Jump to ContentJump to Main Navigation
The Quality of Life$

Martha Nussbaum and Amartya Sen

Print publication date: 1993

Print ISBN-13: 9780198287971

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198287976.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy).date: 16 January 2017

1 Equality of What? on Welfare, Goods, and Capabilities

1 Equality of What? on Welfare, Goods, and Capabilities

(p.9) 1 Equality of What? on Welfare, Goods, and Capabilities
The Quality of Life

G. A. Cohen

Oxford University Press

1 Introduction

In his Tanner Lecture of 1979 called ‘Equality of What?’ Amartya Sen asked what metric egalitarians should use to establish the extent to which their ideal is realized in a given society. What aspect(s) of a person's condition should count in a fundamental way for egalitarians, and not merely as cause of or evidence of or proxy for what they regard as fundamental?

In this study I comment on answers to Sen's question in recent philosophical literature. I take for granted that there is something which justice requires people to have equal amounts of, not no matter what, but to whatever extent is allowed by values which compete with distributive equality; and I describe and criticize what a number of authors who share that egalitarian view have said about the dimension(s) or respect(s) in which people should be made more equal, when the cost in other values of moving towards greater equality is not intolerable.1

The publication of John Rawls's A Theory of Justice in 1971 was a watershed in discussion bearing on the question, derived from Sen, which forms my title. Before A Theory of Justice appeared, political philosophy was dominated by utilitarianism, the theory that sound social policy aims at the maximization of welfare. Rawls found two features of utilitarianism repugnant. He objected, first, to its aggregative character, its unconcern about the pattern of distribution of welfare, which means that inequality in its distribution calls for no justification. But, more pertinently to the present exercise, Rawls also objected to the utilitarian assumption that welfare is the aspect of a person's condition which commands normative attention. Rawls replaced aggregation by equality and welfare by primary goods. He recommended normative evaluation with new arguments (goods instead of welfare quanta) and a new function (equality2 instead of aggregation) from those arguments to values.

Rawls's critique of the welfare metric was undoubtedly powerful, but, as I (p.10) shall argue, his motivation for replacing it by attention to primary goods was not correspondingly cogent. He did not consider, as an alternative to equality of welfare, the claims of equality of opportunity for welfare, which his criticisms of equality of welfare do not touch. What is more, those criticisms positively favour equality of opportunity for welfare as a remedy for the defects in the rejected doctrine.

But while equality of opportunity for welfare survives Rawls's criticisms of equality of welfare, arguments against the welfare metric which were later advanced by Sen also apply against its opportunity‐defined cousin. Sen called for attention to something like opportunity (under the title ‘capability’3), but it was not welfare, or not, at any rate, welfare alone, which Sen thought people should have the opportunity to achieve. Instead, he drew attention to the condition of a person (e.g. his level of nutrition) in a central sense captured neither by his stock of goods (e.g. his food supply) nor by his welfare level (e.g. the pleasure or desire satisfaction he obtains from consuming food). In advancing beyond Rawls, Sen therefore proposed two large changes of view: from actual state to opportunity, and from goods (and welfare) to what he sometimes called ‘functionings’.

In my view, Sen's answer to his own question was a great leap forward in contemporary reflection on the subject. But often a thinker who achieves a revolution misdescribes his own achievement, and I shall argue, at some length, that Sen's work is a case in point. He moved away from Rawlsian and other views in two directions which were orthogonal to each other. If Rawls and welfarists fixed on what a person gets in welfare or goods, Sen fixed on what he gets in a space between welfare and goods (nutrition is delivered by goods supply and it generates welfare), but he also emphasized what a person can get, as opposed to (just) what he does. Sen's misdescription of his achievement lay in his appropriation of the word ‘capability’ to describe both of his moves, so that his position, as he presented it, is disfigured by ambiguity. I shall here expose the ambiguity in Sen's use of ‘capability’ (and cognate terms), and I shall also propose an answer to his (my title) question which departs from his own in a modest way.

2 Rawlsian Criticism of Equality of Welfare

Before examining Rawls's critique of equality of welfare, a word about what will here be meant by ‘welfare’. Of the many readings of ‘welfare’ alive (if not well) in economics and philosophy, the present inquiry recruits two: welfare as enjoyment, or, more broadly, as desirable or agreeable state of consciousness, which (p.11) I shall call hedonic welfare; and welfare as preference satisfaction, where preferences order states of the world, and where a person's preference is satisfied if a relevant state of the world obtains, whether or not he knows that it does.4 It will sometimes be necessary to say which of those two I mean by ‘welfare’, but not always. Often the debates on which I comment have a similar shape under either interpretation of welfare, so that I shall have both in mind (by which I do not mean some amalgam of the two) at once. Unless I indicate otherwise, my contentions are meant to hold under either of the two readings of ‘welfare’ that I have just distinguished.

Rawls advances two objections to equality of welfare, and they apply against both its hedonic and its preference interpretations. I shall call his objections the ‘offensive tastes’ and ‘expensive tastes’ criticisms. I believe that each criticism can be accommodated by a welfare egalitarian through a natural modification of his original view. In the case of the offensive tastes criticism, that would probably be conceded by Rawls (and by Ronald Dworkin, who develops the criticism more systematically and at some length5). But the second criticism is supposed by Rawls and Dworkin to justify an abandonment of the terrain of welfare altogether; as I shall indicate, I do not think that it does. The second criticism also creates a problem for Rawls's system, which I shall describe in a brief digression.

Rawls adverts to offensive tastes in the course of his critique of utilitarianism, but, as Amartya Sen notes,6 he is at that point really criticizing welfarism as such, where welfarism is the view that just distribution is some or other function of nothing but the welfares of individuals. It follows logically that the offensive tastes criticism also applies against a conception of justice in which equality of welfare is the only principle. And although a believer in equality of welfare who also affirms further principles need not be a welfarist other than with respect to the metric of equality in particular, it is extremely unlikely that a good criticism of welfarism proper will not also apply to that restricted welfarism which acknowledges the relevance of no information but welfare in the context of (p.12) equality, even if its proponent admits non‐welfare information elsewhere. In any case, the offensive tastes criticism strikes me as powerful against the welfare‐egalitarian claim, whether or not other claims are conjoined with it.

The offensive tastes criticism of welfarism is that the pleasure a person takes in discriminating against other people or in subjecting others to a lesser liberty should not count equally with other satisfactions in the calculus of justice.7 From the point of view of justice, such pleasures deserve condemnation, and the corresponding preferences have no claim to be satisfied, even if they would have to be satisfied for welfare equality to prevail. I believe that this objection defeats welfarism, and hence equality of welfare. But the natural course for a welfare egalitarian to take in response to the offensive tastes criticism is to shift his favour to something like equality of inoffensive welfare. The criticism does not seem to necessitate abandoning equality of welfare in a more fundamental way.8

The ‘expensive tastes’ criticism is thought to necessitate such an abandonment. It occurs in the context of Rawls's advocacy of primary goods as the appropriate thing to equalize. Now the phrase ‘primary goods’ covers a set of different things in Rawls,9 but for our purposes, we can, as Rawls does in the texts I shall quote, take it narrowly, as referring to goods in the economist's sense, or the power to purchase them.

Prosecuting his case against welfare and for primary goods, Rawls asks us to ‘imagine two persons, one satisfied with a diet of milk, bread and beans, while the other is distraught without expensive wines and exotic dishes. In short one has expensive tastes, the other does not.’ A welfare egalitarian must, ceteris paribus, provide the epicure with a higher income than the person of modest tastes, since otherwise the latter might be satisfied while the former is distraught. But Rawls argues powerfully against this implication of the welfare egalitarian principle:

as moral persons citizens have some part in forming and cultivating their final ends and preferences. It is not by itself an objection to the use of primary goods that it does not accommodate those with expensive tastes. One must argue in addition that it is unreasonable, if not unjust, to hold such persons responsible for their preferences and to require them to make out as best they can. But to argue this seems to presuppose that ‘citizens’ preferences are beyond their control as propensities or cravings which simply happen. Citizens seem to be regarded as passive carriers of desires. The use of primary goods . . . relies on a capacity to assume responsibility for our ends.

People with expensive tastes could have chosen otherwise, and if and when they (p.13) press for compensation, then others are entitled to insist that they themselves bear the cost of ‘their lack of foresight or self‐discipline’.10

I believe that this objection defeats welfare egalitarianism but that it does not, as Rawls supposes, also vindicate the claims of the primary goods metric. The right way for an erstwhile welfare egalitarian to respond to the objection seems to me to be the following: ‘To the extent that people are indeed responsible for their tastes, the relevant welfare deficits do not command the attention of justice. We should therefore compensate only for those welfare deficits which are not appropriately traceable to the individual's choices. We should replace equality of welfare by equality of opportunity for welfare. It would be utterly unjustified to adopt a primary goods metric because of the expensive tastes counter‐example.’

Equality of opportunity for welfare,11 unlike equality of welfare, permits and indeed enjoins departures from welfare equality when they reflect the choices of relevant agents, as opposed to deficient opportunity for welfare. If a person's welfare is low because he freely risked a welfare loss in gambling for a welfare gain, then, under the opportunity form of the principle, he has no claim to compensation. Nor has a person who frittered away welfare opportunities which others seized. Nor, to take a different kind of example, has a person who chose to forgo welfare out of devotion to an ideal which (expressly, or merely as it happened) required self‐denial. A person like Rawls's epicure who, ex hypothesi, could have controlled himself and maintained more austere tastes, has, in consequence, no grievance at the bar of equality of opportunity for welfare if his actual welfare is substandard because his tastes are expensive. The eligibility of equality of opportunity for welfare as a reply to Rawls shows that his conviction that the reasons he gives for rejecting equality of welfare are also reasons for affirming equality of primary goods is misplaced.

Before leaving Rawls, I want to indicate a serious problem for his political philosophy which his remarks about expensive tastes raise. The problem would be less interesting than it is in fact were it merely a problem in the Rawlsian architectonic and not also expressive of a general tension in left‐egalitarian thinking.

The problem, as it appears in Rawls, is that the picture of the individual as (p.14) responsibly guiding his own taste formation is hard to reconcile with claims Rawls uses elsewhere in a fundamental way to support his egalitarianism. These claims, which are quite standard on the left, express scepticism about special efforts as a ground for rewarding people unequally. Here is how Rawls formulates this fairly familiar scepticism:

the effort a person is willing to make is influenced by his natural abilities and skills and the alternatives open to him. The better endowed are more likely, other things equal, to strive conscientiously, and there seems to be no way to discount for their greater good fortune. The idea of rewarding desert is impracticable.12

Now there are two ways of taking this passage. One way is as I think Rawls intended it, and the other is as Robert Nozick took it, and on the basis of which he directed strong criticism against Rawls. Nozick, I am sure, misread the passage, but his misreading of it constitutes a correct reading of what many socialists and egalitarians say about effort, so it will be worth our while to pause here, to attend to Nozick's criticism. On either reading of the passage, it is hard to reconcile with what Rawls says about foresight, self‐discipline, and expensive tastes. But I shall come to that point in a moment, for the passage, can also be criticized independently, and I want to do that first.

The two readings of the passage divide with respect to how they take the word ‘influenced’ in Rawls's use of it here. On my reading of it, it means ‘influenced’, that is, ‘affected, but not necessarily wholly determined’. On Nozick's reading of the passage, ‘influenced’ means something like ‘wholly determined’. There is a difficulty for Rawls whichever way we take the passage, but not the same difficulty in each case.

On my reading of Rawls, in which he means ‘influenced’ by ‘influenced’, he does not say that those who make a lot of effort have no control over, and therefore deserve no credit for, the amount of effort they put in. His different point is that we cannot reckon the extent to which their above‐par effort is attributable not to admirable striving but to ‘greater good fortune’: there is no ‘way to discount’ for the latter. That is a practical objection to trying to reward effort that deserves reward, not a claim that there is no such effort: see the final sentence of the passage.

If Rawls is right that not all effort is deserving, then, we might agree, not all effort deserves reward. But why should it follow that effort deserves no reward at all? The practical difficulty of telling how much of it merits reward hardly justifies rewarding it at a rate of 0 per cent, as opposed to at a rate of somewhere between 0 and 100 per cent, for example through a taxation scheme whose shape and justification escapes, because of its deference to effort, the writ of the Difference Principle. (To be sure, it is hard to tell how much effort a person supplies, but Rawls's claim is not that total effort is indiscernible, but that we cannot identify the contribution to effort of fortunate endowment.)

(p.15) But that criticism of Rawls is mild by comparison with the one to which he is exposed on Nozick's reading of his remarks. The plausibility of that reading is enhanced by Nozick's careless or mischievous omission of what follows ‘conscientiously’ when he exhibits the Theory of Justice passage quoted above. Nozick thereby creates the impression that Rawls is presenting a familiar egalitarian determinist doctrine. Nozick's response to that doctrine is very powerful:

So denigrating a person's autonomy and prime responsibility for his actions is a risky line to take for a theory that otherwise wishes to buttress the dignity and self‐respect of autonomous beings . . . One doubts that the unexalted picture of human beings Rawls' theory presupposes and rests upon can be made to fit together with the view of human dignity it is designed to lead to and embody.13

Nozick is pressing a dilemma: either people have real freedom of choice, in which case they may be credited (at least to some extent) with the fruits of their labours; or there is no such thing as free choice, in which case liberals should take the purple out of the passages in which they set forth their conception of humanity, and — we can add — socialists should stop painting inspiring pictures of the human future (unless they believe that people lack free will under capitalism but will get it after the revolution).

On Nozick's reading of the ‘effort’ passage, it is clearly inconsistent with the responsibility for taste formation with which Rawls credits citizens. That does not matter too much, since Nozick's reading is a misreading. But it is not easy to reconcile what Rawls says about effort with what he says about tastes, even on my less creative reading of his text. On my reading of it, effort is partly praiseworthy and partly not, but we cannot separate the parts, and the indicated policy consequence is to ignore effort as a claim to reward. Now, the passage about tastes begins with the thought that ‘citizens have some part in forming and cultivating their final ends and preferences’, though it ends by assigning a more wholesale responsibility for preferences to citizens. If we stay with the opening thought, then we can wonder why partial responsibility for effort attracts no reward at all while (merely) partial responsibility for expensive taste formation attracts a full penalty (and those who keep their tastes modest reap a welfare reward). And if we shift to the wholesale responsibility motif, then we can wonder why beings who are only in a limited way responsible for the effort they put in may be held wholly responsible for how their tastes develop.

The most natural view about these matters is that people are partly responsible and partly not, both for how much effort they put into production and for whether or not they develop expensive tastes.14 This does not mean that it is easy to tell where responsibility begins and ends or, consequently, what the dictate of distributive justice is in particular instances — but there is no antecedent (p.16) reason whatsoever for supposing that judgements about justice, at a fine‐grained degree of resolution, are easy.

3 Sen and Capability

a. The foregoing critique of Rawls does not prove that quantity of primary goods is the wrong metric for egalitarian evaluation; it just proves that a major reason Rawls offered for favouring primary goods points, instead, to equality of opportunity for welfare. For a more thorough refutation of the primary goods proposal, I turn to Sen's ‘Equality of What?’ That seminal article also argues persuasively against the welfare metric, and, while Sen did not address equality of opportunity for welfare, his argument against equality of welfare readily extends itself to the former view. After presenting and endorsing Sen's negative arguments, I shall argue that his positive replacing proposal, capability equality, suffers from a severe expositional obscurity.

Sen's argument against the primary goods metric was simple but powerful. It was that differently constructed and situated people require different amounts of primary goods to satisfy the same needs, so that ‘judging advantage in terms of primary goods leads to a partially blind morality’.15 It is, Sen rightly said, a ‘fetishist handicap’ to be concerned with goods as such, to the exclusion of what goods ‘do to human beings’.16 Or, as Sen later expressed the point: ‘what people get out of goods depends on a variety of factors, and judging personal advantage just by the size of personal ownership of goods and services can be very misleading . . . It seems reasonable to move away from a focus on goods as such to what goods do to human beings’.17 The principle of equality condemns equal goods provision to a sound‐limbed person and a paraplegic, because greater resources are necessary to enable the latter to achieve mobility, a desideratum to which a metric of stock of wealth is blind.18

Sen also used the example of the needy cripple to good effect against the welfare alternative to primary goods. For the egalitarian response to his plight is not determined by a judgement that he suffers a welfare deficiency. Perhaps, indeed, he suffers no such thing, ‘because he has a jolly disposition. Or because he has a low aspiration level and his heart leaps up whenever he sees a rainbow in the sky.’19 So while both hedonic and preference‐satisfaction welfarists are free of the goods theorist's fetishistic neglect of what goods do to human beings, Sen criticized them for their too‐narrow view of what people get from goods, for focusing ‘not on the person's capabilities but on his mental reaction’, not, for example, on how much nourishment a person gets from food, but on how much (p.17) utility, which is a matter of mental reaction or attitude, he derives from such nourishment.20 Utility is an unsuitable guide to policy, if only because a person may adjust his expectations to his condition. The fact that a person has learned to live with adversity, and to smile courageously in the face of it, should not nullify his claim to compensation.21

His high welfare score is thus not a decisive reason for not assisting someone who labours under a severe disadvantage, which is recognizable as such from an objective point of view. His equanimity may, after all, reflect admirable and reward‐worthy striving to overcome a natural reaction of misery. But even if no such striving is necessary, because the person is blessed from birth with an extra‐sunny disposition, the requirement of compensation retains intuitive force. And that means that not only equality of welfare but also equality of opportunity for welfare falls before the case of the cripple. Consider the poor and lame but sunny‐spirited Tiny Tim. Tiny Tim is actually happy, by any welfarist standard. And we may also suppose that, because of a fortunate innate disposition, he is blessed with abundant opportunity for happiness, that he need not do much to get a lot of it. Yet egalitarians would not on that account strike him off the list of free wheelchair receivers. Hence they do not think that wheelchair distribution should be controlled exclusively by the welfare opportunity requirements of those who need them. They need wheelchairs to be adequately resourced, whether or not they also need them to be, or to be capable of being, happy.

b. In the course of making the critical points reported and endorsed above, Sen used the term ‘capability’, and he appropriated that term to denote his own positive counter‐proposal. I shall now argue that, in ‘Equality of What?’, Sen brought two distinct aspects of a person's condition under that single name, and that this unnoticed duality has persisted in his subsequent writings. Both aspects, or dimensions of assessment, should attract egalitarian interest, but one of them is not felicitously described as ‘capability’. The identification of that latter dimension constitutes a particularly striking contribution to normative understanding, but just that dimension is hard to perceive in Sen's exposition, because of the unfortunate and ambiguous nomenclature.

As we have seen, Sen arrived at what he called ‘capability’ through reflection on the main candidates for assessment of well‐being that were in the field when he gave his 1979 lecture, to wit, utility, or welfare, and Rawlsian primary goods.22 Sen pleaded for a metric of well‐being which measured something (p.18) falling between primary goods and utility, in a sense that will presently be explained–a something which had, amazingly, been largely neglected in previous literature. He called that something ‘capability’: ‘what is missing in all this framework23 is some notion of “basic capabilities”: a person being able to do certain basic things’.24 But that characterization of the missing dimension was different from another which Sen offered in the same text, and which was more in keeping with his argument for the new perspective.

According to that argument, as we have seen, it is necessary to attend to what goods do to (or for) human beings, in abstraction from the utility they confer on them. But to call what goods supply to human beings ‘capability’ was a mistake. For even when utility has been set aside, it is not true that all that goods do for people is confer capability on them–provide them, that is, with the capacity to do things–or that that is the uniquely important thing they do for them, or that that is the one thing they do for them that matters from an egalitarian point of view. In naming his view ‘Basic Capability Equality’ Sen failed to delineate the true shape and size of one of the dimensions he had uncovered, and which I shall now try to describe.

It is indeed false that the whole relevant effect on a person of his bundle of primary goods is on, or in virtue of, his mental reaction to what they do for him. There is also what welfarists ignore: what they do for him, what he gets out of them, apart from his mental reaction to or personal evaluation of that service. I shall call that non‐utility effect of goods midfare, because it is in a certain sense midway between goods and utility. Midfare is constituted of states of the person produced by goods, states in virtue of which utility levels take the values they do. It is ‘posterior’ to ‘having goods’ and ‘prior’ to ‘having utility’.25

Midfare is a heterogeneous collocation, because goods do categorially various things for people: (1) they endow them with capabilities properly so called, which they may or may not use; (2) through people's exercise of those capabilities, goods contribute to the performance of valuable activities and the achievement of desirable states; and (3) goods cause further desirable states directly, without any exercise of capability on the part of their beneficiary: an example would be the goods which destroy the insects that cause malaria. Capability (properly so called) is, then, a part of midfare, for it certainly cannot be excluded from the range of things that goods confer on people, yet, equally certainly, it does not exhaust that range.

Each terminus of the goods–midfare–utility sequence has seemed to some the right focus for assessment of a person's situation from an egalitarian point of view. Rawlsians look at the beginning of the sequence and welfarists look at (p.19) its end. Welfarists think that the Rawlsian measure is too objective, that it takes too little account of distinguishing facts about individuals. Rawlsians think that the welfare measure is too subjective, that it takes too much account of just such facts. The reasons each side gives for disparaging its opponent's dimension suggest that each should prefer midfare to the dimension favoured by its opponent. Welfarists draw attention to utility because, so they say, people do not care about goods as such but about the utility they provide. But, since people also care more about midfare than about goods as such (save where they are themselves being fetishistic), the welfarist reason for preferring welfare to goods is also a reason for preferring midfare to the latter. Advocates of goods oppose the welfare metric because, they say, the welfare consequences of goods consumption are (1) too subject to volition (Rawls, sometimes26), (2) too much a matter of people's (not necessarily chosen) identifications (Rawls at other times,27 Dworkin28), or (3) too idiosyncratic (Scanlon29). On all three grounds midfare arguably scores better than utility does.

Given that each side in the foregoing division has reason to prefer the midfare dimension to the one favoured by its opponents, it is extraordinary that midfare had not been uncovered, and Sen's reorienting proposal was consequently profound and liberating, albeit remarkably simple. For it simply says that, in the enterprise of assessing a person's well‐being, we must look to her condition in abstraction from its utility for her. 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.30

But this significant and illuminating reorientation is not equivalent to focusing on a person's capability, in any ordinary sense. Capability, and exercises of capability, form only one part of the intermediate midfare state. What goods do to people is identical neither with what people are able to do with them nor with what they actually do with them (and it is also not identical with all or part of the combination of these two things). To be sure, it is usually true that a person must do something with a good (take it in, put it on, go inside it, etc.) in order to be benefited by it, but that is not always true, and, even where it is true, one must distinguish what the good does for the person from what he does with it. The colloquial question ‘How are you doing?’ can be used to ask after a person's midfare (especially when this pedantic rider is attached to it: ‘by which I do not mean how do you feel about how you are doing’), but the usual answer will not be (just) a list of capacities, activities, and results of activities, because not all midfare is capability or an exercise of capability or a result of exercising capability. And many midfare states which are indeed a result of (p.20) exercising capability have a (non‐utility) value which is unconnected with their status as effects of exercising capability, and which is not clearly exhibited in its true independence of capability by Sen.

The case of food, which has, of course, exercised Sen a great deal, illustrates my claims. The main good thing that food does for people is that it nourishes them. Typically, of course, people become nourished by nourishing or feeding themselves, by exercising the capability of nourishing themselves which ownership of food confers on them. But the fact that food gives a person the capability to nourish himself is not the same fact as (and is usually less important than) the fact that it enables him to be nourished. To say that food enables him to be nourished is to say that it makes it possible for him to be nourished. That he characteristically actualizes that possibility himself is a further (and usually less important) fact. When, moreover, we ask how well nourished a person is, we are not asking how well he has nourished himself, even though the answer to the two questions will usually be the same; and we are usually primarily interested in the answer to the first question.31

The difference between midfare and capability (properly so called) will perhaps become more evident if we reflect a little about small babies. Small babies do not sustain themselves through exercises of capability. But it is false that, in the case of babies, goods generate utility and nothing else worth mentioning. When food is assigned for the consumption of either a baby or an adult, each is enabled to be nourished. The fact that only the adult is able to nourish himself does not mean that he alone gets midfare. The baby gets it too. Hence 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.

If food does not make my case strongly enough, since babies do suck and chew, think instead of clothes. No collaboration on the baby's part is needed when its parent confers the midfare of warmth and protection on it by dressing it. Or consider the midfare supplied by the nutriment in a hospital drip, to baby and adult alike, or, for that matter, by the rays of the sun. There is no relevant exercise of capability by benefited agents in these instances, but there is an important benefit to be described in non‐welfarist–midfare–terms. Hence the concept of capability is insufficiently general to capture one of the things that Sen wants to identify.

There are two powerful motivations for pointing to something other than either goods or utility when concerning oneself with egalitarian policy, but the motivations point at different things. There is good reason to look at what a (p.21) person can achieve, independently of his actual state; and there is good reason not to reduce the evaluation of that actual state either to an examination of his stock of resources or to an assessment of his utility level. But these are distinct points, and the language of capability felicitously covers the first one only.

The ambiguity I have tried to expose appears in a number of Sen's dictions, including the apparently harmless phrase ‘what people get out of goods’.32 On one reading of it, under which ‘get out of’ means (roughly) ‘extract’, getting things out of goods represents an exercise of capability. But ‘get out of’ can also mean, more passively, ‘receive from’, and it does not require capability to get things out of goods in that sense. Goods (and welfare) theorists ignore (some of) what people get out of goods in both senses of the phrase, but while only the first sense relates to capability, the second denotes something at least as important.

c. In Sen's discourse, to have a capability is to be capable of achieving a range of what he calls ‘functionings’. But Sen characterizes functionings differently at different times, and thereby adds further imprecision to the presentation of his view.

Sometimes, in keeping with the ordinary meaning of ‘functioning’, and in line with Sen's original gloss on ‘capability’ as ‘being able to do certain basic things’,33 a functioning is by definition an activity, something that a person does.34 The questions ‘Can they read and write? Can they take part in the life of the community?’35 inquire into people's functionings in this familiar sense of the term. But at other times, functionings are not by definition activities but all (desirable) states of persons, and ‘being well nourished’, ‘being free from malaria’, and ‘being free from avoidable morbidity’36 are consequently entered as examples of functionings, although, not being activities, they are not functionings in the ordinary sense of the term. (Even though ‘I am free from malaria now’ can be part of the answer to the question ‘How are you doing?’ in its colloquial use.)

When Sen writes that ‘Functionings are . . . personal features; they tell us what a person is doing’,37 he places his incompatible broad and narrow definitions of ‘functioning’ on either side of the semi‐colon. For not all personal features, and not all of the personal features that Sen wishes to encompass, are things that a person is doing. Unlike reading and writing, being free from malaria is not something that one does. Elsewhere, a broader definition of ‘functionings’ (p.22) is offered, under which ‘they tell us what the person is doing or achieving’,38 and it is true that being free from malaria is something that one may achieve. But it is surely of supreme (midfare) importance even when one cannot be credited with achieving it.

Sen himself notes that being free from malaria may be entirely due to ‘anti‐epidemic public policy’.39 What he fails to note is the consequent impropriety of regarding it, in that instance, as something the person achieves, as the exercise of a capability of any kind. Yet Sen would surely not want to exclude heteronomously obtained freedom from malaria from the balance sheet of how a person ‘is doing’. And that proves that he has a concern to promote forms of midfare which does not derive from his concern to promote the claims of capability as such. Indeed, one may go further: the lacks in people's lives which Sen is most concerned to draw to our attention are midfare lacks which are not lacks in capability proper, and the alleviation of which need not always proceed through an enhancement of the sufferer's capability. He is concerned with people who are ‘ill‐fed, undernourished, unsheltered and ill’,40 who lack ‘basic clothing, ability to be housed and sheltered, etc.’41 Being able to be housed is not the same thing as being able to house oneself. Entitlements to goods make desirable states possible for people. They generally realize these possibilities themselves, by exercising a capability to do so. But, with respect to the lacks which most exercise Sen, it is the possibilities that matter, and the corresponding capabilities matter only derivatively.

At one point42 Sen extols the importance of ‘a person's ability to function without nutritional deficiency’ and of ‘the capability of avoiding nutritional deficiency’. Such functionings and avoidings are genuine activities,43 but the generative desideratum here is not activity but simply lack of nutritional deficiency, the fundamental desirability of which is lost in these athletic phrasings. It is not hitting the nail on the head to say that food is desirable because it enables a person to avoid nutritional deficiency, as though performing that activity is the (one) important thing here. Decent living space, to change the example, is a primary good which helps to maintain a person in good health, and it often does that when it would be false to say that it helps him to maintain himself in good health. Whether he is doing that is an exquisitely subtle question, a negative answer to which is consistent with decent living space delivering its hygienic boon. More generally, the ‘kind of life I am living’ cannot be identified (p.23) with what I am ‘succeeding in “doing” or “being” ’,44 unless we put scare‐quotes around ‘succeeding’ as well. There are many benefits I get which I do not literally succeed in getting.

It is true that the better nourished I am, the larger is the number of valuable activities of which I shall be capable. But that capability, the important capability which food confers, is a result of eating it. It is not the Sen capability associated with food, which is a capability to use food to achieve various ‘functionings’: being nourished, conducting a ceremonial, entertaining friends.45 One cannot infer from the central place in life–and midfare–of action and capacity that capability spreads across the entire space of midfare. And, as we have seen, not everything which merits attention under the broad midfare construal of Sen's contention is an activity or achievement.

We may conclude that, while Sen's focus on what goods do for people apart from the mental reaction they induce is original and illuminating, it is unnecessarily narrowed when the object of the focus is described in functioning/ capability language. Comprehending as it does everything which ‘goods do for people’,46 midfare cannot be identified either with capability or with what Sen calls ‘functioning’, nor can it be factored into the two without a confusing stretching of the meanings of words.

d. Why did Sen use the language of capability and functioning to express claims which that language fits quite imperfectly? Because, I hypothesize, he had something in addition to midfare in mind, to wit, freedom, and he wrongly thought that attending to a person's midfare–to what he gets from goods apart from the utility upshot of getting it–is attending to how much freedom he has in the world. Both the misrepresentation of all desirable states as a result of the exercise of capability and the tendency to represent all desirable states as activities reflect an interest in freedom distinct from, but not clearly distinguished by Sen from, the move from both utility and goods to midfare.

There is a case for installing the notion of freedom within egalitarian discourse, but that is a different exercise from vindicating the claims of midfare as such. There are two powerful motivations for pointing to something other than either goods or utility in a comprehensive characterization of well‐being, but the motivations justify two distinguishable deviations from each of those metrics: possession of goods and enjoyment or utility are not the only actual states that matter, (p.24) and — here is the freedom motivation — it is not only actual states, but the range of states the agent can attain, that matter.

According to Sen, ‘the category of capabilities is the natural candidate for reflecting the idea of freedom to do’,47 since ‘capability to function reflects what a person can do’.48 Hence ‘the concept of capabilities is a “freedom” type notion’,49 and the functioning vectors accessible to a person determine her ‘well‐being freedom’.50 All that may be true of capability (more or less) strictly so called, but it is not true of ‘capability’ where the term is used to denote the entire midfare dimension between goods and utility. Sen intends capability to have an athletic character. He associates it with the Marxist idea of a person fulfilling his potential through activity, which is to be contrasted with the idea of a person finding his summum bonum in passive consumption.51 But, in Sen's wider construal of it, as midfare, capability covers too much to provide ‘the perspective of “freedom” in the positive sense’.52

The ambiguity between capability as a form of freedom and capability as midfare was not resolved in Sen's contribution to the July 1988 WIDER symposium (‘Capability and Well‐Being’), to which, with characteristic generosity, he has allowed me to refer. Instead, and as I shall explain, his ambiguous use of ‘capability’ was matched by an ambiguous use of ‘freedom’.

At p. 5 of his 1988 typescript, Sen says that ‘capability reflects a person's freedom to choose between different ways of living’. That formulation more or less identifies capability with freedom of choice (how much it does so depends on what ‘reflects’ means here: it might mean ‘is’). In line with that characterization of capability is Sen's description of the rich faster, who ‘has the capability to be well nourished, but chooses not to [be]’.53

Elsewhere, however, something very different from the freedom to choose whether or not to eat, namely, freedom from hunger, is denominated a ‘capability’.54 In fact, though, freedom from hunger is being well nourished. It is not the ability to choose which the rich faster has: it is what he chooses not to have. Freedom from hunger is a desirable absence or privation, the sort of freedom which even beings that are not agents can have. Healthy plants have freedom from greenfly, and sound houses are free from dry rot. (Note that a person might even be described, in a special context, as free from nourishment, for example, when he wants to fast, or by captors who want him to starve.)

(p.25) Unlike the freedom to choose whether or not to eat, freedom from hunger is not constitutively freedom to do anything. Sen speaks of exercising such ‘capabilities’ as freedom from hunger and freedom from malaria.55 But they are not freedoms that are exercised. Sen's application of the term ‘capability’ both to the freedom to avoid morbidity56 and to freedom from morbidity57 shows that, in the attempt to bring the very different issues with which he is concerned under the single rubric of ‘capability’, he is led to make equivocal use of the term ‘freedom’.58

When Sen introduced capability equality in ‘Equality of What?’, he was modest about its claims. It was ‘a partial guide to the part of moral goodness that is associated with the idea of equality’.59 Five years later, his claim for the new perspective was much stronger. For, in the Dewey Lectures, Sen said that ‘the primary feature of well‐being can be seen in terms of how a person can “function”, taking that term in a very broad sense’, and that ‘the accounting of functioning vectors’ provides ‘a more plausible view of well‐being’ than competing conceptions do.60 Elsewhere, we are advised that, in assessing ‘well‐being and advantage’, we should focus ‘on the capability to function, i.e. what a person can do or be’. His utility is only evidence of a person's advantage in that central sense,61 and the goods at his disposal (here called his ‘opulence’) are only causes of that advantage.62 The position of midfare between primary goods and utility, thus construed, is given as a reason for treating it as the central dimension of value.

These are strong claims, but they are easier to accept in that functionings are now explicitly described as ‘doings and beings’ so that both ‘activities’ and ‘states of existence or being’ come under the ‘functioning’ rubric.63 What I cannot accept is the associated athleticism, which comes when Sen adds that ‘the central feature of well‐being is the ability to achieve valuable functionings’.64 That overestimates the place of freedom and activity in well‐being. As Sen (p.26) writes elsewhere, ‘freedom is concerned with what one can do’ and ‘with what one can do’:65 midfare fails, on both counts, as a representation of freedom.

e. I said earlier that there are two powerful motivations for pointing to something other than either goods or utility when concerning oneself with egalitarian policy: there are other actual states that count, and it is not only actual states that count. In the last section I have shown how confusion of those two points is visible in the attempt to express both in the language of freedom, which is appropriate to the second point only.

Under one exegetically plausible disambiguation of Sen's formulations, they recommend equality of capability to achieve functionings, where ‘capability’ carries something like its ordinary sense (and is therefore not confused with midfare), and where ‘functionings’ denote all desirable states, and not desirable activities only. So disambiguated, Sen's theory displays two departures from equality of welfare: there is a change of modality, in that capability or opportunity, rather than final achievement, is key; and there is an enrichment of the conception of what opportunities are for–not welfare alone, but more broadly conceived good states of the person. In this reconstruction, the error of forcing the concept of capability to denote both the element of opportunity and the move to a broader conception of advantage is eliminated.

When Sen first invoked capability, it was in the context of a proposal that we attend to ‘basic capability equality’.66 The relevant capability was of a fundamental sort, capability whose absence disables the person from satisfying his basic needs. Such need satisfaction is, while clearly related to the achievement of welfare, also irreducible to the latter: one may need something for which one has no desire and one may desire something which does not constitute a need. At the basic level, we can, with some confidence, rank capabilities in importance without paying attention to people's tastes. But, as Sen points out, capability rankings are more moot once we pass beyond the basic desiderata of a normal human life:

when there is diversity of taste, it becomes harder to surmise about capability by simply observing achievement. For extreme poverty this problem is less serious. Valuing better nourishment, less illness and long life tend to be fairly universal, and also largely consistent with each other despite being distinct objectives. But in other cases of greater relevance to the richer countries–the informational problems with the capability approach can be quite serious.67

For capabilities which go beyond need satisfaction, it is hard to see how rankings are possible without recourse to utility valuations of the relevant states. In a critical comment on Sen, Richard Arneson tries to exploit the dependence on preference of the value of ‘higher’ capabilities: (p.27)

I doubt that the full set of my functioning capabilities [matters] for the assessment of my position. Whether or not my capabilities include the capability to trek to the South Pole, eat a meal at the most expensive restaurant in Omsk . . . matters not one bit to me, because I neither have nor have the slightest reason to anticipate I ever will have any desire to do any of these and myriad other things.68

Arneson infers that, in so far as the capability approach claims our attention, it is only as a different way of presenting the idea of equality of opportunity for welfare. But that conclusion is hasty. For one might hold that objective (non‐welfare) assessment of capability is possible at the basic level, even though, beyond that level, we evaluate capability according to the range of desires which it enables a person to satisfy. The capability which matters as such (that is, independently of its welfare consequences) is capability definitive of a normal human existence, capability whose absence spells non‐satisfaction of need. This answer to Arneson is anticipated by Sen:

The index of capabilities can be sensitive to the strength of desires without converting everything into the metric of desires. The welfarist picture drops out everything other than desires. A non‐welfarist over‐all index of capabilities may not drop out desires and may well be sensitive to the strength of desires without ignoring other influences on the indexing.69

And, one might add, the sensitivity of the capability index to desire is inversely related to the degree of ‘basicness’ of the region of capability space under exploration.

Still, if capability in its higher reaches waits on utility for its significance, it is in its more basic reaches that it makes its distinctive normative contribution, as Sen acknowledges: ‘The issue of capabilities–specifically “material” capabilities–is particularly important in judging the standard of living of people in poor countries–it is also important in dealing with poverty in rich countries.’70 And even if utility and opulence offer more general, non‐dependent (on other metrics) assessments of people's conditions, because of not being restricted to the basic, the notion of basic capability equality may provide an apter reading of the egalitarian impulse than they do. The problem of characterizing well‐being in general is not the same as the problem of the priorities of egalitarian justice, and basic midfare, if not basic capability as such, rather than goods bundles or utility quanta, surely is the first priority of justice.

f. In the last sentence of subsection e, I reintroduced the equivocation between capability and midfare. Here I shall explain why I did so, and why, more (p.28) particularly, capability as such is not, in my view, the right thing for an egalitarian to focus on.

I have elsewhere proposed that the right thing to equalize is what I called ‘access to advantage’.71 In that proposal, ‘advantage’ is, like Sen's ‘functioning’ in its wider construal, a heterogeneous 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. That special construal of access is motivated by the thought that egalitarians have to consider states of a person which he neither brought about nor ever was in a position to bring about, states which fall within category (3) of midfare, as it was sub‐classified above (desirable states caused directly, without any exercise of capability by the beneficiary). Under the disambiguation of Sen's position articulated in subsection e above, such states go unconsidered in the egalitarian reckoning (though Sen himself is, of course, supremely concerned about them).

Under equality of access to advantage, the normative accent is not on capability as such, but on a person not lacking an urgent desideratum through no fault of his own: capability to achieve the desideratum is a sufficient but not a necessary condition of not suffering such a lack. My own proposal strikes me as better attuned than capability equality to the true shape of the egalitarian concern with such things as health, nourishment, and housing.

Equality of access to advantage is motivated by the idea that differential advantage is unjust save where it reflects differences in genuine choice (or, more or less, capability) on the part of relevant agents, but it is not genuine choice as such (or capability) which the view proposes to equalize. The idea motivating equality of access to advantage does not even imply that there actually is such a thing as genuine choice. Instead, it implies that if there is no such thing–because, for example, ‘hard determinism’ is true–then all differential advantage is unjust. The fact that my view tolerates the possibility that genuine choice is a chimera makes salient its difference from Sen's. In my view, Sen has exaggerated the indispensability of the idea of freedom in the correct articulation of the egalitarian norm. No serious inequality obtains when everyone has everything she needs, even if she did not have to lift a finger to get it. Such a condition may be woeful in other ways, but it is not criticizable at the bar of egalitarian justice.

(p.29) Bibliography

Bibliography references:

Arneson, R. (1989). ‘Equality and Equality of Opportunity for Welfare’, Philosophical Studies, 56.

Cohen, G. A. (1989). ‘On the Currency of Egalitarian Justice’, Ethics, 99.

Dworkin, R. (1981a). ‘Equality of Welfare’, Philosophy and Public Affairs, 10.

—— (1981b). ‘Equality of Resources’, Philosophy and Public Affairs, 10.

Landesman, B. (1983). ‘Egalitarianism’, Canadian Journal of Philosophy, 13.

Nozick, Robert (1974). Anarchy, State and Utopia. New York: Basic Books.

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

—— (1975). ‘Fairness to Goodness’, Philosophical Review, 84.

—— (1982). ‘Social Unity and Primary Goods’, in A. K. Sen and B. Williams, (eds.), Utilitarianism and Beyond. Cambridge: Cambridge University Press.

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

—— (1986). ‘Citizens' Needs and Primary Goods’, unpublished.

Scanlon, T. M. (1975). ‘Preference and Urgency’, Journal of Philosophy, 72.

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

—— (1981). ‘Ethical Issues in Income Distribution: National and International’, in S. Grassman and E. Lundberg (eds.), The World Economic Order: Past and Prospects. London: Macmillan. Repr. in Sen (1984a).

—— (1982). Choice, Welfare and Measurement. Blackwell, Oxford.

—— (1983). ‘Development: Which Way Now?’, Economic Journal, 93.

—— (1983). ‘Economics and the Family’, Asia Development Review, 1.

—— (1984a). Resources, Values and Development. Oxford: Blackwell.

—— (1984b). ‘The Living Standard’, Oxford Economic Papers, 6.

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

—— (1985b). ‘Rights and Capabilities’, in T. Honderich (ed.), Morality and Objectivity: A Tribute to J. L. Mackie. London: Routledge and Kegan Paul. Repr. in Sen (1984a).

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

—— (1987). ‘Well‐Being and Agency’, unpublished.

—— et al. (1987). The Standard of Living. Cambridge: Cambridge University Press.

—— (1988). ‘Capability and Well‐being’, unpublished draft of the paper that follows in this volume.


For their excellent criticism of a previous draft of the material in this article, I thank Richard Arneson, John Baker, Gerald Barnes, Will Kymlicka, David Lloyd‐Thomas, John McMurtry, Thomas Scanlon, Amartya Sen, and Philippe van Parijs.

(1) I discuss conflict between equality and other values in section 1 of ‘On the Currency of Egalitarian Justice’, Ethics, 99 (July 1989). That article represents part of the contribution which I prepared for the July 1988 Helsinki WIDER symposium on ‘The Quality of Life’, and the present article is also drawn from that contribution. There is overlap between the articles in that a similar critique of Rawls appears in both. Discussions of Dworkin and Scanlon which appear in the Ethics article are not reproduced here.

(2) Or, strictly, the maximin function, which enjoins departures from equality when the worst off benefit as a result. But that complication is of no significance here.

(3) Immediately after introducing the notion of ‘capability to function’ in the Dewey Lectures, Sen shifts to the alternative language of ‘opportunity’ to express the same idea. See ‘Well‐Being, Agency and Freedom: The Dewey Lectures 1984’, Journal of Philosophy, 82 (April 1985), 200–1. Cf. Commodities and Capabilities (Amsterdam, 1985), 59; The Standard of Living (Cambridge, 1987), 36.

(4) These two readings of welfare correspond to Sen's ‘happiness’ and ‘desire fulfilment’ readings and exclude his ‘choice’ reading: see ‘Well‐Being, Agency and Freedom’, 187 ff. It is reasonable to ignore the ‘choice’ reading, since, as Sen shows, it reflects confusion about the relationship between preference and choice. My two readings also correspond to Ronald Dworkin's ‘conscious state’ and ‘relative success’ conceptions: see ‘Equality of Welfare’, Philosophy and Public Affairs, 10 (Summer 1981), 191–4, 204–9, 220–1. I do not consider welfare as ‘overall success’ (ibid., 209 ff.) because it is very hard to handle, and in any case it is, arguably, undermotivated: see my ‘On the Currency of Egalitarian Justice’, n. 34. I also set aside so‐called ‘objective theories of welfare’ (Dworkin, ‘Equality of Welfare’, 224–6), since most philosophers would consider them alternatives to any sort of welfare theory: Thomas Scanlon, for whom welfare is preference satisfaction, would describe his theory as anti‐welfarist, yet it is an objective theory of welfare in Dworkin's sense. Finally, to mention an author whose work is salient in this study, Richard Arneson has the same understanding of welfare as Scanlon and Rawls has not specified a particular conception–which is not to say that he should have done.

(5) See ‘Equality of Welfare’, 197–201.

(6) ‘Equality of What?’ in S. McMurrin (ed.), Tanner Lectures on Human Values i (Cambridge, 1980), 211.

(7) A Theory of Justice (Cambridge, Mass., 1971), 30–1.

(8) In fairness to Rawls, one should recall that he presented the offensive tastes criticism as an objection not to equality of welfare but to utilitarianism, and for utilitarians a move to ‘inoffensive welfare’ would no doubt constitute a pretty fundamental shift. From the fact that the same criticism applies against both views, and that each should be revised in the same way in the face of it, it does not follow that the distance between the original and the revised view is the same in both cases.

(9) See A Theory of Justice, 62.

(10) ‘Social Unity and Primary Goods’, in A. K. Sen and B. Williams (eds.), Utilitarianism and Beyond (Cambridge, 1982), 168–9. Cf. ‘Fairness to Goodness’, Philosophical Review, 84 (April 1975), 553; ‘Justice as Fairness, Political not Metaphysical’, Philosophy and Public Affairs, 14 (Summer 1985), 243–4. For a somewhat different explanation of why justice ignores expensive tastes, with less (not no) emphasis on the idea that they are subject to the agent's control and more on the idea that it is appropriate to hold him accountable for them, see the reply to Arrow in Rawls's unpublished ‘Citizens’ Needs and Primary Goods'.

For interesting comment on and sympathetic development of Rawls's views on responsibility for preference, see Bruce Landesman, ‘Egalitarianism’, Canadian Journal of Philosophy, 13 (March 1983), 37.

(11) For an elegant exposition and defence of equality of opportunity for welfare, in its preference interpretation, see Richard Arneson, ‘Equality and Equality of Opportunity for Welfare’, Philosophical Studies, 56 (1989).

(12) A Theory of Justice, 312.

(13) Anarchy, State and Utopia (New York, 1974), 214.

(14) I press that natural view against Dworkin's development of the theme of expensive tastes (see ‘Equality of Welfare’, 228–40) in section IV of ‘On the Currency of Egalitarian Justice’.

(15) ‘Equality of What?’ 216.

(16) Ibid., 218; see also ‘Ethical Issues in Income Distribution’, in Resources, Values and Development (Oxford, 1984), 294; Commodities and Capabilities, 23; The Standard of Living, 15–16, 22.

(17) ‘Introduction’, in Choice, Welfare and Measurement (Oxford, 1982), 29–30.

(18) ‘Equality of What?’ 218.

(19) Ibid., 217.

(20) Ibid., 218. ‘Mental reaction’ must here cover not only a kind of experience but also a subjective valuation, to cater for the preference form of welfarism.

(21) See, for further development of this point, Commodities and Capabilities, 21–2, 29; ‘Introduction’ in Resources, Values and Development, 34; ‘Rights and Capabilities’, in ibid., 308–9; ‘Goods and People’, in ibid., 512; The Standard of Living, 8–11.

(22) A notable further candidate not yet then in print is Dworkin's equality of resources. It would be a worthwhile–and difficult–exercise to distinguish each of the two Sen dimensions I shall describe from the Dworkin resources dimension. (For pertinent remarks, see Sen's excellent rebuttal, all of which strikes me as correct, of Dworkin's criticism of Sen's view: ‘Rights and Capabilities’, 321–3.)

(23) That is, the framework of discussion restricted to the rival claims of primary goods and utility as measures of well‐being, and, within ‘primary goods’, to goods in the ordinary sense. It is that subset of primary goods which is pertinent here.

(24) ‘Equality of What?’ 218.

(25) Commodities and Capabilities, 11. Midfare is the ‘state of a person’ in the sense of ibid., 23.

(26) See pp. 12–13 above.

(27) See the reference to ‘Citizens’ Needs and Primary Goods' in n. 10 above.

(28) See ‘Equality of Welfare’, 228–40; ‘Equality of Resources’, Philosophy and Public Affairs, 10 (Fall 1981), 302–3, and, for criticism of the latter, see section III c of my ‘On the Currency of Egalitarian Justice’.

(29) See his ‘Preference and Urgency’, Journal of Philosophy, 72 (Nov. 1975), 659–66, and see section V b of my ‘On the Currency of Egalitarian Justice’ for criticism of Scanlon's idiosyncrasy claim.

(30) ‘Introduction’ in Choice, Welfare and Measurement, 30.

(31) In one place (‘Goods and People’, 510), Sen makes one of the distinctions on which I am insisting: ‘while goods and services are valuable, they are not valuable in themselves. Their value rests on what they can do for people, or rather, what people can do with these goods and services.’ But why does Sen here reject his first and, in my view, superior suggestion in favour of the second? Because, I suggest, of an interest in advocating freedom, which is a desideratum different from midfare: see subsection d below.

(32) See p. 16 above, text to n. 17.

(33) ‘Equality of What?’, 218 (italics added).

(34) ‘ “Functionings” are what the person succeeds in doing with the commodities . . . at his or her command’ (Commodities and Capabilities, 10).

(35) ‘The Living Standard’, Oxford Economic Papers, 6 (1984), 84.

(36) Ibid., 84, and see ‘Well‐Being, Agency and Freedom’, 197. These examples fall under the characterization of functionings as ‘activities . . . or states of existence or being’ (ibid., 197). (In one place, Sen describes ‘being well nourished’ not as a functioning but as a capability, but that is probably a slip: see The Standard of Living, 18).

(37) ‘Rights and Capabilities’, 317.

(38) ‘The Living Standard’, 84.

(39) Commodities and Capabilities, 16.

(40) Ibid., 21

(41) Ibid., 73.

(42) I am embarrassed to say that I have lost my record of where these phrases occur.

(43) The ability to function without nutritional deficiency is trickier. I have that ability if and only if there is something I am able to do and I lack nutritional deficiency. But, in general, the characterization of the result as an ability is owing to the first clause in that statement of necessary and sufficient conditions, rather than to my nutritional good order.

(44) Commodities and Capabilities, 28. Cf. ibid., 51, where ‘well‐being’ is described as depending on ‘the particular achievements of the person–the kind of “being” he or she succeeds in having’. This is entirely implausible when ‘achievements’ and ‘succeeds’ are taken literally, and (see subsection d below) Sen has a reason to want their literal meanings to resonate.

(45) At one point Sen writes that ‘the essence of the capabilities approach is to see commodity consumption as no more than a means to generating capabilities’ (‘Goods and People’, 522 (italics added), but that is either an aberration or an unsignalled change of doctrine. For elsewhere commodity entitlement generates capability, which is the power to use or consume the commodity in a variety of ways, each of which uses is a functioning. In the different conceptualization just quoted, capability is a consequence of what elsewhere is called a ‘functioning’.

(46) See the catalogue characterization of midfare at p. 18 above.

(47) ‘Rights and Capabilities’, 316. Cf. ‘Economics and the Family’, in Resources, Values and Development, 376.

(48) ‘Rights and Capabilities’, 317.

(49) Commodities and Capabilities, 14.

(50) ‘Well‐Being, Agency and Freedom’, 201.

(51) For relevant citations of Marx, see ‘Development: Which Way Now?’, in Resources, Values and Development, 497; ‘Goods and People’, 512; The Standard of Living, 37.

(52) ‘Economics and the Family’, 376. The sentence continues: ‘who can do what, rather than who has what bundle of commodities, or who gets how much utility.’ My point is the simple one that what people can do with their commodities is not identical with what their commodities (can) do for them.

(53) ‘Capability and Well‐Being’, 38.

(54) Ibid., 41.

(55) Ibid.

(56) Ibid., 6.

(57) Ibid., 41.

(58) There is further evidence of the persisting ambiguity at pp. 17–19 of ‘Capability and Well‐Being’. At pp. 17–18, and in line with the definition of capability on p. 5, a capability set is characterized by the ‘various alternative combinations of beings and doings any one (combination) of which the person can choose’ (italics added). But at pp. 18–19, it is allowed, in seeming contradiction of that characterization, that the realized combination in a capability set may or may not be chosen. What is ‘achieved’ might not be ‘achieved on the basis of choice’ (p. 19).

(59) ‘Equality of What?’ 220.

(60) ‘Well‐Being, Agency and Freedom’, 197 (italics added), 226. See further ibid., 195, where there is an implied identification of ‘having “well‐being” ’ with functionings. Cf. Commodities and Capabilities, 25, 51; The Standard of Living, 16.

(61) And often it is rather unreliable evidence, since people tend to adjust to adverse conditions: see pp. 16–17 above.

(62) Commodities and Capabilities, Preface. Cf. ibid., 52. (Strictly, opulence is a magnitude which supervenes on command of goods: see ibid., 58).

(63) ‘Well‐Being, Agency and Freedom’, 197 (italics added).

(64) Ibid., 200.

(65) ‘Rights and Capabilities’, 318.

(66) That was the title of section 4 of ‘Equality of What?’

(67) ‘The Living Standard’, 87.

(68) ‘Equality and Equality of Opportunity for Welfare’, 93.

(69) ‘Rights and Capabilities’, 319. Not only does Sen allow strength of desire to condition capability evaluation, but, somewhat curiously, he is also disposed to classify happiness itself as a functioning. ‘Being happy’ is described as a ‘major functioning’ at p. 13 of ‘Well‐Being and Agency’ (unpublished, 1987) and as a ‘momentous’ one at p. 200 of ‘Well‐Being, Agency and Freedom’. See also ‘Goods and People’, 512; The Standard of Living, 8, 11, 14. See, too, Commodities and Capabilities (15, 52) for more tentative statements of happiness's credentials as a functioning.

(70) ‘The Living Standard’, 85.

(71) See ‘On the Currency of Egalitarian Justice’, 920–1.