Fairness: Political not Metaphysical’, Philosophy and Public Affairs, 14.

—— (1987). ‘The Idea of an Overlapping Consensus’, Oxford Journal of Legal Studies, 7.

—— (1988a). ‘Priority of Right and Ideas of the Good’, Philosophy and Public Affairs, 17.

—— (1988b). ‘Reply to Sen’, mimeographed, Harvard University.

—— (1993). Political Liberalism. New York: Columbia University Press.

—— et al. (1987). Liberty, Equality and Law, S. McMurrin, ed. Cambridge, and Salt Lake City: Cambridge University Press, and University of Utah Press.

Sen, A. K. (1970). Collective Choice and Social Welfare. San Francisco: Holden‐Day. Republished Amsterdam: North‐Holland, 1979.

—— (1980). ‘Equality of What?’, in S. McMurrin (ed.), Tanner Lectures on Human Values, i. Cambridge: Cambridge University Press;

repr. in Sen, Choice, Welfare and Measurement. Oxford, and Cambridge, Mass.: Blackwell, and MIT Press (1982); and in Rawls et al. (1987).

—— (1985a). ‘Well‐being, Agency and Freedom: The Dewey Lectures 1984’, Journal of Philosophy, 82.

—— (1985b). Commodities and Capabilities. Amsterdam: North‐Holland.

—— (1989). ‘Women's Survival as a Development Problem’, Bulletin of the American Academy of Arts and Sciences, 43; shortened version in The New York Review of Books, Christmas, 1990.

—— (1990a). ‘Welfare, Freedom and Social Choice: A Reply’, Recherches Economiques de Louvain, 56.

—— (1990b). ‘Justice: Means versus Freedoms’, Philosophy and Public Affairs, 19.

—— (1990c). ‘Gender and Cooperative Conflicts’, in Tinker (1990).

—— (1991a). ‘Welfare, Preference and Freedom’, Journal of Econometrics, 50.

—— (1991b). ‘On Indexing Primary Goods and Capabilities’, mimeographed, Harvard University.

(p.273) —— (1992a). Inequality Reexamined. Oxford, and Cambridge, Mass.: Clarendon Press, and Harvard University Press.

—— (1992b). ‘Missing Women’, British Medical journal, 304.

—— (1993). ‘Well‐being and Capability’, in Nussbaum and Sen (1993).

Tinker, I. (1990). Persistent Inequalities. New York: Oxford University Press.


(1) A revised version of a paper presented at the WIDER conference on Human Capabilities: Women, Men and Equality, in Helsinki, August 1991. In revising the paper. I have benefited from the comments of David Crocker, Jonathan Glover, Martha Nussbaum, and Ruth‐Anna Putnam.

(2) See Sen (1989; 1992b). See also Coale (1991); Klasen (1994) for other bases of estimates, and Harriss and Watson (1987) for a general discussion of the underlying issues.

(3) I have tried to discuss the available evidence in Sen (1990c); and also in my joint work with Jean Drèze; Drèze and Sen (1989), ch. 4. See also Boserup (1970); Lincoln Chen et al. (1981); Kynch and Sen (1983); Sen (1985b).

(4) Indeed, sometimes even social analysts tend to treat the absence of any perceived sense of unjust inequality as ‘proof’ that any suggestion of real conflict is mistaken—‘an import of foreign ideas into the harmony of traditional rural living’. For a critique of this tradition of interpretation, see Kynch and Sen (1983) and Sen (1990c).

(5) For a characterization and analysis of ‘co‐operative conflicts’, see Sen (1990c). This is an extension of what Nash (1950) called ‘the bargaining problem’.

(6) In this paper I am concerned specifically with the situation in the ‘Third World’, but I believe that the problem of gender inequality even in the economically advanced countries of Europe and North America can be better understood by bringing in conceptions of justice and legitimacy as determinants of individual behaviour.

(7) It is sometimes thought that the ‘desire‐fulfilment’ theory of utility is radically different from a ‘mental metric’ approach, since it examines the extent of fulfilment of what is desired, and the objects of desire are not themselves mental magnitudes: for this and related arguments see Griffin (1982; 1986). But the utilitarian formula requires interpersonally comparable cardinal utilities, and this demands comparisons of intensities of desires for different objects, by different people. Thus, in effect, the dependence on mental metrics is extensive also in the desire‐fulfilment formulation of utilitarian calculus.

(8) On this see Kynch and Sen (1983). It is, of course, a different issue as to how these ‘objective’ matters relate to human perceptions generally (including those of professional doctors), and I am not addressing here the foundational question of objective–subjective divisions. On that issue, see Hilary Putnam (1987; 1991).

(9) Relativism raises many different types of issues. There are questions of cultural relativism, which are sometimes invoked to dispute criticisms of traditional societies. There is also the question of a separate ‘feminist’ approach to justice. These is, in that context, the methodological problem as to whether the advantages of men and women in a theory of justice can be judged in the ‘same’ standards. On these matters and also on their bearing on theories of justice, see Okin (1987; 1989), Nussbaum (1988a; 1988b) and Ruth Anna Putnam (1992).

(10) I am referring particularly to the use of ‘the original position’ in Rawls (1958; 1971). See also Rawls (1985; 1993). In his later presentations Rawls has integrated the reasoning based on ‘the original position’ with a constructivist programme inspired by Kant (1785).

(11) Equality is valued in Rawls's first principle (demanding ‘equal liberty’) as well as the second (of which the Difference Principle particularly brings out the concern with the worse off members of the society). The special concern with liberty, which is a part of the first principle, is also an attractive feature of justice, even though the lexicographic priority that liberty gets over other human concerns can be disputed. On this see Hart (1973).

(12) I have discussed this issue in Sen (1980; 1990b; 1992a).

(13) On related matters see Putnam (1992).

(14) On this see Sen (1980; 1985a; 1985b; 1993). For an excellent review of discussions relating to this perspective, see Crocker (1991b). See also Griffin and Knight (1989), Crocker (1991a), and Anand and Ravallion (1993).

(15) On some technical issues in evaluating freedom, see Sen (1990a; 1991a; 1992a). It is important to emphasize that the freedom to choose from alternative actions has to be seen not just in terms of permissible possibilities, but with adequate note of the psychological constraints that may make a person (e.g., a housewife in a traditional family) desist from taking steps that she could, in principle, freely take. On this and related issues, see Laden (1991).

(16) A distinction made between ‘agency objectives’ in general and ‘well‐being objectives’ in particular is relevant here. The capability to function is closely related to well‐being objectives but the approach (of which this outlook is a part) encourages us to look beyond this space when we are concerned with a person's ‘agency freedoms’ (see Sen, 1985a).

(17) See also Crocker (1991b).

(18) This is discussed particularly in Sen (1992a).

(19) See also Crocker (1991b).

(20) On this see Sen (1970; 1990b; 1992a).

(21) On that problem see Plott (1978); Gibbard (1979); Blair (1988) and Sen (1991b).

(22) One of the most important fields of investigation in this context is the role of the freedom to accept remunerative employment on the part of women. On this see Sen (1990c) and Martha Chen (1992).