Basedon thethreepremisesforeffective unreservedgroupsandthesedifferencesare

reservationpoliciesthatthis articlebegan likely to have implicationsfor the extent The Problem:

with,is therea case for OBCreservations to which students who enter through What Is at Stake in Mandal II?
in highereducation?Thelimitedavailable reservationssuccessfully complete their
evidence on averageoutcomesof OBCs higherdegrees. Ml It is hardly surprising that the recent
relativeto othergroupssuggestssomedis- decision to introduce 27 per cent reserva-
advantagerelativeto unreservedHindus, Email:[email protected] tion for the OBCs in elite institutions of
butthesedifferencesaresmallin compari- higher andprofessional education met with
son to thoseof the SCs andSTs. The large References such determinedand vociferous resistance.
numberof groups in a complex social Banerjee, Abhijit and Rohini Somanathan:'The Nor is it surprisingthatthe anti-reservation
hierarchyimplies a greatdeal of hetero- Political Economy of Public Goods: Some views that dominated the media described
geneityamongthe broadcategoriessuch EvidencefromIndia',forthcoming,Journalof the move as motivated by "vote bank
as the OBCs.This, togetherwith the fact Development Economics. politics" designed to benefit a particular
that some communities have, through Coate, Stephen and Glenn Loury (1993): 'Will caste-bloc which is also an electorally
Affirmative Action Eliminate Negative
effectivepoliticalmobilisation,been able Stereotype', American Economic Review, powerful constituency. Although in this
to comera disproportionate shareof state December, pp 1220-40. particularinstance the Congress Party on
resourcessuggeststhataffirmativeaction Deliege, Robert(1999):TheUntouchablesof India, the whole appearsto be more of a bemused
Berg Publishers,Oxford.
policies thattargetbroadsocialgroupsare Jaffrelot, Christophe (2003): India's Silent spectatorthana wily conspirator,the charge
notgoingto actas powerfultoolsof social Revolution: The Rise of the Lower Castes in may well be true in the larger sense.
justice - too many of the disadvantaged North India, Columbia University Press, However, to begin the story here is to begin
will be excluded in favour of the more New York. in the middle; but this suits the anti-
Mendelson, Oliver and MarikaVicziany (2000):
privileged.Lastly, alternativeaffirmative TheUntouchables:Subordination,Povertyand reservationists very well, for they would
actionpoliciesmighthavedifferenteffects the State in Modern India, Cambridge much rather forget the beginning.
on the effort levels of reserved and University Press, Chapter 1.
In the Beginning
The table shows where the "Mandal II"

Exclusive Inequalities story really began. It shows the number

of graduates and postgraduates (including
diplomas and other technical qualifica-
tions) in urban India who were identified
Merit,Caste and Discriminationin in a sample survey done by the National
IndianHigherEducationToday Sample Survey Organisation (NSSO).
Done in 1999-2000. this was one of the
"big sample" five-yearly surveys of the
This essay suggests that questions of merit, caste and NSSO covering the entire country, but the
discrimination in Indian higher education can be usefully data shown here are for urbanIndia alone,
since that is where higher education is
analysed in a framework defined by "exclusive inequalities". concentrated. The survey covered about
Beginning with a discussion of continuing caste inequalities 2.24 lakh people, which when adjustedfor
in higher education, the argument outlines the specificities the relative weight of the particular seg-
of this sector and its peculiarities in the Indian context. The ment of population covered, amounts to
idea of merit and the modalities of the examination are about 1.51 lakh persons. The first column
evaluated in terms of their contributionto the legitimationof lists castes and communities, while the last
column gives their percentage share in the
higher education. total urbanpopulation as estimated by the
survey. The middle columns give the caste-
SATISHDESHPANDE stake in the recent conflicts over other community-wise number (in normal let-
backwardclass(OBC)reservationsinelite tering) and the percentage share (in bold-
he mostacuteandextensiveexami- higher and professional education. italics) of graduates and postgraduates in
nation of compensatory discrimi- SectionII outlines the reasons why the the agriculturalsciences; engineering and
nation policies in independent specificitiesof highereducationrequireus technology; medicine and related fields;
India describes them as being framed by to think in termsof inequalitiesand ex- andall otherfields, which means the natural
"competing equalities" (Galanter 1984).1 clusion(ratherthanequalitiesandcompe- sciences, social sciences and humanities.
This essay argues that the notion of "ex- tition),andwhatimplicationsthis has for You need to read only one row - that
clusive inequalities" provides a compa- affirmative actionpolicies.SectionIIItakes for the "Hindu upper castes" (UCs) - to
rable heuristic framework for analysing upthekeyideasof meritandcasteandtheir get the basic story line. Looking at the
similar policies in the specific context of centralityto the ideologicalcontestations percentage share figures first, and reading
higher education in 21st century India. andthepracticalmanoeuvring goingon in right to left from the last column, this row
Section I begins to spell out this argument publicly funded higher and professional tells us that according to the NSSO, the
with a description of what precisely is at educationtoday. Hindu UC formed almost 37 per cent of

2438 Economic and Political Weekly June 17, 2006

This content downloaded from 131.172.36.29 on Sat, 09 Jan 2016 20:51:02 UTC
All use subject to JSTOR Terms and Conditions
the population of urban India. But they persons with engineering degrees, of which different and unequal in terms of their
accounted for almost 66 per cent of all non- as many as 908 belonged to Hindu UC; access to higher education. Since differ-
technical subject graduates, more than 65 the next highest figure is 202 for Hindu ences of this magnitude appear to have
per cent of medical graduates, almost 67 OBCs - less than one-fourth the former survived for so long after the "abolition"
per cent of engineering and technology figure. Similarly, Hindu OBCs account for of caste, thetablealso provesthatthese must
graduates,and about 62 per cent of gradu- 56 doctors out of the sample total of 535, be the productof durable,self-reproducing
ates in agricultural sciences. In sum, the but the Hindu UCs have as many as 350 mechanisms that are systematic (i e, not
Hindu UC are a little more than one-third doctors, or more than six times the OBC accidental or random) and systemic
of the total urban population, but around figure! Among the non-technical gradu- (i e, relating to system properties rather
two-thirds of professional and higher ates 11,529 out of 17,501 are Hindu UC, than to the attributes of individuals).
education degree holders: their share in the which is more than four and a half times In other words, looking at this phenom-
highly educated is about twice their share the figure for the OBCs at 2,402. The enon from another angle, it is clear that
in the general population. nature and extent of this dominance is there is something in the gate-keeping
We already know from this information reflected in the fact that the Hindu UCs mechanism which regulates entry into
that since the Hindu UC are heavily over- alone have roughly double the total number higher education that makes it discrimi-
represented among the highly educated, of graduates among all other castes and nate in favour of the "upper"and against
some other castes and communities have communities put together. And yet this the "lower" castes.
got to be under-represented.This is indeed measure may be an understatement be- However, the existence of inequality and
true for all the rows above the Hindu UC cause the table includes all types of edu- discrimination may be necessary, but it is
row - the Hindu scheduled tribes (STs) and cational institutions and all types of de- not sufficient to prove the existence of
scheduledcastes (SCs), Muslims andHindu grees from the humble BA upwards. If we injustice. For there are many kinds of
OBCs are under-represented among the were to do a survey of postgraduate and inequalities and of discrimination that are
highly educated relative to their share of professional education in elite institutions, considered just and desirable. Most people
the total urbanpopulation. Hindu SCs are it is a safe bet that the extent of Hindu UC believe, for example, that those who work
the most severely affected - almost 13 per dominance would be much more. more should be paid more, and vice versa;
cent of the urbanpopulation, they are less This is where the story of Mandal II i e, they believe in unequal pay for unequal
than4 percent in all fields, and only around really begins - in the undeniable fact that, work. (This is actually a corollary of the
2 per cent in engineering and medicine. It more than half a century after the formal popular slogan, "Equal pay for equal
is clear that Muslims and Hindu OBCs too adoption of a Constitution that explicitly work".) Similarly, Indian cricket fans
are severely under-represented in higher forbids recognition of caste (except, ironi- would hope and pray that the national
education in urbanIndia. On the other side, cally, to provide compensatory discrimi- selection committee systematically dis-
the rows below the Hindu UC are all over- nation to the lower castes), the dominance criminates in favour of more talented
represented,whetherslightly like the "other of the HinduUCs in Indianhighereducation players and against less talented players.
religions" (Parsis, Jains, Buddhists), or is still substantial, while the lower castes In fact, the word "discrimination" bears
quite significantly like the Sikhs and spe- and Muslims are significantly under- both good and bad meanings - prejudiced
cially theChristians.However, though over- represented. The story begins here, but, as or malicious bias, as well as discernment
representedin proportionalterms all these they say, this is only the beginning. or the ability to distinguish better from
communities are very small in absolute worse, etc. (The contrast with "indiscrimi-
terms and together account for under 6.5 From Inequality to Injustice nate" also brings this out.)
per cent of the urban population. By So there may be inequality and discrimi-
contrast, the under-represented groups - The table (or other evidence of this sort) nation behind the dominance of the Hindu
the Hindu "lower" castes and tribes (STs, surely proves that Hindu UCs and the UCs in higher education, but how do we
SCs, OBCs) plus Muslims - account for under-represented groups are definitely know that it is not legitimate or "good"
well over half (about 57 per cent) of the Table 1: Sample Number and Proportion of Persons with Graduate Degrees,
urban population. Indeed the caste divide NSSO 1999-2000
in urban higher education is even worse
Castes and Numberand Percentage Share of Graduatesin Various Caste/Comm
than what it looks like here because the Communities Disciplinesin the Sample Share of Total
table ignoresthe well known caste divisions UrbanPopulation
within the non-Hindu communities.2 Agriculture Engineering Medicine OtherSubjects (Per Cent)
But the story is not just about the caste HinduST 26 2.4 18 1.3 10 1.8 229 1.3 2.6
divide - the punchline is the overwhelm- HinduSC 41 3.8 30 2.2 10 1.8 629 3.6 12.9
ing dominance of the Hindu UC in higher AllMuslim 101 9.4 68 5.0 54 10.0 1,006 5.7 17.0
HinduOBC 108 10.0 202 14.9 56 10.4 2,402 13.7 24.2
education. This is broughtout very starkly HinduUC 669 62.1 908 66.8 350 65.3 11,529 65.9 36.9
by the figures for the sample numbers. AllChristian 90 8.4 70 5.2 35 6.6 707 4.0 2.8
Although the point is already made by the AllSikh 18 1.7 30 2.2 11 2.1 419 2.4 1.6
Allothers 25 2.4 33 2.4 10 1.9 581 3.3 2.0
percentage share figures, the sheer weight Total 1,078 100.0 1,359 100.0 535 100.0 17,501 100.0 100.0
of the Hindu UC is palpable when you
compare its absolute numbers with those Notes: (i) Figures in bold-italicsshow caste/communityshare of graduates in urbanIndia.
of the next biggest category, the Hindu (ii)Includespersons withpostgraduatedegrees. (iii)Cells show roundedand multiplier
weighted
sample numbers and proportions.(iv) Columns may not add up due to rounding.(v) Total
OBC. In the total sample (about 1.51 lakh unweightedand multiplierweightedsample sizes are 2.24 and 1.51 lakhpersons respectively.
persons in urban India), there were 1,359 Source: ComputedfromNSSO data on CD.

Economic and Political Weekly June 17, 2006 2439

This content downloaded from 131.172.36.29 on Sat, 09 Jan 2016 20:51:02 UTC
All use subject to JSTOR Terms and Conditions
discrimination? Ifthegate-keepingmecha- need much larger numbers than what exclusion by discrimination. However,
nism is favouringthe uppercastes, is it we currentlyhave, it is neverthelessin- despite the extensive domain of beliefs,
becauseit is doingitsjob well or because tuitively clear that only a very small values and norms that merit invokes, in
it is malfunctioning? Thisis essentiallythe (sometimesminuscule)proportionof the concrete practice it depends on the mun-
natureof the splitbetweenthe two oppos- populationwill ever be neuro-surgeons, dane mechanism of the examination or
ing sides in MandalII. While the fact of space scientists, architects,or even ge- some variant thereof. In modern societies,
uppercaste dominanceis undeniableand nericnon-technicalPhDs.Howeverlarge degree of.dependence on the examination
implicitlyor explicitlyconcededby both (relativeto presentlevels) the ideal num- increases in direct proportion to the pres-
sides, one claims that this dominanceis ber or proportionis, it will still be small sure of demand for scarce higher educa-
perfectlyjustifiedandtheotherclaimsthat relativeto thepopulation.We do notneed tional opportunities.The more sought after
it is unjust.Seen in reverse,the institution more;we cannotaffordmore;and there and competitive a field, the greater the
of reservationfor OBCs in elite higher do not seem to be any obvious moral- likelihood that entry will be regulated by
educationis seen by the uppercastes as ethical reasons why we should want to examination. The role of merit and ex-
a travestyof justice - indeed,as the per- change this state of affairs.4 aminations will be discussed in detail in
petrationof injustice.The lowercasteson If the patternof economicrewardsand Section III below.
the otherhandsee it as the long arm of social prestige associated with higher A third general feature of higher edu-
justicefinallycatchingupwiththeunfairly education-basedprofessions(relative to cation in all societies is its role as the
privilegeduppercastes. the rest of society) does not changedras- institutional context responsible for creat-
We are now at the point where the tically,it seemsreasonableto supposethat ing and nurturingan intellectual vanguard
anti-reservationistswanted to begin the the numberof aspirantswho wish to enter entrustedwith the taskof thinkingon behalf
story of MandalII. In orderto proceed such fields will not decline and is likely of society and preparing the present to
furtherwith the story we need to under- to rise.Incombinationwithwhathasbeen meet the future. This abstract and rather
standthe specificitiesof highereducation saidabove,thisleadslogicallyto the con- grand description provides a telegraphic
as a sector and their consequencesfor clusion thathighereducationwill neces- summary of what is expected of "criti-
affirmativeaction. sarilyremaina selectiveor elite sector.In cism", "research" and related activities.
otherwords,therewill alwaysbe a funnel Not all of higher education is devoted to
II effect here- morewill wantto get in than such pursuits, nor need all such pursuits
canbe accommodated, so atleastsomeand necessarily be located within higher edu-
Specificities of Higher Education
probablyverymanywill haveto be turned cation. But there is undeniably an institu-
It maybe usefulto dividethe specifici- away. The importantpoint is that this is tional affinity between higher education
ties of highereducationas a sector into trueandwill remainmoreorless trueeven and the research function, regardless of
thosethatapplyin any context,andthose if Indiamiraculouslyturnsinto a richand whether and how well higher education
thatare peculiarto a poor and populous overdevelopednationtomorrow. manages to support it. The main implica-
countrylike ours.These are discussedin It would seem thereforethat because tion that is commonly drawn from this is
turnbelow. highereducationis inherentlyanexclusive that higher education, or at least some
field, modesof exclusionarebuiltintoits segment of it, can legitimately claim ex-
General Features of fundamentalstructureas a matterof prin- emption from the "normal"rules and re-
Higher Education ciple. Discriminationin the sense of prin- sponsibilities imposed on other sectors of
cipledexclusionis thusa definingfeature public life. I will return to this point later
First of all, unlike primaryeducation, of highereducation.This in turnmeans in the argument.
healthcareorsimilar"basicneeds",higher thatconcretemodalitiesfor selectionand
educationis not a matterof right, leave rejectionare a criticalcomponentof its
alone a fundamentalright.No personof institutionaldesign.Howthesemodalities Higher Education in a Big
and Poor Country5
any caste or communityhas a right to work, how they are percievedby differ-
become a doctor,engineeror otherkind entlyplacedparties,andhowtheyrespond First, in countries like India, higher
of highly educatedperson.Everyonehas to thepressureof conflictandcontestation education is almost entirely state funded
therightto aspireto suchstatusandto fair are thus important questions of both and is still among the most important
andequalconsiderationin the admissions principle and practice. avenues of mobility for all classes includ-
process,accordingto specified normsof Giventhecentralityandcontentiousness ing the affluent class. This is particularly
fairnessand equality.But no one has an of the modalitiesof discriminating (rather trueof elite professional education in India,
a prioriright to actualadmission.3 than indiscriminate)selection to higher despite the growing importance of private
Second,by its very naturehigheredu- education,it is not surprisingthat their actors both domestic and foreign. In India,
cation is a selective field - its elitism is design follows the same generalpattern. the affluent (largely uppercaste) elite have
an integralaspect of its nature,not nec- In most societies and contexts, institu- seceded from school education long ago
essarilyoronlytheperversionof thisnature. tional mechanismsregulatingentry into and are beginning to send their children
Fromthe pointof view of bothefficiency highereducationare basedin practiceon abroad for general undergraduateeduca-
and ethics, higher education is not an some formof scholasticexamination,and tion. The middle classes (with a more mixed
"universalisable" resource.Althoughthe in principleon some notionof merit.The caste composition, but still tilted towards
relative numbers and proportions that idea of meritis particularlyimportantas the upper end of the status hierarchy) are
determineidealor desirablelevels will be it bears the heavy ideologicalburdenof now abandoning state schools, but they
differentin each case, and althoughwe legitimisinga system explicitlybasedon still need state-runcolleges anduniversities.

2440 Economic and Political Weekly June 17, 2006

This content downloaded from 131.172.36.29 on Sat, 09 Jan 2016 20:51:02 UTC
All use subject to JSTOR Terms and Conditions
But institutions like the IITs, IIMs and decades of independence got to occupy an eagerto claimandto demonstrate thatthey
AIIMS are still in demand by everyone empty and expanding state sector. This practise merit discrimination,they are as
from the affluent elite downwards because kind of historic opportunity is a once only anxiousto denythattheyarecomplicitous
they offer credentials encashable in the phenomenon, and no other generation will in the practiceof resourcediscrimination.
"first world" at "thirdworld" prices. That get it. While this may be treated as "his- Similarly,victimsof meritdiscrimination
is why, unlike primary or secondary or torical luck", it should be kept in mind may claim to be victims of resourcedis-
even general tertiary education, institu- when trying to interpret the strife and crimination,or beneficiariesof resource
tions of specialised and professional edu- acrimony that marks the present. discrimination mayclaimto be meritorious.
cation are being subjected to enormous A third important feature of higher Finally, a fourth featureof the higher
political and social pressure.6 education in poor and populous countries educationsector in poor countrieswith
This is happening because the long- is thatit tends to be associated with various large populationsis that it must take on
standing monopoly of the upper caste and kinds of discrimination in practice. These the additionalburdenof supportingaspi-
upper class elite over these resources is may be collectively called "resource dis- rations for mobility. This is somewhat
now being challenged by politically resur- crimination" i e, discrimination born out different from the point made above.
gent lower castes and classes. Previously, of inadequateendowments of the resources Because in our context the numberof
this monopoly worked through something required to access and succeed in higher desirable formal sector jobs is always
akin to the "silent compulsion of economic education. It is well known, for example, much less thanthenumberof job seekers,
relations"thatMarx spoke of. The modali- that higher education is biased against the applicants mustconstantly seektoimproved
ties of merit went with the grain of society poor, and against the lower castes or other their credentials.One way of doing this
so to speak: they "naturally"favoured the groups who suffer from social disadvan- is to acquiremoredegreesand this leads
privileged and in effect handed over elite tages in society. This is true in developed to "credentialinflation". A significant
education to them by default. This status countries as well, but is more starkly proportionof thosein highereducationare
quo, consisting of a de facto monopoly relevant in countries like India. By its very theresimplyto improvetheircareerpros-
masked by the de jure presence of open nature, higher education presupposes ac- pects in non-researchrelatedfields.Thus,
competition, is now being questioned "by cess to a minimum level of economic, in a poor country,highereducationmust
any means possible". culturalandpolitical resources. Only those also accommodatethese legitimateaspi-
The second feature is very closely re- who already possess such resources can rationswhichareforcedto takethe route
lated to the first but is important enough realistically expect to benefit from it. That of highereducationbecauseof prevailing
to be considered a distinct point. This is is why the "creamy layer"argumentneeds marketand social conditions.7
the fact that in a poor country with limited to be made with care. Providing access to One consequenceof this featureis that
avenues for capital accumulation, state- higher education is not a method for tack- it counterbalances thespecialclaimsmade
fundedhighereducationprovidesthe safest, ling poverty; by the same token, poverty on behalfof highereducationbecauseof
most legitimate and least regulated method cannot be made a qualifying condition for its responsibilityto pioneercuttingedge
of privatising public resources. The end granting special access to it. Indiscrimi- researchfor the future.To the extentthat
product produced by the IITs, AIIMS and nate use of the creamy layer argumentthus highereducationfunctionsas an avenue
similar institutions is a "credentialled" risks disqualifying precisely those seg- of mobility,it mustbe subjectto thesocial
individual who is a free agent but by virtue ments of socially disadvantagedcastes and justice or other obligationsimposed on
of the state's investment in him/her now communities who have a good chance of publicinstitutions.Exemptionfromthese
owns a kind of capital that is perfectly succeeding. Conversely, with minor exag- obligationsmeans that this routeto mo-
portable and (in this instance) internation- geration,one could claim thatheavy handed bility is embeddedin a system that is
ally encashable. Being a free agent, the use of creamy layer arguments would end inaccessiblefor many.
credentialledindividualcan putthis capital up admitting students whose cumulative
to (almost) any kind of use (almost) any- disadvantages make it highly probable Ill
where - the particularstate and society that that they will fail, thus discrediting the
Meanings and Roles of Merit
made the investment may or may not gain. affirmative action programme itself.
Compare this to land or industrial capital, This combinationof merit-discrimination, We are now in a positionto returnto
and the differences become clear. or discrimination in principle (discussed the questionof meritand its criticalrole
Although it is obvious that the impor- above), and discrimination in practice, or as the principalsourceof legitimationin
tance of credential capital has increased resource-discrimination, produces a situ- a field inevitablymarkedby discrimina-
greatly in the era of globalisation, we must ation where one kind can in fact masquer- tion andexclusionof variouskinds.What
not forget the special historical role of this ade as another, or where claims to this does meritmeanin the contextof Indian
process. Massive expansion of state funded effect can be made regardless of the facts highereducation?
higher education in the Nehruvian era saw of the matter.The permanentpotential for
the uppercaste middle classes convert their misrecognition of one kind of discrimina- Examinations and Merit:
landed capital into credential capital. The tion for the other kind makes the issue very Denotations
state at that time fuelled both the demand contentious and also very hard to resolve
and supply sides of the higher education because merit-discriminationis considered We are now in a positionto returnto
equation - it provided educational oppor- legitimate and desirable while resource- the questionof meritand its criticalrole
tunities and training, as well as employ- discrimination is considered illegitimate as the principalsourceof legitimationin
ment. The castes and classes who were in and undesirable. While higher educational a field inevitablymarkedby discrimina-
the right place at the right time in the early institutions and administrators are very tion andexclusionof variouskinds.What

Economicand PoliticalWeekly June 17, 2006 2441

This content downloaded from 131.172.36.29 on Sat, 09 Jan 2016 20:51:02 UTC
All use subject to JSTOR Terms and Conditions
does merit mean in the context of Indian The dreaded "cut-off point" is the defensible method of saying "No" to large
higher education? guillotine thatsevers the candidatepool numbers.
At the denotative (concrete, literal) level into the mutuallyexclusive categoriesof
merit usually refers to a certification of "meritorious"and "withoutmerit".But Examinations and Merit:
competence, aptitude or knowledge ac- how is this cut-off point determined? Connotations
quired through an examination of some By the number of places available. In
kind. In most instances, what is actually short, the numberof meritoriouscandi- At the connotative (symbolic, figura-
involved is the relative rank obtained in datesis pre-determined; the exam is only tive, ideological) level, merit functions as
the examination. What really matters is not a means to identify who they will be. a kind of entitlement, a moral claim on
really how "well" one does in the exam, How is this identification to be made? society. It is simultaneously a claim in the
but how much "better"(or worse) one does By ranking the candidates. The first sense of an assertion about myself (my
compared to others taking the same exam. social functionof the exam is to produce capabilities, competence, and at the broad-
The matterdoes not end here. In the context or elicit evidence of inequalityfrom the est level, moral worth); and a claim in the
of entrance exams for professional educa- candidates.Theexamis thusanimplacable sense of an expectation or demand ad-
tional institutions, for example, the critical device for generatinginequalityalong a dressed to the rest of the world. Merit at
factor is getting a high enough rank to continuousscale, the measurementunits this broad symbolic level functions as the
qualify for admission. Suppose an insti- of which can be infinitesimal - three raison d'etre of the examination. Or, to put
tution called XIT has 3,000 places avail- decimalplacesarenowcommonlyreported. it the other way around, the third social
able to be filled through an entrance exam. Buta curiousreversaltakesplaceoncethis function of the examination is to identify
Then "merit"for XIT - and therefore for inequalityis successfully generatedand merit. But where the first two functions
the candidates aspiring to enter it - means the rankingdone - then, the obsessively were latent (i e, unrecognised by or opaque
all ranks from 1 to 3,000. From the point continuous scale suddenly transforms to the actors involved), this third function
of view of XIT, ranks lower than 3,000 into a dichotomy with the guillotine of is a manifest function (i e, explicitly
are all equal or the same in the sense that the cut-off point creatingtwo internally recognised or stated).
they all belong to the category of "Did not homogeneous but mutually exclusive We can now resume the discussion about
qualify", which is indistinguishable from groups.The secondsocial functionof the how higher education inherently involves
the category "without merit".8 examinationis to providean ideologically exclusion, and how the merit examination

Call for Papers

Conference on Multidimensions of Urban Poverty in India, jointly organised by Indira Gandhi Institute of Development Research
(IGIDR),Mumbai,and the Centre de Sciences Humaines (CSH), New Delhi, on October 6th and 7th, 2006 at IndiraGandhi
Institute of Development Research, Mumbai.

Deadline for submission of paper/extended abstract of 1500 words: July 31st, 2006
Email address for submission: [email protected]

Researchers are encouraged to submit original papers based on rigorous case studies and/or empirical or theoretical research
work with an India focus. The organising committee wishes to reach a balance between research on large cities and metropolises
and that on small and medium towns. Papers assessing the impact of public policies and specific programmes are also welcome.

The identified sub themes are:

(i) Characteristics
and determiningfactorsof urbanpoverty(at the nationalscale or at a cityscale, linkswithsize of cities,
role of migration, structure of labour markets etc)
(ii) Access to social (health, education) and physical (water, sanitation) infrastructure
(iii) Inequalities in access to credit markets
(iv) Formal and informal housing markets
(v) Urban Livelihoods

Further informationis available on the website of the institutes:

http://www.igidr.ac.in and http://www.csh-delhi.com
All submissions will be refereed and authors will be informed on the status of their papers no later than August 18th. Authors
of accepted papers are requested to send the final version by September 15th. Since we are able to provide only limited
travel support, authors are encouraged to seek their own travel funding. Outstation participants will be provided with
accommodation at IGIDR. For further information,please send an email to [email protected]

Postal Address: Attn: S. Chandrasekhar & Marie-Helene Z6rah

Organising Committee, Multidimensionsof Urban Poverty in India,
Indira Gandhi Institute of Development Research,
Gen A. K. Vaidya Marg, Goregaon (E), Mumbai- 400 065, India

2442 Economic and Political Weekly June 17, 2006

This content downloaded from 131.172.36.29 on Sat, 09 Jan 2016 20:51:02 UTC
All use subject to JSTOR Terms and Conditions
combination bears the burden of - do entrance tests really measure what resultsin competitiveexamsthatqualifyas
legitimising such necessary exclusion. they are supposed to measure, i e, aptitude, indicatorsof merit:(a) economicresources
From the point of view of allocation of likelihood of succeeding, etc? Here we are (for prior education,training,materials,
scarce higher education resources, merit- on familiar ground, for it is well known freedomfrom work, etc); (b) social and
discrimination is justified as the method how difficult it is to devise tests with good culturalresources(networksof contacts,
of identifying persons who (a) are best predictive power in this sense. We may confidence,guidanceandadvice,informa-
preparedand qualified to enter specialised know a good doctor or engineer when we tion,etc);(c) intrinsicabilityandhardwork.
higher education; and (b) will produce the see her at work, but we do not really know It is somecombinationof thesethatallows
best returnon this social investment. These how to predict this before the fact. people to "acquiremerit".
are actually two distinct claims, but are To argue that a system is arbitraryis not When it is said that merit alone is
frequently conflated - it is possible, but to say that it is useless, or more important, responsibleforthedominanceof theupper
not necessary, that the best prepared per- that obvious alternatives exist. The pur- castes,whatis meantis thateconomicand
son in the sense of scholastic aptitude will pose of this argument was to show that the culturalresourcesare not important,but
also be the person who will make the best moral weight that is placed on merit is in it is differencesin sheer intrinsicability
doctor, engineer or whatever. practice borne by examinations, and that alone thatmakefor the inter-castediffer-
In principle,examinations identify merit, examinations cannot but be arbitraryunder ences. This is a position that is indefen-
and merit provides sufficient justification the conditions imposed on them. Appre- sible, for it cannot be arguedtoday that
for discriminating in favour of its bearers ciation of this arbitrarinessshould temper largegroupsnumberingin themillionsare
and awarding them admission in prefer- one's opposition to reservations or similar more or less intrinsicallyable thanother
ence to others who do not have merit. In proposals that appearto interfere with this suchgroups.We have to look to inequali-
practice, examinations coercively gener- system. When opponents of reservations ties in the other factors to explain the
ate inequality expressed in a rankordering, (who have themselves survived such a difference.
and they help to persuade both the system) use emotive language like Once we recognisethe causalcontribu-
"selected" and the "rejected" that the di- "murderof merit",they are trying to lever- tion of other inequalities towards the
vision is fair. However, as long as they age the moral potency of merit to foreclose unequaldistributionof meritandhenceof
succeed in practice, examinations are tempered responses. They are suppressing highereducational opportunities,thisopens
presumed to have succeeded in principle. or disowning theirown intimateknowledge the door to consideringinterventionsfor
In other words, the only thing an exam of the heartbreakingarbitrarinessof merit theirredressal.(Becausethesereasonsfor
must produce is a rank ordering that is not discrimination. They are endorsing the inequalitycannotbe called"justdiscrimi-
disputed by candidates; this is a necessary guillotine mentality and refusing to ac- nation";it is to foreclosethisthatthe"merit
andalso asufficientcondition forthe system knowledge that the ranking game starts as the only criterion"argumentis made.)
to succeed. Everything else is an optional with the play of infinitesimal gradations. Once we begin to talk in termsof grada-
matter of assertion and counter assertion. How much "compromise" with merit a tions - as we must - it becomes possible
(Imagine what would happen if, with only reservation scheme will entail ought to be to movetowardsa morerealisticandtrans-
50 seats available, the AIIMS entrance recognised for what it is - an empirical parentpolicy frameworkwhere we can
exam produced a result where the top 100 question. How much further down the discussthe differentsocial objectivesthat
candidates had the same mark...) rankings will we have to go? What is the highereducationcan accommodatewith-
To put the argument sharply, the merit- substantive meaning of this distance in out excessive costs or damages being
discriminationsystem functionson the basis marks or ranks, i e, how much of a dif- imposedon it. The presentframeworkof
of formal differentiation of a candidate ference does it make in terms of the quality debate- markedby Manicheandichoto-
pool through an examination; it is insti- of candidates? It is only after we have miesbetweenmeritandincompetence(as
tutionally requiredto ensure such a differ- asked and answered such questions thatwe thoughtherewere nothingin between,or
entiation. It is not required to ensure or will be in a position to respond in a rea- thateach was such a singularmonolithic
defend a substantive differentiation of soned manner to proposals like the recent category)- will not take us forward.
candidates. What must be produced is a one for OBC reservations. By wayof conclusion,it wouldseemthat
differentiated ranking; it is not necessary while more conceptualwork definitely
to explain what meaning the differentia- Merit and Resource-Discrimination needsto be doneon thesequestions,what
tions carry. More accurately, as long as we need even more perhapsis more and
they are present, it is permissible to simply The preceding argumenthas kept within thickerempiricaldescriptions.Mostof all,
assume that the differentiations mean the limits of the "merit-only" position; it we needgooddescriptionsof theeveryday
whatever they are supposed to mean. This is time now to move beyond this self- practicesthathelp produceandreproduce
is the underlying system that, under the imposed limitation. social capitaland link caste to privilege
pressure of large numbers of aspirants, If we return now to the argument in or disprivilegein durableways.Bi
produces the arcane world of thirddecimal Section I and the figures in the table, the
point differences and cut-offs that are skewed distribution of access to higher Email:[email protected]
accepted as justifying large claims about educationwas explained as being due solely
the presence or absence of merit. to merit by the anti-reservationists. How- Notes
In effect, one could say thatthe preceding ever, the flaws in this argument are too
1 The titleandthisessay itself area tributeto Marc
argumentestablishes thatthe examination- obvious to need much rebuttal. As Marc
Galanter'sclassic work, CompetingEqualities:
rankmethod is of dubious reliability. This Galanterhas pointed out, three broadkinds LawandtheBackwardClasses in India(Oxford,
still leaves open the question of validity of resources are necessary to produce the New Delhi, 1984). It is to his credit - and to

Economic and Political Weekly June 17, 2006 2443

This content downloaded from 131.172.36.29 on Sat, 09 Jan 2016 20:51:02 UTC
All use subject to JSTOR Terms and Conditions
 the collective embarrassmentof my "caste"of even-handedapplicationof this logic to the two alleviating the condition of OBCs via
social scientists and specially my "sub-caste" sides in Mandal II.
increasing their share in public sector jobs
of sociologists andsocial anthropologists- that 4 Unless we reach a stage of social development
where the wishes of individualsor groupsto be
and institutions of higher education, then
morethantwo decades afterits publication,his
involved in higher education for its own sake this lack of preparationnot only worsens
superbsurveystillremainsby farthebestanalysis
of "compensatory discrimination" in are considered sufficient to justify investment its case but actually casts a doubt over the
independent India. Its own example and the to expand this field. degree of its commitment to the reduction
events of the "Mandal decade" should have 5 Thefeaturesnotedherearenotnecessarilyabsent of caste-based barriers.
been advantageenough for us to at least match in a richsmallcountry,buttheireffects in a poor
the analyticalprecisionand careful scholarship big country are of much greaterconsequence. Having said that, it would be a mistake
of CompetingEqualities.Whateverthe reasons 6 This argument is made by Mary John; see to believe that the opposition to this move
why this has not yet happened, I hope that 'Schooled in Inequality'in TheHindu,May 30, is only or mainly because of lack of data
"MandalII" will providecollective inspiration 2006. and/orpreparedness.The dominationof an
to overcome them. This is a very preliminary 7 This also happens in rich countries, but to a
versionof work in progressand should be read muchlesserextent.No one needsa postgraduate
upper-caste mindset in the media, the
as such. For more immediate incitement, degree merely to reachthe averagestandardof academy and the wider public, is evident
encouragementand critical engagement I am living.
in most of the anti-reservation com-
grateful to Mihir Shah, Yogendra Yadav and 8 This is because, after the rankingis done and mentaries. Concerns about merit declining
Mary E John. results declared,the XIT cannot meaningfully because "they", rather than "us", will
2 ThesedifferencesaresharpestamongChristians differentiate between different "amounts"of
andSikhs,andleast amongMuslims.The Other merit that are less than whatever was needed
swamp educational institutions; the com-
Religions in this table include Buddhist dalits for being the 3,000th candidate.In otherwords, pletely mythical nightmare of "we" or
(since SCs are limited to Hindus),and they too it has nothing (or the same thing) to say to all "our children" being reduced to demean-
would be sharplydifferentiatedfrom the Jains, rankshigherthan3,000 - "Goodbye,betterluck ing jobs like shining shoes and sweeping,
other Buddhists, and Parsis. if you try next time". There is thus no whereas it is only fair that "they"continue
3 The strict implications of this are often demonstrabledifference between saying "Not to do so, etc, are only some of the indi-
glossed over, as was evident in the less than enough merit"and "No merit". cators of a latent casteism that constitutes
the reality of the supposedly caste-neutral
contemporary urban India. Thus, the
The Eternal Debate status quo, in which the share of the
upper castes in both good jobs (in the
public and the private sectors) and seats
in higher educational institutions is far
Caste still remainsan indicator of disadvantageas distributionof in excess of its share of population, is seen
both income and wealth are skewed along caste lines. Thoughthe as essentially fair. While state-imposed
data on OBCs is scanty, there exists a clear disparitybetweenthese reservations are seen as a murder of
castes and others in terms of educationalattainment,occupational merit, the much more widespread "natu-
success and standardof living. The mechanismsfor perpetuating ral reservations" that come from birth in
business (upper caste) families which
inter-caste inequalityare still strong and alive in contemporary ensure that no outsiders are ever consid-
India. Quotas, however, should not be seen as the beginningand ered for the top jobs in family-owned
the end of affirmativeaction. businesses; other manifestations of caste-
class privilege being expressed via dona-
ASHWINI DESHPANDE are no firm estimates of the distribution tions, capitation fees, etc, are never seen
of OBCs in jobs and institutes of higher as threatening merit. Thus, the operation
rT he debate over reservations - the education. of caste barriers and a beief in caste
Indian version of affirmative It is entirely true that the government, hierarchies as natural underlies the
action - is in the foreground once whether in 1990 or now, ought to have dominant thinking so completely that,
again. Since this time the debate, at least preceded this major move with building ironically, these beliefs are not seen as
ostensibly, is about the other backward a comprehensive case for this particular casteist, but their questioning is. Indeed,
classes (OBCs), one of the first stumbling component of affirmative action, with this mindset asserts thatcaste was virtually
blocks in any informed intervention is the adequate backing of data that could be extinct before it was resurrected by
paucity of data. Most of what has appeared released to the public. Also, while this V P Singh or Mandal or Arjun Singh. The
in newspapers and television commentar- move is supposed to be coupled with an possibility that the invisibility of caste
ies are a set of pre-conceived notions increase in the non-reserved seats, very from "our" lives is a privilege that "we"
intertwined with a virulent opposition to little thought seems to have been given to enjoy because "we" belong to a small
affirmative action. Prima facie, this could the logistics and practicality of increasing minority of urban, upper-caste elite that
be attributedto the lack of data: reserva- seats. Most departments and colleges al- has the luxury of leading a casteless
tions for OBCs could possibly be the first ready face huge constraints in the form of existence, something that the majority is
instance of affirmative action for a cat- inadequate infrastructure and high far from attaining, are issues that are vir-
egory that even the national census does student-teacher ratios and have very little tually never discussed. However,
not enumerate. The National Commission scope for increasing student intake, with- distinct from this mainstream is the rela-
for Backward Classes only lists the jatis out the concomitant structuralchanges that tively small body of opinion thatrecognises
that comprise OBCs without any demo- would be needed to sustain a higher intake. the reality of caste-based barriers and
graphic data about these jatis. Also, there If the government is indeed serious about disadvantages and has been engaged with

2444 Economicand PoliticalWeekly June 17, 2006

This content downloaded from 131.172.36.29 on Sat, 09 Jan 2016 20:51:02 UTC
All use subject to JSTOR Terms and Conditions