Capability and Well‐Being
Capability and Well‐Being
Abstract and Keywords
Capability is not an awfully attractive word. It has a technocratic sound, and to some it might even suggest the image of nuclear war strategists rubbing their hands in pleasure over some contingent plan of heroic barbarity. The term is not much redeemed by the historical Capability Brown praising particular pieces of ...
Capability is not an awfully attractive word. It has a technocratic sound, and to some it might even suggest the image of nuclear war strategists rubbing their hands in pleasure over some contingent plan of heroic barbarity. The term is not much redeemed by the historical Capability Brown praising particular pieces of land–not human beings–on the solid real‐estate ground that they ‘had capabilities’. Perhaps a nicer word could have been chosen when some years ago I tried to explore a particular approach to well‐being and advantage in terms of a person's ability to do valuable acts or reach valuable states of being.1 The expression was picked to represent the alternative combinations of things a person is able to do or be–the various ‘functionings’ he or she can achieve.2
The capability approach to a person's advantage is concerned with evaluating it in terms of his or her actual ability to achieve various valuable functionings as a part of living. The corresponding approach to social advantage–for aggregative appraisal as well as for the choice of institutions and policy–takes the sets of individual capabilities as constituting an indispensable and central part of the relevant informational base of such evaluation. It differs from other approaches using other informational focuses, for example, personal utility (focusing on pleasures, happiness, or desire fulfilment), absolute or relative opulence (focusing on commodity bundles, real income, or real wealth), assessments of negative freedoms (focusing on procedural fulfilment of libertarian rights and rules of non‐interference), comparisons of means of freedom (e.g. focusing on the holdings of ‘primary goods’, as in the Rawlsian theory of justice), and comparisons of resource holdings as a basis of just equality (e.g. as in Dworkin's criterion of ‘equality of resources’).
(p.31) Different aspects of the capability approach have been discussed, extended, used, or criticized by several authors, and as a result the advantages and difficulties of the approach have become more transparent.3 There is, however, a need for a clearer and more connected account of the whole approach, particularly in view of some interpretational problems that have arisen in its assessment and use. This paper is an attempt at a clarificatory analysis at an elementary level. I shall also try to respond briefly to some interesting criticisms that have been made.
2 Functionings, Capability, and Values
Perhaps the most primitive notion in this approach concerns ‘functionings’. Functionings represent parts of the state of a person–in particular the various things that he or she manages to do or be in leading a life. The capability of a person reflects the alternative combinations of functionings the person can achieve, and from which he or she can choose one collection.4 The approach is based on a view of living as a combination of various ‘doings and beings’, with quality of life to be assessed in terms of the capability to achieve valuable functionings.
Some functionings are very elementary, such as being adequately nourished, being in good health, etc., and these may be strongly valued by all, for obvious reasons. Others may be more complex, but still widely valued, such as achieving self‐respect or being socially integrated. Individuals may, however, differ a good deal from each other in the weights they attach to these different functionings–valuable though they may all be–and the assessment of individual and social advantages must be alive to these variations.
In the context of some types of social analysis, for example, in dealing with extreme poverty in developing economies, we may be able to go a fairly long distance with a relatively small number of centrally important functionings and the corresponding basic capabilities (e.g. the ability to be well nourished and well sheltered, the capability of escaping avoidable morbidity and premature mortality, and so forth). In other contexts, including more general problems of economic development, the list may have to be much longer and much more diverse.
(p.32) Choices have to be faced in the delineation of the relevant functionings. The format always permits additional ‘achievements’ to be defined and included. Many functionings are of no great interest to the person (e.g. using a particular washing powder–much like other washing powders).5 There is no escape from the problem of evaluation in selecting a class of functionings in the description and appraisal of capabilities. The focus has to be related to the underlying concerns and values, in terms of which some definable functionings may be important and others quite trivial and negligible. The need for selection and discrimination is neither an embarrassment, nor a unique difficulty, for the conceptualization of functioning and capability.
3 Value‐Objects and Evaluative Spaces
In an evaluative exercise, we can distinguish between two different questions: (1) What are the objects of value? (2) How valuable are the respective objects? Even though formally the former question is an elementary aspect of the latter (in the sense that the objects of value are those that have positive weights), nevertheless the identification of the objects of value is substantively the primary exercise which makes it possible to pursue the second question.
Furthermore, the very identification of the set of value‐objects, with positive weights, itself precipitates a ‘dominance ranking’ (x is at least as high as y if it yields at least as much of each of the valued objects). This dominance ranking, which can be shown to have standard regularity properties such as transitivity, can indeed take us some distance–often quite a long distance–in the evaluative exercise.6
The identification of the objects of value specifies what may be called an evaluative space. In standard utilitarian analysis, for example, the evaluative space consists of the individual utilities (defined in the usual terms of pleasures, happiness, or desire fulfilment). Indeed, a complete evaluative approach entails a class of ‘informational constraints’ in the form of ruling out directly evaluative use of various types of information, to wit, those that do not belong to the evaluative space.7
The capability approach is concerned primarily with the identification of value‐objects, and sees the evaluative space in terms of functionings and capabilities to function. This is, of course, itself a deeply evaluative exercise, but (p.33) answering question (1), on the identification of the objects of value, does not, on its own, yield a particular answer to question (2), regarding their relative values. The latter calls for a further evaluative exercise. Various substantive ways of evaluating functionings and capabilities can all belong to the general capability approach.
The selection of the evaluative space has a good deal of cutting power on its own, both because of what it includes as potentially valuable and because of what it excludes. For example, because of the nature of the evaluative space, the capability approach differs from utilitarian evaluation (more generally ‘welfarist’ evaluation8) in making room for a variety of human acts and states as important in themselves (not just because they may produce utility, nor just to the extent that they yield utility).9 It also makes room for valuing various freedoms–in the form of capabilities. On the other side, the approach does not attach direct–as opposed to derivative–importance to the means of living or means of freedom (e.g. real income, wealth, opulence, primary goods, or resources), as some other approaches do. These variables are not part of the evaluative space, though they can indirectly influence the evaluation through their effects on variables included in that space.
4 Capability and Freedom
The freedom to lead different types of life is reflected in the person's capability set. The capability of a person depends on a variety of factors, including personal characteristics and social arrangements. A full accounting of individual freedom must, of course, go beyond the capabilities of personal living and pay attention to the person's other objectives (e.g. social goals not directly related to one's own life), but human capabilities constitute an important part of individual freedom.
Freedom, of course, is not an unproblematic concept. For example, if we do not have the courage to choose to live in a particular way, even though we could live that way if we so chose, can it be said that we do have the freedom to live that way, i.e. the corresponding capability? It is not my purpose here to brush under the carpet difficult questions of this–and other–types. In so far as there are genuine ambiguities in the concept of freedom, that should be reflected in corresponding ambiguities in the characterization of capability. This relates to a methodological point, which I have tried to defend elsewhere, that if an underlying idea has an essential ambiguity, a precise formulation of that idea (p.34) must try to capture that ambiguity rather than hide or eliminate it.10
Comparisons of freedom raise interesting issues of evaluation. The claim is sometimes made that freedom must be valued independently of the values and preferences of the person whose freedom is being assessed, since it concerns the ‘range’ of choice a person has–not how she values the elements in that range or what she chooses from it. I do not believe for an instant that this claim is sustainable (despite some superficial plausibility), but had it been correct, it would have been a rather momentous conclusion, driving a wedge between the evaluation of achievements and that of freedoms. It would, in particular, be then possible to assess the freedom of a person independently of–or prior to–the assessment of the alternatives between which the person can choose.11.
How can we judge the goodness of a ‘range’ of choice independently of–or prior to–considering the nature of the alternatives that constitute that range? Some comparisons can, of course, be made in terms of set inclusion, for example, that reducing the ‘menu’ from which one can choose will not increase one's freedom.12. But whenever neither set is entirely included in the other, we have to go beyond such ‘subset reasoning’.
One alternative is simply to count the number of elements in the set as reflecting the value of the range of choice.13 But this number‐counting procedure leads to a rather peculiar accounting of freedom. It is odd to conclude that the freedom of a person is no less when she has to choose between three alternatives which she sees respectively as ‘bad’, ‘awful’, and ‘gruesome’ than when she has the choice between three alternatives which she assesses as ‘good’, ‘excellent’, and (p.35) ‘superb’.14 Further, it is always possible to add trivially to the number of options one has (e.g. tearing one's hair, cutting one's ears, slicing one's toes, or jumping through the window), and it would be amazing to see such additions as compensating for the loss of really valued options.15 The assessment of the elements in a range of choice has to be linked to the evaluation of the freedom to choose among that range.16
5 Value‐Purposes and Distinct Exercises
While the identification of value‐objects and the specification of an evaluative space involve norms, the nature of the norms must depend on precisely what the purpose of the evaluation is. Assessing well‐being may take us in one direction; judging achievement in terms of the person's overall goals may take us in a somewhat different direction, since a person can have objectives other than the pursuit of his or her own well‐being. Judging achievement of either kind may also differ from the evaluation of the freedom to achieve, since a person can be advantaged in having more freedom and still end up achieving less.
We can make a fourfold classification of points of evaluative interest in assessing human advantage, based on two different distinctions. One distinction is between (1.1) the promotion of the person's well‐being, and (1.2) the pursuit of the person's overall agency goals. The latter encompasses the goals that a person has reasons to adopt, which can inter alia include goals other than the advancement of his or her own well‐being. It can thus generate orderings different from that of well‐being. The second distinction is between (2.1) achievement, and (2.2) the freedom to achieve. This contrast can be applied both to the perspective of well‐being and to that of agency. The two distinctions together yield four different concepts of advantage, related to a person: (1) ‘well‐being achievement’, (2) ‘agency achievement’, (3) ‘well‐being freedom’, and (4) ‘agency freedom’. These different notions, which I have tried to discuss more extensively elsewhere, are not, of course, unrelated to each other, but nor are they necessarily identical.17
(p.36) The assessment of each of these four types of benefit involves an evaluative exercise, but they are not the same evaluative exercise. They can also have very disparate bearings on matters to which the evaluation and comparison of individual advantages are relevant. For example, in determining whether a person is deprived in a way that calls for assistance from others or from the state, a person's well‐being may be, arguably, more relevant than his agency success (e.g. the state may have better grounds for offering support to a person for overcoming hunger or illness than for helping him to build a monument to his hero, even if he himself attaches more importance to the monument than to the removal of his hunger or illness). Furthermore, for adult citizens, well‐being freedom may be more relevant to state policy, in this context, than well‐being achievement (e.g. the state may have reason to offer a person adequate opportunities to overcome hunger, but not to insist that he must take up that offer and cease to be hungry). Interpersonal comparisons can be of many distinct types, with possibly dissimilar evaluative interests. Despite the interdependences between the different valuepurposes, they can generate quite distinct exercises with partly divergent concentration and relevance.
6 Well‐Being, Agency, and Living Standards
The well‐being achievement of a person can be seen as an evaluation of the ‘wellness’ of the person's state of being (rather than, say, the goodness of her contribution to the country, or her success in achieving her overall goals). The exercise, then, is that of assessing the constituent elements of the person's being seen from the perspective of her own personal welfare. The different functionings of the person will make up these constituent elements.
This does not, of course, imply that a person's well‐being cannot be ‘other‐regarding’. Rather, the effect of ‘other‐regarding’ concerns on one's well‐being has to operate through some feature of the person's own being. Doing good may make a person contented or fulfilled, and these are functioning achievements of importance. In this approach, functionings are seen as central to the nature of well‐being, even though the sources of well‐being could easily be external to the person.
The functionings relevant for well‐being vary from such elementary ones as escaping morbidity and mortality, being adequately nourished, having mobility, etc., to complex ones such as being happy, achieving self‐respect, taking part in (p.37) the life of the community, appearing in public without shame (the last a functioning that was illuminatingly discussed by Adam Smith18). The claim is that the functionings make up a person's being, and the evaluation of a person's wellbeing has to take the form of an assessment of these constituent elements.
If the value‐purpose is changed from checking the ‘well‐ness’ of the person's being to assessing the person's success in the pursuit of all the objectives that he has reason to promote, then the exercise becomes one of evaluation of ‘agency achievement’, rather than of well‐being achievement. For this exercise, the space of functionings may be rather restrictive, since the person's goals may well include other types of objective (going well beyond the person's own state of being). Also, the difference between agency achievement and well‐being achievement is not only a matter of space (the former taking us beyond the person's own life and functionings), but also one of differential weighting of the shared elements (i.e. for the functionings that are pertinent both to one's wellbeing and to one's other objectives, possibly different weights may be attached in agency evaluation vis‐à‐vis well‐being appraisal).
The assessment of agency success is a broader exercise than the evaluation of well‐being. It is also possible to consider ‘narrower’ exercises than the appraisal of well‐being. A particularly important one is that of evaluating a person's standard of living. This, too, may take the form of focusing on the person's functionings, but in this case we may have to concentrate only on those influences on well‐being that come from the nature of his own life, rather than from ‘other‐regarding’ objectives or impersonal concerns. For example, the happiness generated by a purely other‐regarding achievement (e.g. the freeing of political prisoners in distant countries) may enhance the person's well‐being without, in any obvious sense, raising his living standard.
In the ethical context, the explicit recognition that one's well‐being may often be affected by the nature of other people's lives is not, of course, new. Even Emperor Asoka, in the third century BC, noted the distinction clearly in one of his famous ‘rock edicts’ in the process of defining what should count as an injury to a person: ‘And, if misfortune befalls the friends, acquaintances, companions and relations of persons who are full of affection [towards the former], even though they are themselves well provided for, [this misfortune] is also an injury to their own selves.’19 The inability to be happy, which will be widely recognized as a failure of an important functioning (even though not the only important one, except in the hedonist version of utilitarianism), may arise either from sources within one's own life (e.g. being ill, or undernourished, or otherwise deprived), or from sources outside it (e.g. the pain that comes from sympathizing with others' misery). While both types of factor affect one's well‐being, the case (p.38) for excluding the latter from the assessment, specifically, of one's living standards would seem fairly reasonable, since the latter relates primarily to the lives of others, rather than one's own.20
7 Why Capability, not Just Achievement?
The preceding discussion on the achievement of well‐being and living standards has been related to functionings rather than to capabilities. This was done by design to introduce distinct problems in sequence, even though eventually an integrated view will have to be taken. In fact, the capability approach, as the terminology indicates, sees the capability set as the primary informational base. Why should we have to broaden our attention from functionings to capability?
We should first note that capabilities are defined derivatively from functionings. In the space of functionings any point, representing an n‐tuple of functionings, reflects a combination of the person's doings and beings, relevant to the exercise. The capability is a set of such functioning n‐tuples, representing the various alternative combinations of beings and doings any one (combination) of which the person can choose.21 Capability is thus defined in the space of functionings. If a functioning achievement (in the form of an n‐tuple of functionings) is a point in that space, capability is a set of such points (representing the alternative functioning n‐tuples from which one n‐tuple can be chosen).
Note further that the capability set contains information about the actual functioning n‐tuple chosen, since it too is obviously among the feasible n‐tuples. The evaluation of a capability set may be based on the assessment of the particular n‐tuple chosen from that set. Evaluation according to the achieved functioning combination is thus a ‘special case’ of evaluation on the basis of the capability set as a whole. In this sense, well‐being achievement can be assessed on the basis of the capability set, even when no freedom‐type notion influences that achievement. In this case, in evaluating the capability set for the value‐purpose of assessing well‐being achievement, we would simply have to identify the value of the capability set with the value of the achieved functioning n‐tuple in it. The procedure of equating the value of the capability set to the value of one of the elements of that set has been called ‘elementary evaluation’.22
(p.39) Clearly, there is at least no informational loss in seeing well‐being evaluation in terms of capabilities, rather than directly in terms of the achieved, or chosen, or maximal functioning n‐tuple. While this indicates that the informational base of capability is at least as adequate as that of achieved functionings, the claim in favour of the capability perspective is, in fact, stronger. The advantages of the extension arise from two rather different types of consideration.
First, we may be interested not merely in examining ‘well‐being achievement’, but also ‘well‐being freedom’. A person's actual freedom to live well and be well is of some interest in social as well as personal evaluation.23 Even if we were to take the view, which will be disputed presently, that well‐being achievement depends only on the achieved functionings, the ‘well‐being freedom’ of a person will represent the freedom to enjoy the various possible well‐beings associated with the different functioning n‐tuples in the capability set.24
Second, freedom may have intrinsic importance for the person's well‐being achievement. Acting freely and being able to choose may be directly conducive to well‐being, not just because more freedom may make better alternatives available. This view is contrary to the one typically assumed in standard consumer theory, in which the contribution of a set of feasible choices is judged exclusively by the value of the best element available.25 Even the removal of all the elements of a feasible set (e.g. of a ‘budget set’) other than the chosen best element is seen, in that theory, as no real loss, since the freedom to choose does not, in this view, matter in itself.
In contrast, if choosing is seen as a part of living (and ‘doing x’ is distinguished from ‘choosing to do x and doing it’), then even ‘well‐being achievement’ need not be independent of the freedom reflected in the capability set.26 In that case, both ‘well‐being achievement’ and ‘well‐being freedom’ will have to be assessed in terms of capability sets. Both must then involve ‘set evaluation’ in a non‐elementary way (i.e. without limiting the usable informational content of capability sets through elementary evaluation).
(p.40) There are many formal problems involved in the evaluation of freedom and the relationship between freedom and achievement.27 It is, in fact, possible to characterize functionings in a ‘refined’ way to take note of the ‘counterfactual’ opportunities, so that the characteristic of relating well‐being achievement to functioning n‐tuples could be retained without losing the substantive connection of well‐being achievement to the freedom of choice enjoyed by the person. Corresponding to the functioning x, a ‘refined’ functioning (x/S) takes the form of ‘having functioning x through choosing it from the set S’.28
Sometimes even our ordinary language presents functionings in a refined way. For example, fasting is not just starving, but starving through rejecting the option of eating. The distinction is obviously important in many social contexts: we may, for example, try to eliminate involuntary hunger, but not wish to forbid fasting. The importance of seeing functionings in a refined way relates to the relevance of choice in our lives. The role of the choice involved in a capability set has been discussed above in the context of well‐being only, but similar arguments apply to the assessment of agency achievement and the standard of living.29
8 Basic Capability and Poverty
For some evaluative exercises, it may be useful to identify a subset of crucially important capabilities dealing with what have come to be known as ‘basic needs’.30 There tends to be a fair amount of agreement on the extreme urgency of a class of needs. Particular moral and political importance may well be attached to fulfilling well‐recognized, urgent claims.31
It is possible to argue that equality in the fulfilment of certain ‘basic capabilities’ provides an especially plausible approach to egalitarianism in the (p.41) presence of elementary deprivation.32 The term ‘basic capabilities’, used in Sen (1980), was intended to separate out the ability to satisfy certain crucially important functionings up to certain minimally adequate levels. The identification of minimally acceptable levels of certain basic capabilities (below which people count as being scandalously ‘deprived’) can provide a possible approach to poverty, and I shall comment on the relation of this strategy to more traditional income‐focused analyses of poverty. But it is also important to recognize that the use of the capability approach is not confined to basic capabilities only.33
Turning to poverty analysis, identifying a minimal combination of basic capabilities can be a good way of setting up the problem of diagnosing and measuring poverty. It can lead to results quite different from those obtained by concentrating on inadequacy of income as the criterion of identifying the poor.34 The conversion of income into basic capabilities may vary greatly between individuals and also between different societies, so that the ability to reach minimally acceptable levels of basic capabilities can go with varying levels of minimally adequate incomes. The income‐centred view of poverty, based on specifying an interpersonally invariant ‘poverty line’ income, may be very misleading in the identification and evaluation of poverty.
However, the point is sometimes made that poverty must, in some sense, be a matter of inadequacy of income, rather than a failure of capabilities, and this might suggest that the capability approach to poverty is ‘essentially wrong‐headed’. This objection overlooks both the motivational underpinning of poverty analysis and the close correspondence between capability failure and income inadequacy when the latter is defined taking note of parametric variations in income–capability relations.
Since income is not desired for its own sake, any income‐based notion of poverty must refer—directly or indirectly—to those basic ends which are promoted by income as means. Indeed, in poverty studies related to less developed countries, the ‘poverty line’ income is often derived explicitly with reference to nutritional norms. Once it is recognized that the relation between income and capabilities varies between communities and between people in the same community, the minimally adequate income level for reaching the same minimally acceptable capability levels will be seen as variable—depending on personal and social characteristics. However, as long as minimal capabilities can be achieved (p.42) by enhancing the income level (given the other personal and social characteristics on which capabilities depend), it will be possible (for the specified personal and social characteristics) to identify the minimally adequate income for reaching the minimally acceptable capability levels. Once this correspondence is established, it would not really matter whether poverty is defined in terms of a failure of basic capability or as a failure to have the corresponding minimally adequate income.35
Thus, the motivationally more accurate characterization of poverty as a failure of basic capabilities can also be seen in the more traditional format of an income inadequacy. The difference in formulation is unimportant. What is really important is to take note of the interpersonal and intersocial variations in the relation between incomes and capabilities. That is where the distinctive contribution of the capability approach to poverty analysis lies.
9 Midfare, Functionings, and Capability
In this paper, I have so far been primarily concerned with clarifying and integrating the basic features of the capability approach, though I have taken the opportunity to address, in passing, some criticisms that have been made of this approach. In this section and in the next, I discuss two different lines of criticism — presented respectively by G.A. Cohen and Martha Nussbaum—arguing for different ways of analysing and assessing the problems of well‐being and quality of life.
In his paper in this volume, and elsewhere (Cohen, 1989, 1990), G.A Cohen has provided a critical assessment of my writings on capability (and also of the theories of others—utilitarians, John Rawls, Ronald Dworkin, Thomas Scanlon, et al.), at the same time presenting his own answer to the question ‘equality of what?’. Cohen is generous in giving credit where he reasonably can, and his assessment is positive in many ways, but the criticisms he makes, if sustained, would indicate a major motivational confusion as well as a conceptual inadequacy underlying the capability approach as I have tried to present it.
Cohen's main thesis is that in my paper ‘Equality of What?’ (Sen, 1980), I ‘brought two distinct aspects of a person's condition under that single name [capability] and that this unnoticed duality has persisted in [my] subsequent writings’. ‘Both aspects, or dimensions of assessment, should attract egalitarian interest, but one of them is not felicitously described as “capability” ’ (p. 17). One aspect is concerned with ‘a person being able to do certain basic things’. The other is what Cohen calls ‘midfare’, because ‘it is in a certain sense midway between goods and utility’. ‘Midfare is constituted of states of the person (p.43) produced by goods, states in virtue of which utility levels take the values they do’ (p. 18).
Cohen finds the dimension of midfare important for normative understanding and he notes, rightly, that I had put some emphasis on the state of the person, distinguishing it both from the commodities that help to generate that state, and from the utilities generated by the state. ‘We must look, for example, at her nutrition level, and not just, as Rawlsians do, at her food supply, or, as welfarists do, at the utility she gets out of eating food.’ ‘But’, Cohen argues, ‘this significant and illuminating reorientation is not equivalent to focusing on a person's capability.’ ‘Capability, and exercises of capability, form only one part of the intermediate midfare state’ (p. 19); ‘midfare, the product of goods which, in turn, generates utility, is not co‐extensive with capability, and “capability” is therefore a bad name for midfare’ (p. 20).
Is the distinction correct? I believe it is. The first thing to note is that Cohen's ‘midfare’ corresponds to what I have called a person's functionings, and not to capability. The two are related, but not meant to be the same. That distinction is, in fact, a basic part of the capability approach, and there is no embarrassment in acknowledging it. The real issue lies elsewhere, to wit, whether the capability set can have any relevance in analysing well‐being, given the obvious connection between well‐being and functionings (or midfare)—a connection that Cohen finds adequate for the analysis of well‐being. This is an issue that was addressed in a less specific form earlier on in this paper (in Section 7). Cohen's preference for the perspective of midfare or functionings over that of capabilities relates to that substantive issue.
In Section 7, the relevance of the capability set for the analysis of well‐being was defended on two different grounds, namely, (1) its connection with well‐being freedom (even if well‐being achievement depends only on the achieved functioning n‐tuple), and (2) the possible importance of freedom (and thus of the capability set) for well‐being achievement itself.
The second claim is the more controversial of the two. I believe it is correct, but I should also assert that even if it were incorrect, the capability approach would still be quite untarnished. As was discussed in Section 7, assessing well‐being according to the achieved functioning n‐tuple (or midfare) is a special case of the use of the capability perspective based on ‘elementary evaluation’ (focusing only on one distinguished element—the achieved functioning n‐tuple—in the capability set). This point is obscured in Cohen's analysis by his conviction that ‘the exercise of capability’ must be a rather ‘active’ operation, and Cohen is misled by this diagnosis when he argues that he ‘cannot accept . . . the associated athleticism, which comes when Sen adds that “the central feature of well‐being is the ability to achieve valuable functionings” ’ (p. 25). Cohen gives examples (e.g. small babies being well nourished and warm as a result of the activities of their parents) that clearly show that having midfare (or enjoying functionings) need not be a particularly athletic activity. I see no reason to object to this, since athleticism was never intended, despite the fact that Cohen has obviously been (p.44) misled by my use of such words as ‘capability’ and ‘achieving’.36
But let use move now from the minimalist defence to the claim that an active exercise of freedom might well be valuable for a person's quality of life and achieved well‐being. Obviously, this consideration would be of no direct relevance in the case of babies (or the mentally disabled), who are not in a position to exercise reasoned freedom of choice (though babies can sometimes be amazingly cogent, choosy, and insistent). For people who are in a position to choose in a reasoned way and value that freedom to choose, it is hard to think that their well‐being achievement would never be affected if the freedom to choose were denied, even though the (unrefined) functioning vector (or midfare) were guaranteed by the actions of others. Even in Cohen's analysis of midfare, I should have thought that room would have to be found to see it in choice‐inclusive terms, in much the same way that the functionings can be redefined in ‘refined’ terms (as discussed in Section 7). And if this is done, that would be isomorphic to including substantive consideration of the capability set, going beyond focusing exclusively on the achieved—unrefined—functioning vector (as was also discussed in Section 7).
Freedom has many aspects. Being free to live the way one would like may be enormously helped by the choice of others, and it would be a mistake to think of achievements only in terms of active choice by oneself. A person's ability to achieve various valuable functionings may be greatly enhanced by public action and policy,37 and these expansions of capability are not unimportant for freedom for that reason. Indeed, I have argued elsewhere that ‘freedom from hunger’ or ‘being free from malaria’ need not be taken to be just rhetoric (as they are sometimes described); there is a very real sense in which the freedom to live the way one would like is enhanced by public policy that transforms epidemiological and social environments.38 But the fact that freedom has that aspect does not negate the relevance of active choice by the person herself as an important component of living freely. It is because of the presence of this element (rather than the absence of others), that the act of choosing between the elements of a capability set has a clear relevance to the quality of life and well‐being of a person.
But suppose we were to accept (wrongly, I believe) that this element of freedom really has no direct impact on the well‐being of a person. In that case, the capability perspective could still be used to relate well‐being achievement to (p.45) achieved (unrefined) functionings (or midfare) through elementary evaluation. The need to relate well‐being freedom to the capability set would also remain. That was indeed the first claim (in Section 7, pp. 38–9) in favour of the use of the capability set for analysing well‐being (in this case, well‐being freedom).
As was discussed earlier in this paper,39 for many problems of individual behaviour and social policy, well‐being freedom is a concept of relevance and importance. If achieved functionings (or midfare), defined in the ‘unrefined’ way, were all that mattered, we might be as worried about the rich person fasting as about the starving poor. If we are more concerned to eliminate the hunger of the latter, it is primarily because the former has the capability to be well nourished but chooses not to, whereas the latter lacks that capability and is forced into the state of starvation. Both may have the same midfare, but they differ in their capabilities. Capability does have importance in political and social analysis.
Motivationally, the focus on capability (in addition to achieved functionings) is, in fact, not altogether different from the concern that Cohen shows elsewhere for ‘access to advantage’. Cohen notes that in his proposal
‘advantage’ is like Sen's ‘functioning’ in its wider construal, a heterogenous collection of desirable states of the person reducible neither to his resources bundle nor to his welfare level. And while ‘access’ includes what the term normally covers, I extend its meaning under a proviso that anything which a person actually has counts as something to which he has access, no matter how he came to have it, and, hence, even if his coming to have it involved no exploitation of access in the ordinary sense (nor, therefore, any exercise of capability). If, for example, one enjoys freedom from malaria because others have destroyed the malaria‐causing insects, then, in my special sense, such freedom from malaria is something to which one has access (Cohen, p. 28),
I do not see any great difficulty in ‘extending’ the meaning of ‘access’ in this way. An ‘access’ I enjoy may not have been created by me. But exactly the same applies to freedom and capability as well. The fact that a person has the freedom to enjoy a malaria‐free life (or, to put it slightly differently, that his choice of a malaria‐free life is feasible) may be entirely due to the actions of others (e.g., medical researchers, epidemiologists, public health workers), but that does not compromise the fact that he can indeed have a malaria‐free life and has the capability (thanks largely to others) to achieve such a life.40
I don't even see that much ‘extension’ of ordinary usage is involved in such use of the terms freedom and capability (even though this is not the central issue in any case).41 Indeed, even the expression ‘freedom from malaria’, used also by Cohen, is a pointer to the fact that ordinary language takes a less narrow view of the use of the term freedom. Similarly, there is no underlying presumption that we have the capability to lead a malaria‐free life only if we have (p.46) ourselves gone around exterminating the malaria‐causing insects.
Turning to a different issue also raised by Cohen, the really interesting question is not whether ‘equality of access to advantage’ coincides with ‘capability’ in general, since capability (as was discussed earlier) is a more versatile concept and its particular characterization has to be related to the ‘evaluative purpose’ of the exercise (e.g., whether ‘agency’ or ‘well‐being’ is the focal concern in that exercise). But if advantage is seen specifically in terms of well‐being (ignoring the agency aspect), then Cohen's ‘equality of access to advantage’ would indeed be very like equality of well‐being freedom, defined in terms of evaluation of capability sets from that perspective.42
Cohen's analysis has brought out the distinctions between a number of different problems which are all addressed in the capability approach but which require separate treatment. While substantive differences may remain between his focus and mine (e.g. about the importance of choosing as a constitutive element in the quality of life), Cohen's analysis has greatly helped to pinpoint some focal issues and concerns, and the need to address them explicitly.
10 The Aristotelian Connections and Contrasts
In earlier writings I have commented on the connection of the capability approach with some arguments used by Adam Smith and Karl Marx.43 However, the most powerful conceptual connections would appear to be with the Aristotelian view of the human good. Martha Nussbaum (1988, 1990) has discussed illuminatingly the Aristotelian analysis of ‘political distribution’, and its relation to the capability approach. The Aristotelian account of the human good is explicitly linked with the necessity to ‘first ascertain the function of man’ and it then proceeds to explore ‘life in the sense of activity’.44 The basis of a fair distribution of capability to function is given a central place in the Aristotelian theory of political distribution. In interpreting Aristotle's extensive writings on ethics and politics, it is possible to note some ambiguity and indeed to find some tension between different propositions presented by him, but his recognition of the crucial importance of a person's functionings and capabilities seems to emerge clearly enough, especially in the political context of distributive arrangements.
While the Aristotelian link is undoubtedly important, it should also be noted that there are some substantial differences between the way functionings and capabilities are used in what I have been calling the capability approach and the way they are dealt with in Aristotle's own analysis. Aristotle believes, as Nussbaum (1988) notes, ‘that there is just one list of functionings (at least at a certain level of generality) that do in fact constitute human good living’ (p. 152).
The capability approach has indeed been used (for example, in Sen, 1983c, 1984) to argue that while the commodity requirements of such capabilities as ‘being able to take part in the life of the community’ or ‘being able to appear in public without shame’ vary greatly from one community to another (thereby giving the ‘poverty line’ a relativist character in the space of commodities), there is much less variation in the capabilities that are aimed at through the use of these commodities. This argument, suggesting less variability at a more intrinsic level, has clear links with Aristotle's identification of ‘non‐relative virtues’, but the Aristotelian claims of uniqueness go much further.45
Martha Nussbaum, as an Aristotelian, notes this distinction, and also points to Aristotle's robust use of an objectivist framework based on a particular reading of human nature. She suggests the following:
It seems to me, then, that Sen needs to be more radical than he has been so far in his criticism of the utilitarian accounts of well‐being, by introducing an objective normative account of human functioning and by describing a procedure of objective evaluation by which functionings can be assessed for their contribution to the good human life.46
I accept that this would indeed be a systematic way of eliminating the incompleteness of the capability approach. I certainly have no great objection to anyone going on that route. My difficulty with accepting that as the only route on which to travel arises partly from the concern that this view of human nature (with a unique list of functionings for a good human life) may be tremendously overspecified, and also from my inclination to argue about the nature and importance of the type of objectivity involved in this approach. But mostly my intransigence arises, in fact, from the consideration that the use of the capability approach as such does not require taking that route, and the deliberate incompleteness of the capability approach permits other routes to be taken which also have some plausibility. It is, in fact, the feasibility as well as the usefulness of a general approach (to be distinguished from a complete evaluative blueprint) that seems to me to provide good grounds for separating the general case for the capability approach (including, inter alia, the Aristotelian theory) from the special case for taking on exclusively this particular Aristotelian theory.
In fact, no matter whether we go the full Aristotelian way, which will also need a great deal of extension as a theory for practical evaluation, or take some other particular route, there is little doubt that the kind of general argument that Aristotle uses to motivate his approach does have a wider relevance than the defence of the particular form he gives to the nature of human good. This applies inter alia to Aristotle's rejection of opulence as a criterion of achievement (rejecting wealth and income as the standards), his analysis of eudaimonia in terms of valued activities (rather than relying on readings of mental states, as in some utilitarian procedures), and his assertion of the need to examine the (p.48) processes through which human activities are chosen (thereby pointing towards the importance of freedom as a part of living).
11 Incompleteness and Substance
The Aristotelian critique points towards a more general issue, namely, that of the ‘incompleteness’ of the capability approach–both in generating substantive judgements and in providing a comprehensive theory of valuation. Quite different specific theories of value may be consistent with the capability approach, and share the common feature of selecting value‐objects from functionings and capabilities. Further, the capability approach can be used with different methods of determining relative weights and different mechanisms for actual evaluation. The approach, if seen as a theory of algorithmic evaluation, would be clearly incomplete.47
It may well be asked: why pause at outlining a general approach, with various bits to be filled in, rather than ‘completing the task’? The motivation underlying the pause relates to the recognition that an agreement on the usability of the capability approach–an agreement on the nature of the ‘space’ of value‐objects–need not presuppose an agreement on how the valuational exercise may be completed. It is possible to disagree both on the exact grounds underlying the determination of relative weights, and on the actual relative weights chosen,48 even when there is reasoned agreement on the general nature of the value‐objects (in this case, personal functionings and capabilities). If reasoned agreement is seen as an important foundational quality central to political and social ethics,49 then the case for the pause is not so hard to understand. The fact that the capability approach is consistent and combinable with several different substantive theories need not be a source of embarrassment.
Interestingly enough, despite this incompleteness, the capability approach does have considerable ‘cutting power’. In fact, the more challenging part of the claim in favour of the capability approach lies in what it denies. It differs from the standard utility‐based approaches in not insisting that we must value only happiness (and sees, instead, the state of being happy as one among several objects of value), or only desire fulfilment (and takes, instead, desire as useful but imperfect evidence–frequently distorted–of what the person herself values).50 It differs also from other–non‐utilitarian–approaches in not placing among value‐objects primary goods as such (accepting these Rawlsian‐focus variables only derivatively and instrumentally and only to the extent that these goods promote capabilities), or resources as such (valuing this Dworkinian perspective only in terms of the impact of resources on functionings and capabilities), and so forth.51
(p.49) A general acceptance of the intrinsic relevance and centrality of the various functionings and capabilities that make up our lives does have substantial cutting power, but it need not be based on a prior agreement on the relative values of the different functionings or capabilities, or on a specific procedure for deciding on those relative values.
Indeed, it can be argued that it may be a mistake to move on relentlessly until one gets to exactly one mechanism for determining relative weights, or — to turn to a different aspect of the ‘incompleteness’ — until one arrives at exactly one interpretation of the metaphysics of value. There are substantive differences between different ethical theories at different levels, from the meta‐ethical (involving such issues as objectivity) to the motivational, and it is not obvious that for substantive political and social philosophy it is sensible to insist that all these general issues be resolved before an agreement is reached on the choice of an evaluative space. Just as the utilization of actual weights in practical exercises may be based on the acceptance of a certain range of variability of weights (as I have tried to discuss in the context of the use of the capability approach52), even the general rationale for using such an approach may be consistent with some ranges of answers to foundational questions.
12 A Concluding Remark
In this paper I have tried to discuss the main features of the capability approach to evaluation: its claims, its uses, its rationale, its problems. I have also addressed some criticisms that have been made of the approach. I shall not try to summarize the main contentions of the paper, but before concluding, I would like to emphasize the plurality of purposes for which the capability approach can have relevance.
There are different evaluative problems, related to disparate value‐purposes. Among the distinctions that are important is that between well‐being and agency, and that between achievement and freedom. The four categories of intrapersonal assessment and interpersonal comparison that follow from these two distinctions (namely, well‐being achievement, well‐being freedom, agency achievement, and agency freedom) are related to each other, but are not identical. The capability approach can be used for each of these different types of evaluation, though not with equal reach. It is particularly relevant for the assessment of well‐being — in the form of both achievement and freedom — and for the related problem of judging living standards.
As far as social judgements are concerned, the individual evaluations feed directly into social assessment. Even though the original motivation for using the capability approach was provided by an examination of the question ‘equality of what?’ (Sen, 1980), the use of the approach, if successful for equality, need not (p.50) be confined to equality only.53 The usability of the approach in egalitarian calculus depends on the plausibility of seeing individual advantages in terms of capabilities, and if that plausibility is accepted, then the same general perspective can be seen to be relevant for other types of social evaluation and aggregation.
The potentially wide relevance of the capability perspective should not come as a surprise, since the capability approach is concerned with showing the cogency of a particular space for the evaluation of individual opportunities and successes. In any social calculus in which individual advantages are constitutively important, that space is of potential significance.
Aristotle (4th c. BC). The Nicomachean Ethics: see Ross (1980).
Arneson, R. (1987). ‘Equality and Equality of Opportunity for Welfare’, Philosophical Studies, S4.
Asahi, J. (1987). ‘On Professor Sen's Capability Theory’, mimeographed, Tokyo.
Basu, K. (1987). ‘Achievements, Capabilities and the Concept of Well‐being’, Social Choice and Welfare, 4.
Beitz, C. R. (1986). ‘Amartya Sen's Resources, Values and Development’, Economics and Philosophy, 2.
Brannen, J., and Wilson, G. (eds.) (1987). Give and Take in Families. London: Allen and Unwin.
Broome, J. (1988). Review of On Ethics and Economics and The Standard of Living in London Review of Books.
Campbell, R. H., and Skinner, A. S. (eds.) (1976). Adam Smith, An Inquiry into the Nature and Causes of the Wealth of Nations. Oxford: Clarendon Press.
Cohen, G. A. (1989). ‘On the Currency of Egalitarian Justice’, Ethics, 99.
—— (1990). ‘Equality of What? On Welfare, Goods and Capabilities’, Recherches Economiques de Louvain, 56.
Culyer, A. J. (1985). ‘The Scope and Limits of Health Economics’, Okonomie des Gesundheitswesens.
Dasgupta, P. (1986). ‘Positive Freedom, Markets and the Welfare State’, Oxford Review of Economic Policy, 2.
—— (1988). ‘Lives and Well‐being’, Social Choice and Welfare, 5.
—— (1989). ‘Power and Control in the Good Polity’, in Hamlin and Pettit (1989).
d'Aspremont, C., and Gevers, L. (1977). ‘Equity and the Informational Basis of Collective Choice’, Review of Economic Studies, 46.
deBeus, Jos (1986). ‘Sen's Theory of Liberty and Institutional Vacuum’, University of Amsterdam.
deVos, K., and Hagenaars, Aldi J. M. (1988). ‘A Comparison between the Poverty Concepts of Sen and Townsend’, mimeographed, University of Leiden.
Delbono, F. (1987). Review article on Commodities and Capabilities, Economic Notes.
Drèze, J., and Sen, A. (1989). Hunger and Public Action. Oxford: Clarendon Press.
Dworkin, R. (1981). ‘What is Equality? Part 2: Equality of Resources’, Philosophy and Public Affairs, 10.
Elster, J., and Hylland, A. (eds.) (1986). Foundations of Social Choice Theory. Cambridge: Cambridge University Press.
Gaertner, W. (1988). Review of Commodities and Capabilities in Zeitschriften fur National Okommnomia, 48.
Goodin, R. E. (1985). Political Theory and Public Policy. Chicago and London: University of Chicago Press.
—— (1988). Reasons for Welfare: Political Theory of the Welfare State. Princeton: Princeton University Press.
Griffin, J. (1986). Well‐being. Oxford: Clarendon Press.
Griffin, K., and Knight, J. (eds.) (1989a). ‘Human Development in the 1980s and Beyond’, special number, Journal of Development Planning, 19.
—— (1989b). ‘Human Development: The Case for Renewed Emphasis’, Journal of Development Planning, 19.
Hamlin, A. P. (1986). Ethics, Economics and the State. Brighton: Wheatsheaf Books.
Hamlin, A., and Pettit, P. (eds.) (1989). The Good Polity: Normative Analysis of the State. Oxford: Blackwell.
Hare, R. (1981). Moral Thinking. Oxford: Clarendon Press.
Harrison, R. (ed.) (1979). Rational Action. Cambridge: Cambridge University Press.
Hawthorn, G. (1987). ‘Introduction’ in Sen (1987b).
Helm, D. (1986). ‘The Assessment: The Economic Border of the State’, Oxford Review of Economic Policy, 2.
Hossain, I. (1990). Poverty as Capability Failure. Helsinki: Swedish School of Economics and Business Administration.
Kakwani, N. (1986). Analysing Redistribution Policies. Cambridge: Cambridge University Press.
Kanbur, R. (1987). ‘The Standard of Living: Uncertainty, Inequality and Opportunity’, in Sen (1987b).
Koohi‐Kamali, F. (1988). ‘The Pattern of Female Mortality in Iran and some of its Causes’, Applied Economics Discussion Paper 62, Oxford Institute of Economics and Statistics.
Koopmans, T. C. (1964). ‘On Flexibility of Future Preference’, in M. W. Shelly and G. L. Bryan (eds.), Human Judgments and Optimality. New York: Wiley.
Kreps, D. (1979). ‘A Representation Theorem for Preference for Flexibility’, Econometrica, 47.
Kumar, B. G. (1987). Poverty and Public Policy: Government Intervention and Level in Kerala, India. D.Phil. dissertation, Oxford University.
Kynch, J., and Sen, A. (1983). ‘Indian Women: Well‐being and Survival’, Cambridge Journal of Economics, 7.
Liddell, H. G., and Scott, R. (1977). A Greek–English Lexicon, extended by H. S. Jones and R. McKenzie. Oxford: Clarendon Press.
Luker, W. (1986). ‘Welfare Economics, Positivist Idealism and Quasi‐Experimental Methodology’, mimeographed, University of Texas, Austin.
McMurrin, S. M. (ed.) (1980). Tanner Lectures on Human Values, i. Salt Lake City: University of Utah Press, and Cambridge: Cambridge University Press.
Muellbauer, J. (1987). ‘Professor Sen on the Standard of Living’, in Sen (1987b).
Nussbaum, M. (1988). ‘Nature, Function, and Capability: Aristotle on Political Distribution’, Oxford Studies in Ancient Philosophy, suppl. vol.
—— (1990). ‘Non‐Relative Virtues: An Aristotelian Approach’, Midwest Studies in Philosophy, 13; revised version in this volume.
Pattanaik, P. K., and Xu, Yongsheng (1990). ‘On Ranking Opportunity Sets in Terms of Freedom of Choice’, Recherches Economiques de Louvain, S6.
Pigou, A. C. (1920). The Economics of Welfare. London: Macmillan.
Rawls, J. (1971). A Theory of Justice. Cambridge, Mass.: Harvard University Press, and Oxford: Clarendon Press.
—— (1988a). ‘Priority of Right and Ideas of the Good’, Philosophy and Public Affairs, 17.
—— (1988b). ‘Reply to Sen’, mimeographed, Harvard University.
—— et al. (1987). Liberty, Equality, and Law: Selected Tanner Lectures on Moral Philosophy, ed. S. McMurrin. Cambridge: Cambridge University Press, and Salt Lake City: University of Utah Press.
Riley, J. (1988). Liberal Utilitarianism: Social Choice Theory and J. S. Mill's Philosophy. Cambridge: Cambridge University Press.
Ringen, J. (1987). The Possibility of Politics: A Study of the Economy of the Welfare State. Oxford: Clarendon Press.
Roemer, J. (1982). A General Theory of Exploitation and Class. Cambridge, Mass.: Harvard University Press.
—— (1986). ‘An Historical Materialist Alternative to Welfarism’, in Elster and Hylland (1986).
Ross, D. (1980). Aristotle, The Nicomachean Ethics. The World's Classics. Oxford: Oxford University Press.
Scanlon, T. M. (1979). ‘Preference and Urgency’, Journal of Philosophy, 72.
—— (1982). ‘Contractualism and Utilitarianism’, in Sen and Williams (1982).
Schokkaert, E., and van Ootegem, L. (1990). ‘Sen's Concept of the Living Standard Applied to the Belgian Unemployed’, Recherches Economiques de Louvain, S6.
Seabright, P. (1989). ‘Social Choice and Social Theories’, Philosophy and Public Affairs, 18.
Sen, A. K. (1970). Collective Choice and Social Welfare. San Francisco: Holden‐Day. Republished Amsterdam: North‐Holland, 1979.
—— (1977). ‘On Weights and Measures: Informational Constraints in Social Welfare Analysis’, Econometrica, 45.
—— (1980). ‘Equality of What?’ (1979 Tanner Lecture at Stanford), in McMurrin (1980); repr. in Sen (1982a) and Rawls et al. (1987).
—— (1982a). Choice, Welfare and Measurement. Oxford: Blackwell, and Cambridge, Mass.: MIT Press.
—— (1982b). ‘Rights and Agency’, Philosophy and Public Affairs, 11.
—— (1983a). ‘Liberty and Social Choice’, Journal of Philosophy, 80.
—— (1983b). ‘Development: Which Way Now?’, Economic Journal, 93; repr. in Sen (1984).
—— (1983c). ‘Poor, Relatively Speaking’, Oxford Economic Papers, 35; repr. in Sen (1984).
—— (1985a). ‘Well‐being, Agency and Freedom: The Dewey Lectures 1984’, Journal of Philosophy, 82.
—— (1985b). Commodities and Capabilities. Amsterdam: North‐Holland.
—— (1985c). ‘Women, Technology and Sexual Divisions’, Trade and Development, 6.
—— (1987a). On Ethics and Economics. Oxford: Blackwell.
—— (1987b). The Standard of Living (1985 Tanner Lectures at Cambridge, with contributions by Keith Hart, Ravi Kanbur, John Muellbauer, and Bernard Williams, edited by G. Hawthorn). Cambridge: Cambridge University Press.
—— (1988a). ‘Freedom of Choice: Concept and Content’, European Economic Review 32.
—— (1988b). ‘The Concept of Development’, in H. Chenery and T.N. Srinivasan (eds.), Handbook of Development Economics. Amsterdam: North‐Holland.
—— (1990a). ‘Justice: Means versus Freedoms’, Philosophy and Public Affairs, 19.
—— (1990b). ‘Gender and Cooperative Conflicts’, in Tinker (1990).
—— (1991). ‘Preference, Freedom and Social Welfare’, Journal of Econometrics, 50.
—— (1992). Inequality Reexamined. Oxford: Clarendon Press.
—— and Williams, B. (eds.) (1982). Utilitarianism and Beyond. Cambridge: Cambridge University Press.
Sircar, D. C. (1979). Asokan Studies. Calcutta: Indian Museum.
Smith, Adam (1776). An Inquiry into the Nature of Causes of the Wealth of Nations: see Campbell and Skinner (1976).
Steiner, H. (1986). ‘Putting Rights in their Place: An Appraisal of Amartya Sen's Work on Rights’, mimeographed, University of Manchester.
Stewart, F. (1988). ‘Basic Needs Strategies, Human Rights and the Right to Development’, mimeographed, Queen Elizabeth House.
Streeten, P. (1984). ‘Basic Needs: Some Unsettled Questions’, World Development, 12.
Streeten, P., et al. (1981). First Things First: Meeting Basic Needs in Developing Countries. New York: Oxford University Press.
Sugden, R. (1986). Review of Commodities and Capabilities, Economic Journal, 96.
Suppes, P. (1987). ‘Maximizing Freedom of Decision: An Axiomatic Analysis’, in G. R. Feiwel (ed.), Arrow and the Foundations of Economic Policy. New York: New York University Press.
Suzumura, K. (1988). Introduction to the Japanese translation of Commodities and Capabilities. Tokyo: Iwanami.
Tinker, I. (ed.) (1990). Persistent Inequalities. New York: Oxford University Press.
Williams, A. (1985). ‘Economics of Coronary Bypass Grafting’, British Medical Journal, 291.
—— (1991). ‘What is Health and Who Creates it?’, in J. Hutton et al. (eds.), Dependency to Enterprise. London: Routledge.
Williams, B. (1985). Ethics and the Limits of Philosophy. London: Fontana, and Cambridge, MA: Harvard University Press.
—— (1987). ‘The Standard of Living: Interests and Capabilities’, in Sen (1987b).
Wilson, G. (1987). Money in the Family. Aldershot: Avebury.
Zamagni, S. (1986). ‘Introduzione’, in A. Sen, Scelta, Benessere, Equita. Bologna: Il Mulino.
For helpful discussions, I am most grateful to G. A. Cohen, Partha Dasgupta, Jean Dreèze, Hilary Putnam, Ruth Anna Putnam, Martha Nussbaum, Derek Parfit, John Rawls, John Roemer, and Thomas Scanlon.
(1) This was in a Tanner Lecture given at Stanford University in May 1979 (‘Equality of What?’), later published as Sen (1980). The case for focusing on capability was introduced here in the specific context of evaluating inequality. I have tried to explore the possibility of using the capability perspective for analysing other social issues, such as well‐being and poverty (Sen, 1982a, 1983c, 1985b), liberty and freedom (Sen, 1983a, 1988a, 1992), living standards and development (Sen, 1983b, 1984, 1987b, 1988b), gender bias and sexual divisions (Kynch and Sen, 1983; Sen, 1985c, 1990b), and justice and social ethics (Sen, 1982b, 1985a, 1990a).
(2) Though at the time of proposing the approach, I did not manage to seize its Aristotelian connections, it is interesting to note that the Greek word dunamin, used by Aristotle to discuss an aspect of the human good, which is sometimes translated as ‘potentiality’, can be translated also as ‘capability of existing or acting’ (see Liddell and Scott, 1977: 452). The Aristotelian perspective and its connections with the recent attempts at constructing a capability‐focused approach have been illuminatingly discussed by Martha Nussbaum (1988).
(3) See the contributions of Roemer (1982, 1986), Streeten (1984), Beitz (1986), Dasgupta (1986, 1988, 1989), Hamlin (1986), Helm (1986), Zamagni (1986), Basu (1987), Brannen and Wilson (1987), Hawthorn (1987), Kanbur (1987), Kumar (1987), Muellbauer (1987), Ringen (1987), B. Williams (1987), Wilson (1987), Nussbaum (1988, 1990), Griffin and Knight (1989a, 1989b), Riley (1988), Cohen (1990), and Steiner (1990). On related matters, including application, critique, and comparison, see also de Beus (1986), Kakwani (1986), Luker (1986), Sugden (1986), Asahi (1987), Delbono (1987), Koohi‐Kamali (1987), A. Williams (1987), Broome (1988), Gaertner (1988), Stewart (1988), Suzumura (1988), de Vos and Hagennars (1988), Goodin (1985, 1988), Hamlin and Pettit (1989), Seabright (1989), Hossain (1990) and Schokkaert and van Ootegem (1990), among others.
(4) If there are n relevant functionings, then a person's extent of achievement of all of them respectively can be represented by an n‐tuple. There are several technical problems in the representation and analysis of functioning n‐tuples and capability sets, on which see Sen (1985b: chs. 2, 4, and 7).
(5) Bernard Williams (1987) raises this issue in his comments on my Tanner Lectures on the standard of living (pp. 98–101); on which see also Sen (1987b: 108–9). On the inescapable need for evaluation of different functioning and capabilities, see Sen (1985b: chs. 5–7). Just as the concentration on the commodity space in real‐income analysis does not imply that every commodity must be taken to be equally valuable (or indeed valuable at all), similarly focusing on the space of functioning does not entail that each functioning must be taken to be equally valuable (or indeed valuable at all).
(8) Welfarism requires that a state of affairs must be judged by the individual utilities in that state. It is one of the basic components of utilitarianism (the others being ‘sum‐ranking’ and ‘consequentialism’); on the factorization, see Sen (1982a) and Sen and Williams (1982).
(9) Being happy and getting what one desires may be inter alia valued in the capability approach, but unlike in utilitarian traditions, they are not seen as the measure of all values.
(10) On this, see Sen (1970, 1982a, 1987a). In many contexts, the mathematical representations should take the form of ‘partial orderings’ or ‘fuzzy’ relations. This is not, of course, a special problem with the capability approach, but applies generally to conceptual frameworks in social, economic, and political theory.
(11) The belief in this possibility seems to play a part in Robert Sugden's (1986) criticism of what he sees as my approach to capability evaluation, namely, a ‘general strategy of trying to derive the value of a set of functioning vectors from prior ranking of the vectors themselves’ (p. 821). He argues in favour of judging ‘the value of being free to choose from a range of possible lives’ before taking ‘a view on what constitutes a valuable life’. This criticism is, in fact, based on a misunderstanding of the approach proposed, since it has been a part of my claim (on which more presently) that the judgement of the quality of life and the assessment of freedom have to be done simultaneously in an integrated way, and, in particular, that ‘the quality of life a person enjoys is not merely a matter of what he or she achieves, but also of what options the person has had the opportunity to choose from’ (Sen, 1985b: 69–70). But the point at issue in the present context is the possibility of judging a range of choice independently of the value characteristics of the elements in that range. It is this possibility that I am disputing.
(12) Even this can be questioned when an expanded menu causes confusion, or the necessity to choose between a larger set of alternatives is a nuisance. But such problems can be dealt with through appropriate characterization of all the choices one has or does not have. This must include the consideration of the overall choice of having or not having to choose among a whole lot of relatively trivial alternatives (e.g. the choice of telling the telephone company to shut out mechanically dialled calls from sales agents offering a plethora of purchasing options). The issues involved in this kind of complex evaluation, incorporating choices over choices, are discussed in Sen (1992).
(14) The unacceptability of this kind of number‐counting evaluation of freedom is discussed in Sen (1985b). For an assessment of the axiomatic foundations of this and other methods of evaluation of freedom, see Sen (1991).
(15) This type of case also shows why the set‐inclusion ranking is best seen as a ‘weak’ relation of ‘no worse than’ or ‘at least as good as’, rather than as the ‘strict’ relation of ‘better than’. Adding the option of ‘slicing one's toes’ to the set of valued options a person already has may not reduce her freedom (since one can reject toe‐slicing), but it is hard to take it to be a strict increase in that person's freedom.
(16) As was argued earlier, the relation is two‐sided, and the evaluation of the freedom to lead a life and the assessment of the life led (including choosing freely) have to be done simultaneously, in a desegregated way.
(17) Since a person's agency objectives will typically include, inter alia, his or her own well‐being, the two will to some extent go together (e.g. an increase in well‐being, other things being equal, will involve a higher agency achievement). In addition, a failure to achieve one's non‐well‐being objectives may also cause frustration, thereby reducing one's well‐being. These and other connections exist between well‐being and agency, but they do not make the two concepts congruent–nor isomorphic in the sense of generating the same orderings. Similarly, more freedom (either to have well‐being or to achieve one's agency goals) may lead one to end up achieving more (respectively, of well‐being or of agency success), but it is also possible for freedom to go up while achievement goes down, and vice versa. We have here four interdependent but non‐identical concepts. These distinctions and their interrelations are discussed more fully in Sen (1985a, 1992).
(20) This view may be disputed by considering a different way of drawing the line between wellbeing and living standards. One common approach is to relate the assessment of living standards only to real incomes and to ‘economic’ or ‘material’ causes. On this see A. C. Pigou (1920); and on the conceptual differences see Bernard Williams (1987). But the Pigovian view has problems of its own. For example, if one has a disability that makes one get very little out of material income or wealth, or if one's life is shattered by an inconvenient and incurable illness (e.g. kidney problems requiring extensive dialysis), it is hard to claim that one's standard of living is high just because one is well heeled. I have discussed this question and related matters in Sen (1987b: 26–9, 109–10).
(22) On this see Sen (1985b: 60–1). The distinguished element can be the achieved one (as in this case), or more specifically the chosen one (if there is a choice exercise in determining what happens), or the maximal one (in terms of some criterion of goodness). The three will coincide if what is achieved is achieved through choice, and what is chosen is chosen through maximization according to that criterion of goodness.
(23) As was argued earlier in dealing with responsible adults, it may be appropriate to see the claims of individuals on society in terms of the freedom to achieve well‐being (and thus in terms of real opportunities) rather than in terms of actual achievements. If the social arrangements are such that a responsible adult is given no less freedom (in terms of set comparisons) than others, but he still ‘muffs’ the opportunities and ends up worse off than others, it is possible to argue that no particular injustice is involved. On this and related matters, see Sen (1985a).
(24) The same capability set can than be used for the evaluation of both ‘well‐being achievement’ (through elementary evaluation, concentrating on the achieved element) and ‘well‐being freedom’ (through non‐elementary set evaluation).
(25) Thus, in standard consumer theory, set evaluation takes the form of elementary evaluation. For particular departures from that tradition, see Koopmans (1964) and Kreps (1979). In the Koopmans–Kreps approach, however, the motivation is not so much to see living freely as a thing of intrinsic importance, but to take note of uncertainty regarding one's own future preference by valuing—instrumentally—the advantage of having more options in the future. On the motivational contrasts, see Sen (1985a, 1985b).
(26) As was argued in an earlier paper, ‘the “good life” is partly a life of genuine choice, and not one in which the person is forced into a particular life—however rich it might be in other respects’ (Sen, 1985b): 69–70).
(30) The ‘basic needs’ literature is extensive. For a helpful introduction, see Streeten et al. (1981). In a substantial part of the literature, there is a tendency to define basic needs in the form of needs for commodities (e.g. for food, shelter, clothing, health care), and this may distract attention from the fact that these commodities are no more than the means to real ends (inputs for valuable functionings and capabilities). On this question, see Streeten (1984). The distinction is particularly important since the relationship between commodities and capabilities may vary greatly between individuals even in the same society (and of course between different societies). For example, even for the elementary functioning of being well nourished, the relation between food intake and nutritional achievements varies greatly with metabolic rates, body size, gender, pregnancy, age, climatic conditions, epidemiological characteristics, and other factors (on these and related matters, see Drèze and Sen, 1989). The capability approach can accommodate the real issues underlying the concern for basic needs, avoiding the pitfall of ‘commodity fetishism’.
(31) The importance of socially recognized ideas of ‘urgency’ has been illuminatingly discussed by Thomas Scanlon (1975).
(32) On this see Sen (1980). To avoid confusion, it should also be noted that the term ‘basic capabilities’ is sometimes used in quite a different sense from the one specified above, e.g. as a person's potential capabilities that could be developed, whether or not they are actually realized (this is the sense in which the term is used, for example, by Martha Nussbaum (1988).
(33) While the notion of basic capabilities was used in Sen (1980, 1983c), in later papers the capability approach has been used without identifying certain capabilities as ‘basic’ and others as not so (see e.g. Sen, 1984, 1985a, 1985b). This point is relevant to G.A. Cohen's distinction between focusing on what he calls ‘midfare’ and on functioning and capabilities. There are more important distinctions to explore (to be taken up in Section 9), but the contrasts look artificially sharper if the capability approach is seen as being confined only to the analysis of basic capabilities.
(35) Technically, what is being used in this analysis is the ‘inverse function’, taking us back from specified capability levels to necessary incomes, given the other influences on capability. This procedure will not be usable, in this form, if there are people who are so handicapped in terms of personal characteristics (e.g. being a ‘basket case’) that no level of income will get them to reach minimally acceptable basic capabilities; such people would then be invariably identified as poor.
(36) Perhaps the word ‘capability’ is misleading, but I am not sure that this should be the case. The pieces of land to which Capability Brown attributed ‘capability’ could not have been much more active in looking after themselves than babies are. The crucial Greek word used in this context (by Aristotle among others), namely dunamin, can be translated as ‘capability of existing or acting’, and presumably ‘existing’ need not be the result of some vigorous ‘exercise of capability’. Nor do I have any great difficulty in saying that the babies in question did achieve the state of being nourished and warm. Perhaps something else in my inept prose misled Cohen.
(41) As was mentioned earlier (in footnote 2), in their well‐known Greek‐English lexicon, even Liddell and Scott (1977) had translated the Greek word dunamin, central to Aristotle's concept of human good, as ‘capability of existing or acting’ (p. 452).
(53) Corresponding to ‘equality of what?’, there is, in fact, also the question: ‘efficiency of what?’