On the Status of Equality Author(s): Amartya Sen Reviewed work(s): Source: Political Theory, Vol. 24, No. 3 (Aug.

, 1996), pp. 394-400 Published by: Sage Publications, Inc. Stable URL: http://www.jstor.org/stable/191919 . Accessed: 09/04/2012 03:24
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp 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.

Sage Publications, Inc. is collaborating with JSTOR to digitize, preserve and extend access to Political Theory.

http://www.jstor.org

ON THE STATUS OF EQUALITY

AMARTYA SEN Harvard University

1. INTRODUCTION The belief thatthereis a close connectionbetween equalityandjustice is quite common. Since justice is a quintessentiallyethical idea, that linkage clearly must have some ethical status. But what is the basis of this status9 On this subject, presumptions diverge-explicitly or by implication-even those who firmlyassociateequalitywithjustice. Some see the linkage among to be mainly political; others take it to be more than that-perhaps even since politicalperspilogically necessary.This distinctionis not unimportant would haveto be contingenton the ideasandcommitmentsof thepeople cuity involved, whereas logical necessity must be independentof that. I want to discuss the difference between the two approaches-the political and the logical-and to examinebrefly therespectivedomainsof analyticalrelations and political contingenciesin connectingequalitywith justice. The immediateoccasion for this exanunationis an interestingand importantpaperby JohnKane,in whichhe disputestheview thatthereis any logical connection between justice and equality.' Kane is not alone to have questioned the presumedlogical link, and this makes it all the more important to addressKane'scogent reasoning.I arguethatKaneis partlynght, buthe does that not give the analyticalrelationsthe due acknowledgement even his own line of reasoningwould require. However, before I proceed further,I must comment on an ambiguityof perhaps even an error-in Kane's interpretation my own position in This is importantto do not just to Reexamined,on this subject.2 Inequality set the recordstraight,but also becausethe substantiveposition I shall argue
AUTHOR'SNOTE:For helpfuldiscussion on the issues in this article, I am grateful to Sudhir are Anand,John Rawls, and ThomasScanlon.Thanks also due to the MacArthur Foundationfor researchsupport. Vol. 1996 394-400 POLITICAL THEORY, 24 No.3, August ? 1996SagePublicatons, Inc. 394

Sen / ON THE STATUSOF EQUALITY

395

for is a developmentof whatI hadproposedin thatmonograph. this sense, In the next section has both a correctiveand a constructivepurpose.In the later sections, I proceed to distinguishbetweenthe respectivedomainsof analytical relations and political contingencies, and also to comment on Kane's substantivearguments.

2. POLITICAL AND BASAL PLAUSIBILITY EQUALITY Kane notes that"thepresumption some necessaryconnectionbetween of the ideas of justice and equality is endemic as Sen and I both believe." But he also contends-suggesting stronglythatthereis a divergencebetween our views here-that "sucha presumption in fact mistaken."3 had not, in is I claimed that one must presume that there is a necessary connection fact, betweenjustice andequality.Rather, hadnotedthat"everynormativetheory I of social arrangement has at all stood the test of time seems to demand that equality of something-something thatis regardedas particularly important in that theory"(p. 12). I did try to clarify whatexactly I was claiming:
We may be persuaded that the basic disputationsare likely to be about "equalityof what?",but it might still be asked whetherthereneed be a demandfor equalityin some importantspace or other. Even if it turns out that every substantivetheory of social in arrangements vogue is, in fact, egalitaran in some space-a space seen as centralin that theory-there is still the need to explain and defend thatgeneralcharactenzation in each case. The issue to address is not so much whether there must be for strctly formal reason (such as the discipline of "thelanguageof morals"),equal consideration for all, at some level, in all ethical theories of social arrangement. I am more concerned with the question whether ethical theories must have this basic feature of equality to have substantiveplausibilityin the world In which we live.4

In the formalstructure hadtriedto develop, equalityof some type-what I was called "basalequality"-is soughtin each such substantive ethicaltheory by arguingfor (or by simply assuming)thepnmacy of some space-incomes, utilities, libertariannghts, and so forth, in the respective actual ideas of justice. The endorsementof the importanceof equality in thatcentralspace (i.e., basal equality) then yields the need to accept inequalityin otherspaces, since these differentequalitiescould not be simultaneouslyguaranteed. (The extensive diversities of humanbeings-a majortheme of my book-makes the congruence of equalities in the differentspaces extremelyunlikely.)For example, those who seek equality of incomes cannot also demandequality of libertarianrghts of ownership, transactions,and bequeathal-and vice

396

THEORY/ August 1996 POLITICAL

in versa."Equality whatis seen as the 'base'is invokedfor a reasoneddefence of the resultinginequalitiesin the far-flung'periphenes.'"5 I hadnot claimedthatit would be logically impossibleto constructa moral theory that would not demand equality in any space whatsoever, nor in particularattach any importanceto equality in some space that is seen as to "central" that moral theory.But in confining my attentionto the class of theories that do value equality in some space, I did emphasize the political force of having "equalconcernfor all at some level"-a level thatis seen as quite centralin that theory (e.g., incomes for Income-egalitanans,liberties and for libertarians, so on).6Political diversitiesof our beliefs permitsome variationin the choice of space in which equality is to be "basally"valued, but they do not easily accommodaterejectingthe need for equalityin every possible space. I did try to read some significance in the fact that "every that normativetheory of social arrangement has at all stood the test of time seems to demand equality of something-something that is regardedas in but important thattheory," the readingwas primarilypolitical particularly claim being made thatthingscould not have been otherwise,or (withoutany thatmere logic would make this "testof time"analyticallyinescapable). To seize the structureof this argument,it is useful to note thatI was not only acknowledgingthe practicalforce of the demandsof equality in some chosen space that is seen as central,but also-the other side of the same need to accept(and,in termsof consistency,contincoin-the corresponding gently require)inequalityin otherspaces thatare seen as less central.There is no necessity in all this of disputing Kane's claim that "justicerequires different things in differentcircumstances,and no a pnon judgments as to are equality or inequalityof treatment justifiable"(p. 389). Indeed,even as mustsurelygo well beyond faras "basalequality"is concerned,the argument the a priori,and well beyond the acceptanceof the need to seek equality in haveto be presentedas to why a particular some space. Substantive arguments ratherthananothershould be seen as centralin thatethical context.7 space

PROPORTIONALITY 3. ARISTOTELIAN While the need to have substantiveargumentsgoing beyond theform of equality is not a bone of contentionbetween us, Kane does, in fact, offer a differentlyformalizedview ofjustice, relatedto "theAristoteliandoctrne of He proportionality." contraststhis formalprnciple with the rule of equality in some space, in outliningwhat he believes the basic conception of justice to be:

Sen / ON THE STATUSOF EQUALITY calculatedis The pnnclple upon which such rghts, penalties,or duties are traditionally the the Anstotelian pnnciple of proportionality: more one has of a certainpropertyX, Y the more one deserves of whatevertreatment is relevantto it. (p. 378)

397

I had, in fact, briefly discussed this prnciple, in terms of particular uses, in InequalityReexamined(pp. 90-2), treatingit as a type of equality of treatment. Since Kane sees equality and proportionalityas two contradictory formal principles, it is worth discussing in what sense they are or are not really contrary. In fact, a kind of formal equality, with well-defined analytical requirements, is implied by any rule of the type of the proportlonality principle:in this case, thatdifferentpersons (no matterwho) thatsharethe same value of the centralpropertyX must be seen as deserving the same value of Y. The basic point here is that the Y that is "due"(to use Kane's terminology) to a person is supervenlenton his or her X, and not on who the person is (or on any othercharactersticof this person).The generalnatureof treatingpeople equally in this sense was extensively discussed in InequalityReexamined, and It applies inter alia to the proportionality prnciple as well. (as Second, it so happensthatif Y andX arestrictlyproportional the name of the rulesuggests andsome versionsof the Arstotelianclaim endorse),then not only does everyonewith the sameX get an equal Y,butalso everyone-no matterwhat X he or she has-gets an equal amountof Y per unit of X. In fact, however, despite the name, it is common to consider monotonlc functions f(X) otherthanproportlonality applyingthe so-called proportlonality in princlple. This involves an additionalproblem of rescaling to characterze the sense in which the "due" values of Y for respective values of X are The supposed to be determinedby the principleof "proportionality." idea of would relateto the "dueness" functionconnectedwith that "equaltreatment" scaling. I shall not pause here to elaborateon this (the relationcan be dealt comwith, mathematically,in similar ways to the problemof interpersonal The crucial issue is the equal treatmentof people in identical parability).8 circumstances (taking, in this case, the relevant circumstance to be the value of X). The fact thatthe proportlonality falls well within the generalcategory rule of egalitarianrules need not come as a greatsurprise,since the formalaspects of equalitycan be very versatile.This is not to deny thatthereis an interesting substantivequestion as to the form in which equalitymay be most plausibly view of sought. Indeed,in InequalityReexamined(91-3), the proportlonality in having one's Y in line with one's X ("equaltreatment" relatingY to X) was contrasted,in this respect, with the rule of working for the same Y for all in ("equaltreatment" giving everyone equal Y). A substantiveexamination

398

POLITICAL THEORY/ August 1996

of "equalityof what"has to addressthis question along with the choice of space. But what is at Issue is not whether or not to have any general rebut should quirementof equalityof treatment, the form thatequal treatment take. Any principle of the kind that Kane himself discusses thus incorof and poratesa requirement "equality," the only remainingissue is the more Reexamined addressed),namely,equalwas general one (to whichInequality ity'of what?

OF 4. REACHAND LIMITS NECESSITY So what do we conclude?As a substantiveproposition,the requirement kind (of incomes, of utilities, of freedoms,etc.) of equality of any particular cannot,of course, be a logically necessarypartof justice underevery possible of that conception. Even the requirement demandingequalityof "something in is regarded as particularlyimportant" the respective theory of justice cannot be fully analytical.If it is indeed the case (as I have triedto claim it that has at all stood is) that "every normativetheory of social arrangement thatcommonality the test of time" does, in fact, satisfy such a requirement, denves largelyfromthepoliticalregulantlesthathaveemergedover certainly time in the beliefs and convictions of people. For this reason,I see no reasonto disagreewith Kane'smainclaim (except with his implicit suggestionthatI hadarguedfor the contraryview; this I had that not done). But thereis a further issue, to wit, the formalrequirements are rule of justice. Any delineation of principles-of imposed by any general needs, of deserts, of entitlements,or whatever-makes the right outcomes supervenienton some specific charactensticsthat a person may possess or acquireor receive, and this cames with it the entailmentthatanyonewho has the same charactenstics-no matterwho he or she is-must receive the same is consideration.That "equalityof treatment" nonspecific, but it does incora substantivedemandfor equality,in a parametrc form. porate This applies to all the classes of justificatoryargumentsthat requirethe use of such rules and regularntes.Kane does not say this, and indeed in his citation of Anstotelian "proportionality" rules, he seems to be denying any connection of such prnciples with equal treatmentat all. But this is largely a matter of presentation.Consider, for example, Kane's argument:"If a of teacherpunishesa whole class forthe misbehaviour one pupil,or an officer shoots an entire village for an act of sabotageby one of its members,surely of the onus is on the perpetrators the equal treatmentto say why they have made no distinctionsbetweenthe guilty andthe innocent"(p. 389). The same

Sen / ON THE STATUSOF EQUALITY

399

substantivequery can be raised by asking the offending teacher(or police officer) why the princlpleshouldnot be thatall misbehavers(or all saboteurs) be treatedequally,ratherthanthatpeople with dissimilarresponsibilitiesbe is given the same punishment.WhatKanecalls "theequal treatment" not the can form thatequal treatment take. only We can still question whetherthereis any necessity to have these general that in classes ofjustificatoryarguments takeus to equalityof treatment some sense (e.g., a person should get this because of some characterstic generic he or she has). Indeed,one can arguethatthe necessity of equalityin classes of substantivemoraltheores was not in itself a necessity. I do not see much merit in spending effort to resist this claim. I have no great difficulty in conceiving of a putativetheory of justice thatremainscompletely arbitrary in its allocation, without any use of a principlethat involves supervenience on some generalcharactensticsof persons (ratherthansimply who is who). I will, of course, resist such a theory of justice, but not on the ground that there is something of a logical mistake here. Its arbltrannesshas to be comparedwith the reasons-plausible to us-that can be given in favourof equal treatmentof one kind or another,and this fits into the discipline of examining a substantivetheoryof justice in political philosophy. The really engaging questions about "the status of equality" concern (a) the choice of the space and the form in which equality is to be sought ("equalityof what?"),and(b) addressingconflicts betweenthe claims of any particular type of equalityand otherconsiderations,includingothertypes of The status of equality is a equality ("whatrelative significance to attach?"). substantiveratherthan a formalissue.

NOTES
1. John Kane, "Justice,Impartiality, Equality:Why the Concept of Justice Does Not and PresumeEquality," Political Theory24, no. 3 (1996, this issue): 375-93. 2. AmartyaSen, InequalityReexamlned(Oxford:ClarendonPress;New York:Russell Sage Foundation; Cambrdge, MA. HarvardUniversityPress, 1992). 3. Kane, "Justice,Impartiality, Equality," and 376. He does not explicitly attribute view this to me (as far as I can see) but proceeds as if I believed in a logically necessaryconnection.He comes closest to attribution the following passage: in Sen suspects that the reason why so many altogetherdifferentsubstantivetheones of social justice have the commonfeatureof demandingequalityof somethingis to be found in a "foundational prnciple" of ethicalreasornng, namely,the pnnciple of equal concern for all at some level. The absence of such concern, he says, makes a theoryarbitrarily discnnmnating. (pp. 375-6)

400

POLITICAL THEORY/ August 1996

But the question to ask here surely is: what lknd of arbitrarnesswould this be-political or logical or both? 4. Sen, InequalityReexanuned, 16-7. I had, m fact, gone on to argue: "A theory may accept-indeed demand-inequality in termsof manyvariables,butin defendingthoseinequalities it would be hardto duck the need to relatethem,ultimately,to equal consideration all in for some adequatelysubstantialway" (p. 17). 5. Sen, InequalityReexanuned,19. force relatesclosely to the idea thathas been illummatingly ?6. Thatargumentative explored by Thomas Scanlon of the need to have "reasonsthat others could not readily reject."Idem, and in "Utilitanamnsm Contractualism," Utilitaranism and Beyond, ed. Amartya Sen and BernardWilliams(Cambridge,UK. Cambrdge Umnversity Press, 1982). 7. Kane simply asserts that the "caste system m Indiais formallyjust," and in this context cites MichaelWalzerto the effect that"ina society wheresharedsocial meaningsare integrated and hlerarchical,'justicewill come to the aid of inequality'" (p. 390). But politicalmovements challenging this view have been, in fact, dominant in contemporaryIndia, and even the constitution of the Indian Umon (adopted on the basis of a consensus immediately after disindependence)bans various practices of caste-basedinequalityand requires"affirmative crimnations"m favor of lower castes in civil service and education(provisionsthathave been furtherexpandedin the last decade). The discipline of political plausibilityis not without its substantivebite. 8. On this, see my CollectiveChoice and Social Welfare (San Francisco:Holden-Day,1970; North-Holland,1979), chaps.7 and 7* republished,Amsterdam:

ProfessorAmartyaSen is LamontUniversityProfessorandprofessor of economicsand philosophyat HarvardUniversity.