Amartya Sen's Unequal World Author(s): G. A. Cohen Source: Economic and Political Weekly, Vol. 28, No. 40 (Oct.

2, 1993), pp. 2156-2160 Published by: Economic and Political Weekly Stable URL: http://www.jstor.org/stable/4400230 . Accessed: 31/12/2010 12:09
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=epw. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to Economic and Political Weekly.

http://www.jstor.org

Amartya

Sen's

Unequal

World

G A Cohen In Section I of this article the authordescribes the leading idea- 'capability'-which AmartyaSen has broughtto the discourse on problems of equalityand its absence in his book InequalityReexamined. He then takes up one of the book's sub-themes,the connection or lack of it betweenfreedom and control. In thefinal section he defends Sen against some scepticism about the practical relevance of his workexpressedin thisjolurnalafew monthsago.
but the first question, aboutthe appropriate type of advantage to focus on, dominates Inequality Reexamined. Accordingly, the present"reexamination"is not, as the titles of the two books might suggest, precisely and centrally of the topic of On Economic ITheearlierwork,one could say, Inequality. has a comparativeadvantage(in the strict, Ricardian,sense) with respect to the interests (as opposed to the needs) of economists, while the recent one has a comparative advantagewith respect to the interests andneeds of political theoristsandphilosophers. Now, although Sen's official topic is inequality, his motivatinginterestis poverty, which appears,when it does, at the downwardend of the spectrumof advantage,and which is a phenomenon distinct from inequality, since everyone might be equally poor, and since there is (at least) money I inequalitybetweenmillionairesandbillionTwo questions arise with regard to the aires.His specialinterestin povertyis shown measurement of inequality. The first.con- in Sen's choice of capabilityas the premier cerns the respecz(in economists' language, space of advantage:capability provides a the space) in w'ch people should be ac- highly suitable measure of the deprivation counted equal cG unequal:what is the right that poverty imposes, but it is not so evitype of advantageto examine whenequality dently serviceable when the object is to and its absence are at issue? Representative identify degrees of inequality as such. answers to this first question are: utility Let me explain.WhatSen calls "capabil(some economists and some philosophers), ity" is determinedby the differentforms of income (other economists and no philoso- life thatarepossible for a person:a person's phers), primary goods (John Rawls), and capability is a disjunctionof the combinaresources, capaciously conceived (Ronald tions available to her of what Sen calls Dworkin).The issue was broaclhed by Rawls "functionings", which arestatesof activity in A Theoryof Justice (1971), which argued and/orbeing. These functioningsvary, Sen thatnot utility or welfare butprimarygoods says, (things everyone needs to pursuetheirgoals from most elementary ones, such as being in life, whateverthose goals may be) constiwell-nourished, avoiding escapable morbidtute the right metric for distributivejustice. ity and premature mortality,etc, to quite But it was in Sen's 1979 Tanner Lecture, complex and sophisticated achievements, suchashavingself-respect beingableto take called 'Equality of Wlhat?',that that quesandso on.2 partinthelife of thecommunity, tion was first put in an expressly general listed form, and it was there thatSen venturedhis Now, the "elementary" functionings now quite well known thesis thatcapability here have ceilings of accomplishment:you is the right thing to look at when judging can get richer and richer, but you cannot how well a person's life is going. keep on rising in the dimensionsof nourishThe second andindependent questionwith ment and health. And something similar is regard to the measurementof inequality is true of the more ''sophisticated' how to compute the degree of inequality functionings:you cannotkeep on addingto that obtains, for given sets of scores of your stock of self-respect. and there is necadvantage (whatever may be the right re- essarily a limit to how muchyou can, or can spect in which to reckon advantage,which want, to take partin the life of the commuwas the first question). That second ques- nity. Theserecurrently cited examplesshow tiondominatedSen's OnEconomnic Inequial- that wlhatSen really cares about is basic ity (1973), and, although it is briefly ad- capability,3 the prerequisite of adequate dressed in the work underdiscussion, not it functioning,which, he rightly complained, INEQUAL17-Y REEXAMiNED exhibits (often, perforce,only in fleeting cameo) the current state of Amartya Sens's decadeslong engagement with problemsof equality andits absence. The book provides not only an exhilarating tour d'horizon of ideas developed at greater ease elsewhere, but also fresh nuances that are designed to accommodate and deflect some of the extensive criticism and comment which Sen's magnetic work has attracted. In Section I of this article, I describethe leading idea-'"capability" -which Sen has broughtto this field of discourse.I thentake up one of the book's sub-themes,regarding theconnection or lackof it betweenfreedom and control.Finally(in Section111), I defend Sen against some scepticism abbutthe practical relevance of his work that was expressed in these pages a few months ago. was a form of advantage neglected in the literature,despite being themost fundamental one of all. Capacities beyond the basic (can I run a mile? can I impress Ukrainians with my impersonationof Russians? can I sew more quickly than you?) seem quite irrelevant to measurement of deprivation, inequality or anything else of urgent concern from the point of view of justice. (This point, that poverty is the key theme of the book, will prove to be consequential whenI come to comment in Section III on some criticisms of Sen made by Andre Beteille.) Whether or not capability deserves its assignedrole as a metric of advantage,Sen's very identification of the capability dimension of assessment was impressive, in the light of its previouscomplete neglect. Capability lies, causally, between incomp or primarygoods or resources on the one hand andutility or welfare on the other. Focus on capabilitymeans emphasising not goods as such, but what they enable a person to do, andit also means disemphasising the (often vagariouslyinduced) utility associatedwith his doing it. The trouble with a metric of goods or resources or income is that the point of goods (and so forth) is to generate possibilities of choice for the individual: much better,then, Sen argued,to look not at their generators but at those possibilities themselves, which do not vary uniformly with what generatesthem, because of variations in people's physical (climatic, topographical, etc) and social circumstances, and in their biological constitutions. And the troublewith a metric of utility is that it is blind to the fact that people adjust their expectations downwardlywhen in poverty and upwardly when in wealth. This and othersubjectivevagaries mean thatutility is not the rightquantityto focus on: it is unfair to a poor person to resource him less because he has developed modest tastes and tlerefore needs less wherewithalto achieve a given level of welfare. What mattersis the causal intermediary,the effect of goods that causes utility: functioning, and capability, as such.4 A person's functionings matter because they are his life, considered apartfrom the utility he gets out of it. And capability matters at least instrumentally, since functionings matter, and adequate functioning can obtain only if it lies within a person's capability set. But capability also matters in tlhreeotlher ways. Freedom to

2156

Economic and Political Weekly

October 2, 1993

-chooseis good in itself, apartfromthegoods it provides access to; freedom to choose with adequatefunctioning within the scope of thatchoice is a person's right; and capability also contributes directly to well-being, becauseaperson's life is "richer" when the "opportunityof reflective choice." appearswithin it: capability is good not only, then, as a space of choice, but also because free choosing, a process thatrequiressuch a space, is itself a good (pp 40-42, 51-52). Capabilityis a form of freedom, the freedom, specifically, to choose a set of functionings. When the value of that freedom is measured in terms of the forms of well-beingthose sets of functioningsconstitute, then what Sen calls a person's "wellbeingfreedom" is displayed.But well-being freedom is not freedom as such. It must be from"agencyfreedom",which distinguished is a person's freedom to achieve whatever goals she has, including goals otherthanher own well-being. It diminishes a persons's agency freedom that she cannot pursue a cause to which she is committed, but the restriction need not commensurately detractfrom her well-being freedom. The capability focus makes a difference tothe analysisof poverty,and,consequently, to anti-poverty policy. Thus, for example, women often require, for biological, social and cultural reasons, higher income than men do to secure the same capability; and factorssuch as age, location, andepidemiological atmosphere also strongly affect a person's power to convert money into the elements of a worthwhile life (p 113). Accordingly, the plea for attentionin poverty analysis not to low income as such but to income inadequateto sustain basic capability in given circumstances has considerable practical significance. To be sure, it may sometimesbe too difficultand/ortooinvasive fora stateauthorityto identify eachperson's andtailorhis orher capabilityrequirements, income support to them. But, even then, it may be possible to identify aggregatedifferences between sub-populations which affect the convertibility of income into capability, differences whichjustify unequalper capita funds dispersal across regions and groups. And apartfrom the sheerly quantitativequestion of what income is needed, in different conditions, to generate a given amountof capability,attentionto capability desiderata suggestsmodesof intervention and whichsecuresubstantial enablement improvements for comparativelylittle expenditure.5 Thebenefitsof theshifttocapabilityshowthat "conceptsmatter"6 inpractice.

control over what happensby the relevant free agent. A supposed case in point is the freedom enjoyed by someone who lives in an environmentwithout malaria. Sen does not mean, by that freedom, the freedom to do things that can only be done when malaria is absent, for, in that consequent free-

domcontrolis manifestlypresent.8 He means the (supposed)freedomthata personenjoys just in thather environmenthas no malaria in it. This, Sen argues,is indeeda partof the person's freedom,even thoughshe does not (and cannot) control whetherthere will be malariain herenvironment.Sen used to call this (supposed) freedom, where control is missing, the freedomof power. Criticsprotested thatthe situationof a person benefiting from a salubriousenvironmentno more manifestsherpowerthanit does hercontrol. Sen now accedes to this criticism, dropsthe word "power", andcalls his theme "effeetive freedom". Sen has here, once again, identified an undoubtedlyimportantand neglected phenomenon, but it is not freedom without control. In this section, I show that Sen's argumentfor the existence of such a freedom is unsuccessful, and I then indicate what I think is the true shape of the importantphenomenonthat he has discerned. The failure of Sen's argumentbecomes whenwe askwhy a personbenefitapparent ing from an environment rid of malaria qualifies in that respect as free. Sen vacillates between two different answersto that question, correspondingto two conditions, one strong and one weak, each of which he takes to be sufficient for "effective freedom". The strongcondition indeed identifies a form of freedom, but one in which, contrary to what Sen requires, control is present, however (literally) remote. The weaker condition indeed involves no control, but also, pari passu, no freedom.

II
I turn to my selected sub-theme: the relationshipbetween freedom and control. Senclaims7thatthereis a significant forn of personalfreedomenjoyment of which does not involve (as freedom usually does, andis

generally thought always

lo

do)excrciseof

Throughfailing to distinguish the two conditions Sen produces his fallacious result, that freedom can obtain without control over what happens. The strongercondition conjoins two elements, corresponding to the two phrases italicised by Sen in his statementof it: "As long as the levers of control are systematically exercised in line with what I would choose andforthat exact reason,my 'effective freedom' is uncompromised, though my 'freedom as control' may be limited or absent'"(pp 64-65). In illustrationof this point, Sen instances the relationship between me and a proof-readerof my book, who correctsthe text as he does because he knows I would want it to be correctedthat way.9 Now it is true, in this example, that "the levers of control" are not "directly operated" by me (p 64), but the fact that Sen emphasises "directly" betraysthatthey are indirectlyoperatedby me, which meansthat I do control what happens. I can be free

withoutexercising the levers of controlprecisely because I can control without exercising the levers of control. (When I tell my obedient chauffeur where and how to drive I do not exercise the levers which controlthe butI neverthelesscontrolwhatit does.) car10 Satisfaction of the strong condition indeed yields freedom, but not freedom without control. But Sen often uses a weakercondition for "effective freedom" or (supposed)freedom without control, which is yielded by preserving the first italicised element butdeleting the second one in his statement of the strong condition which I gave two paragraphs ago. The weaker condition is that whoever controls what happensdoes whatI would choose if I were in control, no matter for what reason, and, in particular,whether or nothe "knowswhatmy instructions would have been if sought" (p 64). So, forexample, it may conform to my will thatmy environment has been rid of malaria, even thoughI did not, and could not have, made it so, and even though those who did make it so were relevantly unmindful of my wishes in the matter. I believe that this phenomenon, in which things conform to my will althoughI do not exercise it, is of great philosophical interest, but that it is not freedom. An indication, apartfrom his actual formulations and examples, that Sen also uses the stated weaker condition for freedom without control is that, as I reported, he acceptedthe criticism thatfreedom without controlcould not be said to qualify as power. For satisfaction of the first condition, as illustrated in the proof-reader case, does entail power: I do not let the proof-reader operateunless I am satisfied that he will act as I want hiim to. (If the proof-reader is imposed on me willy-nilly," then at most the weakercondition holds, andfreedom of choice disappears.) Anotherindication that Sen also relies on the weakerconditionis his description of the relevant form of (supposed)freedomas "outcome-based"(p 135). For the second element in the strongercondition for "effective freedom" ("for that exact reason") is a matter of process, not outcome. Consider the malaria clearance case, in which the controlling agent is the state, or some state body. If the policy of malaria elimination is adopteddemocratically, then people, together, control what happens to them, and therefore exercise freedom and, for good measure, power: but Sen's claim, that there exists freedom without control, remains unillustrated.If, on the otherhand, the malaria clearance is achieved by an undemocratic(but, at least in this respect, benign) administration, or, as Sen at one point suggests (p 65), by an international agency underdistantdirection, then there is neither control nor freedom, but it remains true that what eventuates conforms to what peoplewouldchoose,andispro tanlo com2157

Economic and Political Weekly

October 2, 1993

mendable. Sen says that what he calls "effective freedom" is importantbecause, in a modem complex society, there is much that yve can secure not individually but only collectively.-2As thecontrastbetweendemocratic and other malaria clearance shows, thatis not a reason for concluding that.there exists freedom without control. We have to do, in sum, with two phenomena, one more general than the other, and only the more specific one involves freedom. The specific one, in.whichthingsgo as I will because it is my will, even thoughmy handsare off the levers of control, is politically importantbutphilosophically not very interesting.The more generalone, in which things go in accordance with my will, but truly without my control, is politically less (*, not) interestingbut philosophicallyvery interesting,althoughnotrightlycalled "freedom". I agree with Sen that the fact that a central aspect of a person's situation may conform to his will other than because he himself arrangesor sustainsthatconformity opens a "momentous perspective" (p 69). First, a bit more on why the perspective is not one of freedom. Then, an attemptto say why the perspective is neverthelessmomentous. Contraryto what Sen says, when a person gets an unchosen thing thatshe would have chosen, no "ability" on her pArt "to choose to live as [shel desires" is thereby indicated(p 67). "Ability" is here infelicitousinjust the way its cogrtate"power" was: abilities, like powers, are things that are exercised, and there is, ex hypothesi, no relevant exercise of anything in this example.'3 It is also false, in the pertinent sense, 14 that if peopledo desire a life withouthungeror malaria,the eliminationof these maladies throughpublic policy does enhance their "libertyto choose to live as theydesire"." "Abilitytochoose" and"libety tochoose" obtain only when it is possible to choose, andmuch of the interestof the phenomenon misdescribed in these dictions is that the agent has no choice in the matter (for example, of whetheror not there is malariain his environment).(Especially) when applying the weaker condition, Sen emphasises absenceof control, while insisting on liberty to choose, but liberty to choose entails control over what happcns. So, while I agreewith Sen thatthe issue of whetherpeople havewhatthey wouldchoose "is a momentousperspective",I do notthink thatwhen they have whatthey would choose they are pro tanto free. To see the true significance of the phenomenon to which Sen draws our attention, let us begin by distinguishing between a person's good and a person's will. Following a trafficaccident, my Christian scientist friend lies unconscious on the road. I must decide between taking'himhome, as I know h1e would wish me to, and taking him to hospital, as I thinlk 2158

would do him more good. I do what conforms to his will if and only if I take him home. A person's will is how he would makethingsgo if he could, whetheror nothe is in a position to make them go that way, whetheror not it would be good for him if they go that way, and whateverinterest or lack of it he has in his own good.'6 Notice that the Christian scientist's fate in this example conforms,thankstome, tohis will, but that his will does not "systematically" (Sen's word'7) determine his fate. It just happenedto be me who came along, rather than, for example, a thief, or a differentlymindedfriendwhomighthaveputtheman's welfare first, or not known about his will. Now, standardly,'8 when freedomis exercised, the agent exercising it aims to make the world (in the relevantrespect) conform to his will. But whatSen correctlynotes, and rightly emphasises, is that the world may conform to a person's will other than as a result of his control (or, as we can therefore safely add, of his exercise of a freedom). This is shownby theChristian scientist case, and,indeed, by a case in which, unlikewhat holds in thatone, the worldconforms to my will not only not as a result of my exercise of a freedom but not at all because my will has the character it does: when a public authority,perhapsacting for my good, but notout of respectformy will, deliversgoods that I would choose to get if I could. The real substance of Sen's innovative focus, then, is that the standard aim of exercises of freedomis achievablenot only through exercises of it but also by other means: a frienddoes whathe knows is your will because it is your will, a benign (or otherwise) agency does what happensto be your will. In these cases, the necessary effect of a successful exercise of freedom, conformity of the world to the will, occurs without any such exercise. But althoughthe standardeffect of a successful exercise of freedom, world/will conformity,is thereby wrought,it is not thereforeright to call the personwhose will is satisfied free, or to say that he has freedom without control. The Christianscientist is not more free in being at home than he would be if he were in hospital, though his will is better satisfied. Freedom,the "ability to get what we value andwant" (p 64), obtainsonly whenit is the agent who secures the conformity of the world to its will. My will is how I would make things go if I could do so. If they go thatway withoutmy intervention,then,except in special cases, I will unambivalently welcome thlat. (Special cases areones in which it mattersto me that I be the person who secures what I will: I want it to be me, not someone else, who nurses my spouse back to health.) And the malariacase is not a relevantlyspecial one: I shall not feel thatI have missed an opportunity to eliminate noxious mosquitoes if thlegovernmentdoes it for me (or even if it

does it not strictly for me). There are two

values associated with successful exercises of freedom. One is that the world conforms to my will and the other is that it is I who achieve that result. Sometimes, the second value does not mattermuch, andthe malaria example is a case in point. Thereis apolitical reason why Sen insists on the phrasings that I have stigmatised as inappropriate. "Freedom", he says, "is one of the most powerful social ideas" (p 69), and he is therefore concerned (so I hazard) to prevent ideological enemies of state interventionfrom obscuring the fact thatfreedom is among the benefits that such intervention can bring. Right-wing ideologues regardall state intervention as diminishing freedom (whetheror not they concede thatit might bejustified on othergrounds).Against that, I agree with Sen that freedom is pro
tanto enhanced when the state functions as

an instrumentof the democratic will. But what should be said to the right-wing ideologue who finds no freedom in a malaria clearancewhich is not democratically instituted?Not what Sen would say to her, that she is blind to the fact that there is freedom heretoo. Ratherthis: thatall or most of what would make this situation valuable if it did representan exercise of freedom is present here.The ideologue is blind, specifically, to that. She so makes a fetish of freedom that she fails to notice thata largepartof its value can be present when freedom itself does not
obtain.19

III
I remarked in Section 1 thatSen's animating concern is not so much inequality as poverty,andI indicatedsome ways in which a reorientationfrom both income andutility to capability both helps us to appreciate what the central evil of poverty is and has practicalforce in the struggle against it. It is the aspect of practice on which I shall concentrate in this closing section, which is a reply to AndreBeteille's review of Inequality Reexaminedin the Economic and Political Weeklyof April 17, 1993. Beteille called his review 'AmartyaSen's .21 I think that phrase was both Utopia' ungenerousand unjustified.It suggests that Sen is a head-in-the-clouds theorist yearning after an unattainableegalitarianideal at the expense of what can actually be done to improve the world. Now, that is certainly false at the level of aspiration.In Poverty and Famines, Sen confessed himself "immodest enough to believe that the analysis presented . this monograph has a certain amount6f relevance to mattersof practical concern" [Sen 1981: xi], andthe final paragraph of the book under discussion avows thatits "analysi.s has been verysubstantiallys motivated" by its ' directbearingonmatters of practical concern" (p 152, and see, too, p 11).

F nf me and PoliticskiW'kly

October 2. 1993

I have indicated, sketchily (see the final paragraphof Section I above), how that bearing goes, and thereby how the aspiration to practicalrelevance is in some degree fulfilled, with respect to poverty in general. Butconsider, for a moment, Sen's extensive workon famine. Before I had readany of it, I found what I knew to be its leading proposition puzzling, not because I thought it false, but because it seemed all too obviously true: that famines occur when and because people lose their customarylegitimateentitlementto (sufficient) life-sustaining matter. One may satisfy oneself of the truthof thatgeneralisationaftera shortbout of clear-headed armchairreflection. It is a truism,andnot, as Sen sometimes misleadingly suggests, something to be established by amassing and analysing observational data.21 When I came to readthe relevantwriting, I discovered thatthe intellectual interestof the stated truismlies precisely in its practical implications. For, properly applied, it defeats the unthinking presumption, still widespreadat least in countrieswhere famines areunknown,thatthey occurif andonly if, and because, food supply shrinks. In truth,the immediatecause of loss of access to food is, necessarily, the fracturing or shrinking of individual entitlement to it throughwages, trading relations, personal production,etc, a fracturing or shrinking sometimes,indeed, caused by sheer shortage of food, but often occurring in the face of continuedphysical availability of food, because of price shifts, vagaries of wages, and unintelligent or callous public policy. The identification of entitlement as what finally matters implies that (1) food shortage need not be pivotal with respect to famine (and-this is an empirical fact, of enormouspractical import-in several major faminesit has not actuallyoccurred)and that(2) whetheror not food shortageobtains in a given case of famine, famineprevention and relief requires identifying, sustaining and repairing lines of entitlement whose decay is what actually causes death.22 Now, it might be said, in defence of Beteille's characterisation of Inequality Reexamined,that this particularbook is not about famines. But it does contain some discussion of the topic, and in any case no one should be charged with Utopianism becausehis more practicalwork happensto be laid out in places other than the one that misgeneratesthe stated charge. To call Sen Utopian on the basis of this book is like complaining that a physician's manual of diagnostics tells us little about how to cure; is anyhowbeside the point, andwhich wvlichi is doubly unfair when the same physician has also written a manual of therapeutics. But let us turn from poverty and famine eliminationto the less urgentissue of equality itself. If someone whlofavours less inequality offers a circumspect accoun)t of
Economic and Political Weekly

what equality is, does it follow, that he Utopianlycondemnsanythingthatfalls short of it? If someone explains what it would mean for the Augeanstables to be perfectly clean, is it pertinentto stigmatise him as a Utopian because it is not possible to eliminate all the dirt? There is no rhapsodic depiction of complete equality in this book or elsewhere in Sen's work, and no affirmationthat such a thing is feasible, or even desirable (since increases in equality can reduce aggregate utility, income, and capability)." There is a cogent demand for more equality and an excellent identificationof whereandhow to start.Sen offers not a Utopia but a practicallyfruitfulcriterion. Beteille writes that those, like Sen, who dwell upon ideals tend to be alittle impatient aboutthe little constraints of the actualworldandit is thenthe obligationof the sociologistto bringthese constraints to theirattention.24 But Sen does not need to be advised that, for example, thereare structural tendencies in society to inequalities of status and power.25 Sen re-examinesinequality,buthe does not undertaketo say everything relevant to every aspect of that theme. He describeshis aims clearly in Chapter I of the book, and one cannot say that he does not have enough significant ones, and should therefore also have addressed the rather different issues which engage Beteille. Pursuinghis themethatUtopiais difficult to achieve, Beteille asks: Can the equal evaluationof persons and begenerated positions theconstructhrough tion of a social arrangement in which all personsand all positions will be equally esteemed,moreor less? (p 755) A negative answerto this question is less telling thanBeteille appearsto suppose.We can distinguish thtee types of inequalityof esteem. Thereis, first, the type thatreflects differentialachievements,which is not the same as inequality of esteem that derives from differencesin income andpower, and neitherof those is the same as inequalityof esteem sustainedby explicitly inegalitarian ideology which assigns people to different categories of quality or being. I shall address these threetypes of inequalityof esteem in reverseorder,with a view to removing the stingfromBeteille' s semi-rhetorical question. The thirdtype of inequalityof esteem is well illustratedby the Hinducaste system. Tenacious and savage though it is, it is surelynot this statusinequalitythatBeteille has in mindwhenhe askshis question,since he cannot believe that it is impossible, as opposed to difficult, to eliminate it. Proof that it can be eliminated is the fact that it does not exist in every society, and that analoguesof it disappearedfrom many societies where they were once strong: the anti-feudalbourgeois revolutionin Europe is a salientmacro-example.Nor is it myste-

rious how to proceed against it in incom; pletely bourgeois states like India: by outlawing and punishing the practice of untouclhability, by strugglingagainstdiscriminatory caste taboos, by instituting intelligently designed programmesof affirmative action. Caste inequalities of esteem differ from the two other types distinguished above, in that the latter supervene on other inequalities, representativeexamples of which are inequality of income on the one hand, and inequality of intellectual attainmenton the other. To the extent that it supervenes on income inequality,thereis no sensein pointing at status inequalityas a supposed trump card that the egalitarian carelessly or wilfully forgot was in the deck. For you could not say, on this basis, that even if the inequalities that egalitarians fix on (money, power,etc) were somehoweliminated, there would still be this other big one to reckon with. To the extent, finally, that inequality of esteem superveneson differential achievement, the only ways of suppressing it entirely areby restrictinghumanachievement throughdenial of equality of opportunityor by some awfully complicated and repugnant disinformation programme which spreads lies about people's accomplishments. Accordingly, no one should want to eliminatethe inequalityof esteem thattracks differential achievement. That inequality would appear in the best of all possible social worlds. An old right-wing notion says thatleftists seek to eliminate all significant inequalities, yet cannot hope to eliminate ones like inequality of esteem. But one of the signal merits of Sen's new book is its acknowledgement-indeed, its subtle and circumspect demonstration-that, because of humandiversity, equality in one dimension (e g, of opportunityto develop talent) means inequality in others (thus, here, of achievement and esteem).26So it is especially unfair to parade before Sen a proof that not all inequalities are eliminable (whether or not there exist less reflective egalitarianswho areless awareof thattruth). Since all equalities generate companion inequalities, we have to decide which ones to combat and which to tolerate. If we set aside caste differences, we can say that inequalities of esteem should be tolerated, for at least two reasons. First, the disjunctive one given above, thatthey can be eliminated only by telling lies and/or by suppressing equality of opportunity. Second, that they do not involve transferable resources: maybe you can prevent someone fromobtainingthe highiregardhe wouldget in the normal course of things, but you cannot take it from him and give it to somebody else. Consequently, the case for the (indep)endent27) injustice of status difis extremely weak, and it is thereferenltials
2150

October 2, 1993

pursuedin that article of this and other fore not the right inequality to combat. minordislocations in the conceptual scafUnder capability equality, and inevitably foldingthatSen haserected,or respond to different uses of it, inequality of esteem is Sen'streatment of itin Sen (1992a),or pass both unavoidable and acceptable. judgment on the extentto whichthe critithatwhile we knowthat Beteille remarks cism appliesto formulations offeredin the aredifferent actualsocialarrangements from book underdiscussion. the one [Sen] prefers,[he] has told us very 5 For strikingillustrations of this truth,see little aboutwhatwe oughtto do-or whatwe 8 of Inequality Section3 of Chapter Reexcan do-to bring the preferredsocial arainined. rangementinto being, and at what cost 6 'Do Concepts Matter?' is thetitleof Section (p 756). 4 of Chapter 7 of Inequality Reexamined. Sen paints no picture of the total social 7 As he didin earlieressays:see, in particuthat he would like to see, nor, arrangement lar, Sen (1982), Section5. afortiori (to this extent Beteille's comment 8 See fn 17, p 67, whereSen acknowledges is correct),does he tell us how to realise such that"freedoms thatresultfrom nothaving a thing. Sen's focus, in theory and in pracmalaria[are]notin dispute"betweenhim on tice, is particularevils. His concern is andactualandpotential critics.Thequeswith poverty and hunger, and one could tion is not whetheryou are freer in the hardlymaintain that he has said very little absenceof malaria (because youarefreeto aboutwhat we ought to do about them. The do thingsyou couldotherwise notdo), but detaileddiscussion in this book of appropriwhetherits absenceis itself partof your ate indices of illfare and welfare, and,more freedom. Unlessthisdistinction is grasped, so, the related discussions in Poverty and the criticism of Sen in this sectionwill not Famines, targetedon poverty in particular, be understood. for example in its Chapter 3, are entirely 9 It is notentirelyclearthatSen meansus to taketheproof-reader relevant to appropriateforms, levels, and examplein this fashion, but it must be so construedfor the means of delivery of public provision and phrase"forthatexactreason"to haveapsupport. If 'Utopian' means starry-eyed, plication. Sen is not. If it means wanting the world to be better,but having no idea how to make it 10 As opposedto theleverswhichcontrolthe chauffeur: my commands mightbe saidto so, then, again, that is not Sen. If it means be suchlevers. wanting the world to be better, then he is, 11 It is unclear-see note 9 above-whether commendably, guilty as charged. or notSen wouldregard thatas a relevant variant of his proof-reader example. Notes, 12 and see Sen P65, (1982), pp 216, 218. V KRamachandran [Ithank andArnold Zuboff for fine criticismsof an earlierversionof this 13 People do, of course, have abilities that they do not exercise,but I cannotsee how article.] theperson intheexample couldbe shownto 1 Sen himselfmakesclearat p 4. haveanabilityunlesssheis shownto have 2 P 5. Itis indicativeof Sen's primary focus exercisedone. on povertythatthe list of "functionings" 14 The qualification is necessarybecauseof given hereto illustrate thatconceptin genthe distinction madein note 8 above. eralis virtuallyidenticalwiththe list used 15 P 67, quotingBerlin(1969), p 179. to characterise elsewhere (p 1 O) as 16 The Christian poverty scientist'swill does aim at such. his own(eternal) good,butwe couldimag3 The sectionof 'Equalityof What?'introine a different in whichthe acciexample, ducingthenotionof capability wasentitled dent victim would wish to be elsewhere 'BasicCapability Thephrase has Equality'. than the hospitalfor self-sacrificingreaof thecharacterofan something oxymoron. sons. For people can all have basic capability 17 See the quotation fromp 64 above. withoutbeingequalin capability, income, 18 Thereareexceptionswhichhave no bearoranything else. "Universal basiccapabiling here:when(perhaps becauseI am actity" wouldbebothmorefelicitous andmore ingunder Ihopeto fail,orwhenI am duress) forwhatSen hasin mindwhen appropriate pickingindifferently betweenpossibilities he speaksof "basiccapability equality". and I do notcarewhichone is realised. 4 There are two powerful motivationsfor 19 I amalso unpersuaded by Sen's attempt to pointing to something other than either vindicatethe phrase"freedomfrom magoods or utility when concerningoneself laria"as fitting"intoa broad generalconwithegalitarian policy,butthe motivations ceptof freedom"(p 68) andnotjust signipoint at differentthings. lThereis good fyingabsenceof something undesired, like reasontolookatwhata personcan achieve, a cupboard thatis freeof dirt,butI shallnot of his actualstate;anidthere independently pursuethat relativelyminormatterhere. is good reasonnotto reducethe evaluation See Cohen(1992), pp 24-25. of thatactial stateeitherto anexamination 20 Andhegarnished thattitlewithBrowning's of his stockof resources ortoanassessment couplet: "Ah, but a man's reachshould of hisutilitylevel. Thesearedistinctpoints, exceedhis grasp/or what'sa Heavenfor?" and, so I complainin Cohen(1992), they 21 He frequently counterposes his own "enhave oftenbeen conflatedin Sen's presentitlement" conception to "foodavailability tation.I shall not hereresunme thecriticism decline" as thoughthey were competing

empiricaltheories. 22 See Sen (1981), pp 8, 78-83, 123-29 and Dreze andSen (1989), pp 26-27, for compendiouspresentation of these points. 23 See pp 7-8, 92-93, 136-47. 24 Beteille (1993), p 754. 25 Beteilleis, by the way, unclear,at least to me,as to whether hemeanstoemphasise the intrinsicor the causal importance of such "Inequality inequalities. is at bottom a matter of social esteem" (p 755, my italics) invites,givenitscontext,thefirstconstrual. Butunequalesteemis also described as "a far deeper source of inequalitythan unequalincome (p 755, my italics), andthat mustbe a causalclaim.Beteillemightwant to makeboth claims, but he shouldnevertheless distinguishthem.One can ask, rethecausalclaim,of whatinequality garding are inequalitiesof esteem (and, Beteille adds, power) "deeper sources" thanunequalincomeis? I findno hintof ananswer to thatquestionin Beteille's review.What he perhaps meansis notthatinequalities of statusandpowerareeitherintrinsically or causally more importantthan others but thatthey are especiallytenaciousbecause robustly self-reproducing. (Thisthird interpretation of whatBeteillesaysis suggested moreby the actualtruththanby Beteille's words.) 26 See pp ix-xi, 1-3, 19-21, 27-28, 129-31, 139. Beteille himself highlightsthis element in Sen's book: see Beteille (1993), p 753. 27 Independent, that is, of the injusticesit sometimesreflects,such as thatof grossly unequalincome: recall that the ladderof esteem that goes with caste is not under commentin this paragraph.

References
Berlin,Isaiah(1969): FourEssayson Liberty, OxfordUniversityPress,London. Beteille, Andre(1993): 'Amartya Sen's UtoandPolitical Weekly, pia', Economic April 17, 1993. Cohen, G A (1992): 'Equalityof What?On Welfare, Goods and Capabilities' in MarthaC Nussbaum and AmartyaSen (eds), The Quality of Life, Oxford: ClarendonPress. Dreze,JeanandSen, Amartya (1989):Hunger andPublicAction,Oxford: ClarendonPress. Rawls,John(1971):A Theory ofJustice,Cambridge,Mass:Harvard UniversityPress. Sen, Amartya (1980): 'Equalityof What?' in SterlingMcMurrin (ed), The Tannier Lectures on HumanValues,Vol I, Salt Lake City:Universityof UtahPress. - (1981): Povertyand Faminies: An Essayon Entitlemenitantd Deprivationt,Oxford: Clarendon Press. - (1982): 'Liberty as Control:An Appraisal', Midwest Studiesin Philosophy, Vol 7. - (1992a): 'Capabilityand Well-Being' in MarthaCNussbaumiand AmartyaSen (eds), The Quality of Life, Oxford:Clarendon Press. - (1 992b): Intequality Oxford: Reexam)inted, Clarendon Press.

2160

Economic and Political Weekly

October 2, 1993