Competing Equalities: Law and the Backward Classes in India 9780195615531, 0195615530

Citation preview

Competing Equalities

Competing Equalities

LAW AND THE BACKWARD CLASSES IN INDIA

Marc Galanter “Sf

DELHI OXFORD UNIVERSITY PRESS BOMBAY CALCUTTA MADRAS 1984

Oxford University Press, Walton Street, Oxford OX26DP LONDON GLASGOW NEW YORK TORONTO DELHI BOMBAY CALCUTTA MADRAS KARACHI KUALA LUMPUR SINGAPORE HONG KONG TOKYO NAIROBI DAR ES SALAAM CAPE TOWN MELBOURNE AUCKLAND and associatesin BEIRUT BERLIN IBADAN MEXICO CITY NICOSIA

© Marc Galanter 1984

Edition for South Asia

Filmset by N.K. Enterprises, Daryaganj, New Delhi 110002

Printed by Rekha Printers Pvt. Ltd., Okhla Industrial Area, New Delhi 110020

Published by R. Dayal, Oxford University Press 2/11 Ansari Road, Daryaganj, New Delhi 110002

TO MY FATHER JACOB GALANTER 1896-1972 Justice, justice you shall pursue

yn (Deut. 16:20)

pw

pw

Contents

List of Tables Foreword by M. Hidayatullah Preface Acknowldegements

Note on Citations and References Note on State Names List of Abbreviations Introduction

xiii xvii

xxiii xxiv

XXV

PART ONE: COMPENSATORY DISCRIMINATION: THE SETTING AND THE POLICIES Chapter

1.

The Compartmental Society

Chapter

2.

Reform, Mobility, and Politics under British Rule

Chapter

4.

Reservations in Government Employment: A Closer Look at the Paradigm Program A. The scope of reservations B. The effectiveness of reservations C. Assessing the costs and benefits

Chapter

3.

Compensatory Discrimination Programs and Their Implementation A. Legislative reservations B. Education and other schemes C. Administration D. Critics, costs, and benefits

PART TWO: IDENTIFYING THE BENEFICIARIES Chapter

5.

The Designation of the Scheduled Castes and Scheduled Tribes

A. The invention of the Scheduled Castes B. Scheduled Castes under the Constitution

C. The Scheduled Tribes

105 119 121 122 131 147

viii

Contents

Chapter

Chapter

6.

7.

The Other Backward Classes: The Process of Designation A. The pre-history of the backward classes category B. The constitutional design

154 154 159 167 179

of Beneficiary Classes A. Castes as classes

188 188

C. The failure to centralize policy D. The varied state patterns The Backward Classes and the Judiciary: The Selection B. The permissibility of non-communal classes

C. The utility of communal units

Chapter

8.

D. Forbidden classifications and the burden of proof The Backward Classes and the Judiciary: The Measure of Backwardness: Educational backwardness

mEOMBOOD>

Social backwardness

Cumulative backwardness Under-representation

Absolute vs. relative backwardness

Chapter

The sufficiency of data Economic tests . Geographical criteria ’ Non-inclusion claims

9. Membership in Groups Entitled to Preferential Treatment A. Reading the schedules B. Caste and sect

C. D. . E. . F. .

Tribe Tribe and caste Religion Group integrity: admission, expulsion, re-admittance

222 223 229 240 243 246 249 261 277 279 282 283 290 293 295 305 326

©

Non-communal criteria: income and occupation tests H. Conclusion: pragmatic and formal solutions Social Identity and Judicial Action

215

Chapter 10.

PART THREE: COMPENSATORY DISCRIMINATION AND THE JUDICIAL PROCESS Chapter 11.

The Constitutional Scope of the Compensatory Discrimination Policy

A.

Preferential treatment in the context of

Fundamental Rights

341 348 352

361

363 364

Contents B. The scope of the constitutional provisions for special treatment

368

C. The exceptional character of compensatory discrimination Chapter 12.

377

D. A new constitutional vista? Thomas and after

382

Limits and Judicial Control A. How much reservation? The minimum

396 396

The Scope and Extent of Preferences: Constitutional

B. How much reservation? The maximum

399

C. The intensity of preference

420

F. Benefits other than reservations

41

D. The wider use of the reservation device E. The number of beneficiaries

Chapter 13.

425 437

The Design and Operation of Compensatory Discrimination Programs: Constitutional Limits and Judicial Control

444

A. Rulefulness and discretion

Chapter 14.

B. C. D. The

444

Reservations as guaranteed minimums Compartments and layers Relations between multiple reservations Setting and Incidence of Judicial Intervention

455 463 472 477

A. Courts and judges

477

C. The pattern of litigation D. Lawyers

498 513

B. Authority and precedent: formal structure and operating norms

Chapter 15.

ix

The Uses of Judicial Action

A. The impact on policy and administration B. Judicial review: the doctrinal accomplishment C. The limits of judicial action

490

.

522

522 533 541

CONCLUSION Chapter 16.

“The Little Done, the Vast Undone” A. A Costly Success

547 347

D. Ex Orient Lux?

562

B. Fairness and History _ C. Secularism and Continuity

Appendix: Relevant Sections of the Indian Constitution Bibliography General Index

Index of Cases

552 559

569 575

613

621

Tables

Annual Per Capita Expenditure on Scheduled Castes and Scheduled Tribes for the five Five-Year Plans School Enrolment as a Percentage of Corresponding Age Group,

57

by Communities, 1973-1974

Views of Reservation in Government Employment for Scheduled Castes: Three North Indian Localities in 1968 Alleged Benefits and Costs of Policy of Compensatory Discrimination Scheduled Caste and Scheduled Tribe Employment in Central Government Service, 1953 and 1975 Scheduled Castes, Scheduled Tribes, and Other Backward

70 81 89

Classes in Government Service: Posts Reserved and Posts

Held, about 1976 Approximate Effectiveness of Reservations for Scheduled

Castes and Scheduled Tribes in Central Government

Services over a 21¥2-Year Period (1953-1975)

Comparison of Scheduled Caste Employment in Government Undertakings, Mid-1960s Comparison of Scheduled Caste Employment in Government 10. 1. 12. 13.

Undertakings, 1975

and Scheduled Tribe Service and Public Sector

and Scheduled Tribe Service and Public Sector

Alleged Benefits and Costs of Preferences in Government Employment Prior Educational Performance of Successful Candidates for All-India Higher Non-technical Services, 1960-1965 Percentage of Untouchables in British India

Estimates of Untouchable Population in British India by Various Official Bodies, 1917-1941

14, 15.

Population of Scheduled Castes in 1941 Scheduled Castes Population (in millions) and Percentages,

16. 17. 18. 19.

Scheduled Tribes Population Scheduled Areas Population

1950, 1951, 1956, 1961,1971

Concessions for Other Backward Classes, 1951-1952

Estimates of Population of Other Backward Classes, Early 1950s

102 103 106 112 124 125 131 133 150 1 163 169

Tables xi A Profile of Other Backward Classes in the Late 1970s Relative Success of Selected Castes on Merit and on Selection by Income-cum-Occupation Test in Mysore Engineering Colleges, 1963

25.

Discrimination, 1950-1977

Initiating and Successful Parties in Compensatory 27. 28.

271

458

S01

Discrimination Litigation, 1950-1977

Regional Distribution of Litigation about (Other) Backward Classes Subject Matter of Supréme Court Litigation about Compensatory Discrimination

§ §

24,

Contrasting Results of Different Tests of Membership Possible Methods of Declaring Winners in a Double-Member Constituency A Comparison of Upper and Lower Strata of Backward Classes in Tamil Nadu Subject Matter of Reported Litigation on Compensatory

216

Foreword

PROFESSOR MARC GALANTER has written an informative and educative treatise on the Indian experience with what he terms “compensatory discrimination,” a subject which deals with the Indian Constitutional policies to safeguard the interests of the historically disadvantaged sections of the population. At the present moment there is considerable debate in India whether this compensatory discrimination is not an unfair measure, as it keeps out deserving persons and promotes others not so qualified. In fact, there is growing evidence of a

confrontation between persons who have received the benefit of reser-

vation and those who have not. The issues resemble recent cases in which the United States Supreme Court has attempted to balance preferred treatment for blacks against the claims of whites, complaining

ofa sort of discrimination in reverse. This book is divided into three parts. The first part describes these policies and their background. The second part deals with the problem of identifying the recipients of this discriminatory advantage. The various aspects of these problems are treated very fully, and the concluding chapter summarises ‘the implications of the governmental superimposition of official categories of beneficiaries in the complex process of group and personal identity for the role of the courts in implementing the compensatory discrimination policy.” The third part deals with the Constitutional provisions which embody this policy and with the laws which have implemented it and the judicial decisions which apply these provisions. Finally, the Conclusion tries to assess the achievements of these programmes in India in the light of the problems

that are inherent in such a policy.

The problem of disadvantaged sections of the population has existed in India for centuries, but it acquired special importance after the First World War when a system of protections began to be evolved. Later that system came to be supported by the concept of human rights and

fundamental freedoms in the United Nations Charter and the Universal Declaration of Human Rights.

In India the special position of the minorities, for a long time, has

xiv

Foreword

been an ineluctable fact of political life. From the time of the Indian National Congress session held in Karachi in 1931 and the Nehru and Sapru Reports, it was consistently felt that in order to create a homogenous egalitarian society and to achieve national unity, balance between the various communities in India was essential. Jawaharlal

Nehru posed the problem thus: ‘How shall we promote the unity of India and yet preserve the rich diversity of our inheritance?” The Preamble

of our Constitution

resolves to secure “JUSTICE,

social,

economic and political,” to all its citizens. It further proclaims as one of its aims the promotion among them of “FRATERNITY assuring the dignity of the individual and the unity of the nation.” Minorities in India enjoy the guarantees of equality before the law (Article 14), prohibition of discrimination on grounds of religion, race, caste, sex or place of birth (Article 15); equality of opportunity in

matters of public employment (Article 16), the well-known rights to freedom (Article 19), the right to freedom of conscience and free profession, practice, and freedom in religion or sections thereof (Article 26).

The Constitution also confers certain rights specifically on minority groups: the right to conserve a minority group’s distinct language,

script, or culture, and the right to establish and administer educational institutions of the group’s choice (Articles 29 and 30). In addition to

religious minorities and linguistic minorities the Constitution specifically recognises Scheduled Castes (untouchables), Scheduled Tribes, and socially and educationally backward classes and guarantees to them protection. Article 17 decrees the abolition of untouchability. Article 46 requires the State to promote with special care the educational and economic interests of the backward citizens and in particular of the Scheduled Castes and Scheduled Tribes and imposes a duty on the State to protect them from social injustice and all forms of exploitation. The Governors have been given special responsibility in this connection. The Constitution also provides that the Commissioner for Scheduled Castes and Scheduled Tribes and the Commissioner for Minorities have to submit periodic reports, which are placed before Parliament and which are open to discussion. Reports have been submitted by many commissions, making suggestions for protecting the interests of the minorites. The rights of minorities have been implemented by statutory enact-

ments. These rights have been supported and their scope and ambit clarified by decisions of the Supreme Court and the various High Courts from time to time. In spite of the consistent commitment to minority rights, there is insufficient awareness of the laws on the subject, not only among the people directly affected, but also among some of those who are concerned with the administration of these laws. The

Foreword

xv

absence of discrimination and the protection of minority groups are the tests of the political and moral achievement of the society. The creation

of a just social order as envisaged in our Constitution (Article 38) requires a social consciousness and active cooperation of all the sections

of the society. Professor Marc Galanter has written exhaustively of the history of the

disadvantages and the revolt against these disadvantages. In his account, he has mentioned all the cases decided in the different High

Courts and the Supreme Court which provide valuable help in under-

standing the objectives embodied in our Constitution in the light of our commitments to democracy and socialism.

This book will be of immense use not only to the members of the legal profession but also to those who have at heart the achievement of national integration. It will contribute to the shaping of public opinion as well as to the enhancement of discussion of these matters by professionals and policy-makers. It comes none too soon, as a great debate is

going on, on the various aspects of the reservation policy.

Vice President's Residence New Delhi

16 March 198!

M. Hidayatullah

B.A. (Nag.), M.A. (Cantab.), LL.D.

Bencher, Lincoln’s Inn

Preface

IN went forth on a untouchability. rights” lenses,

1957, NOT LONG after graduating from law school, I Fulbright Scholarship to study India’s efforts to abolish Ready to see these efforts through my American “civil I was struck by the fact that measures to enforce civil

rights of a sort familiar to me were overshadowed

by programs: of

preferential treatment. The principle of systematic preference in favor of historically deprived groups was incorporated into a constitutional order with a strong emphasis on evenhanded application of law. The notion that commitment to equality might entail policies of preferential treatment was at that time hardly conceivable in the American setting. A few years later, it was with a sense of familiarity that I first encountered advocacy of compensatory preference to enable American Blacks (then Negroes) to overcome their historic handicaps. Today such special treatment is a familiar feature of American life— although the legitimacy and efficacy of such programs is still highly controverted. India’s long and elaborate experience with kindred policies may contain lessons for American policy. It was not the hope of extracting lessons for America that drew me to this topic, but curiosity about the course of India’s pursuit of equality by group preference. Surely this concern reflected events in America, but I experienced it first as part ofa challenge to understand India, then (as my perspective shifted) as an intriguing instance of using law to

reshape recalcitrant social patterns. Many interruptions and digressions later, I find that this topic has over the years absorbed far more of my time and energy than I ever planned. This book has been long in the writing, and the discerning reader may find outcroppings of the understandings and perspectives that informed various stages of the work. I

console myself that such disparities are the price of deepening awareness of the perplexities of the subject. The shape the book has taken over these years is not what I would project if I were to start now with my present views and interests. Were I starting now, I would undoubtedly pay more attention to politics and

administration and somewhat less to the courts. As a lawyer, courts

xviii

Preface

were my entrée into Indian society. The use of courts as an observation point introduces a special bias, but the cases give a texture and immediacy to the account of policy. I have tried to deal with legal materials in sufficient depth to preserve the detail and complexity that make them meaningful to the immediate participants (Indians and mostly lawyers,

judges, and government officials) while making them accessible to those at a distance. Technical legal questions, I submit, should not be dismis-

sed as “merely” technical. It is in a host of detailed questions that the tensions among conflicting commitments must be resolved—and this

includes all those commitments which are crystallized into the styles of the participants. At the same time that we see large principles illuminated by their concrete applications, we can observe the interplay

of institutional

routines

and

inventiveness

that gives

kaleidoscopic sense of simultaneous rigidity and fluidity.

the

law

its

The courts proved not only a convenient vantage, but interesting to observe in themselves, giving this book what became its second subject.

Focussing on the intersection between courts and the compensatory discrimination policy, I use the courts as a way of learning about the

policy, and the intersection as one way of learning about the function of courts in Indian society. The courts turned out to be more limited as a

vantage point than I naively assumed at the outset. They act as a balance wheel channelling the compensatory policies and accommodating them to other commitments; but it is the political process that shapes the larger contours of these policies and gives them their motive force. Official doctrine—judicial pronouncement or administrative

regulation— proved an insufficient guide to the shape of the policies in action and the results they produced. For a time I entertained the hope

of including a systematic portrayal and appraisal of the working of each

of the various programs, but it belatedly dawned on me that such an

enormous undertaking would postpone—perhaps permanently— the completion of this study. So I have contented myself with sketching the lines along which an appraisal loyal to the complexities of these multi-

goaled policies might proceed. I have throughout tried to identify the

points at which legal arguments depend on empirical assertions, to tease out what evidence I could find relevant to such assertions, and to indicate the yawning gaps that would have to be filled to arrive at a definitive answer to many of these questions. Compensatory discrimination policies are not only an instrumental

mechanism

but a medium

through

which

the

Indians

who

devise,

apply, evade, and attack them express their understanding of their own

society. This book, then, may be read as a portrayal of some contemporary Indian discourse about the tensions of hierarchy and equality,

group and individual, policy and social reality.

Preface xix To this discourse I, as an outsider, have a curious and ambivalent

relation. I have been intermittently immersed in these questions, my

sympathies engaged and my intellect provoked. By writing about these

policies from time to time, I have become willy-nilly a participant in this discourse—a queer sort of participant who shares neither an insider’s understanding of Indian life nor the necessity of living with the consequences of my observations. Nevertheless, Indian colleagues have been generously supportive and receptive over the years, inducing the con-

ceit that my relative detachment (and my capacity to stumble over what is obvious to the insider) may produce a contribution that usefully complements those of insiders. I do not pretend to a total and aloof

neutrality. I do not refrain here from drawing conclusions and proposing interpretations of the Indian Constitution. These are advanced in the spirit of friendly argument. There are no conventional categories for organizing the discussion of this sector of Indian law. Of the alternatives that suggested themselves,

I preferred to organize the chapters around logical clusters of related

problems, rather than around separate constitutional provisions, specific programs of preferential treatment, or specific problems of administration. Although many sections are organized around a chronological movement, I have made no attempt to adhere to

chronological sequence overall. Instead, there are successive cuts into the same deposits from different points on the surface. Since the various topics are closely interrelated, I was faced with a choice between repetition or copious cross-reference. I have preferred the latter except

where the repetition was small or the matter merited emphasis.

This is a dynamic field which became a focus of political controversy in the late 1970s. There are bound to be new policy developments and

judicial responses before this is printed; some of the things reported here are bound to be dated or proved wrong ina short time. I hope, though, that the delineation of major problems and the building of conceptual tools for approaching them will outlast some of the more detailed conclusions. In the main I have attempted to trace legal developments through materials available to me through 1978, supplemented by more selective resort to new materials through mid-1980. Although the coverage of the last few years is less comprehensive than for earlier years, I believe the book pictures the major contours of policy as they stood at the replacement of the Janata Government by a resurgent Mrs. Gandhi in early 1980.

Between the completion of the manuscript and the checking of the

proofs in mid-1983, controveisy about compensatory discrimination

took on a new intensity. The violent and protracted protests against

xx

Preface

reservations in medical colleges that racked Gujarat in early 1981 gave

sanguinary expression to the widespread resentment of these policies and their beneficiaries. As conflict has mounted, the fundamental—if

not wholly consistuent—lines of policy have remained unchanged. The

courts have elaborated some lines of doctrine in the earlier caselaw. The

second Backward Classes Commission (Mandal Commission) reported (Backward Classés Commission, 1981), proposing a major extension of preferential treatment to a vastly enlarged group of beneficiaries and

reviving the vamna classifications dismissed by its predecessor a quarter century earlier. The report was tabled in Parliament in April, 1982, and subsequently received parliamentary endorsement, but has not yet been implemented. The scope and intensity of the compensatory policy

is very much a live and open question. The Mandal Report revives for replay on the larger national stage all of the questions and perplexities traced in this book. That stage, too, has undergone change. There has

been a significant enlargement of the judicial repertoire by the addition of a more proactive and innovative style associated with the rubric of “public interest litigation.” There have emerged new advocacy groups

with aspirations and capacities to mobilize this kind of judicial response

on behalf of constituencies of the dispossessed (Galanter, 1984). Until now, none pf this ‘‘public interestlaw” has been directly concerned with

compensatory discrimination. The coming major round of adjudication may involve courts and litigants equipped with enhanced capabilities to address the fundamental questions of policy implicit in the commitment to the compensatory principle.

Acknowledgments

THIS BOOK HAS gone through several cycles of growth. In each it has benefitted from many stimuli and encouragements. The United States Educational Foundation in India administered the Fulbright Scholarship which gave me my first taste of Indian law and life. A year working at Stanford Law School provided the occasion to immerse myself in Indian constitutional law. A decade of participation in the vigorous South Asianist community at the University of Chicago enlarged my appreciation for the complexities of India. A Senior Fellowship (1965-66) and a travel grant (1978) from the American Institute of Indian Studies enabled me to explore preferential policies in the field. The Indian Law Institute provided a congenial base during my 1965-66 fieldwork. From time to time, opportunities for helpful exchange were afforded by a number of institutions, including notably the Faculty of Law and the Department of Sociology of the University of Delhi, and the Indian Council of Social Science Research. My studies of this cluster of problems received valuable encouragement from the Joint Committee on South Asia (of the American Council of Learned Societies and the Social Science Research Council) and from the Insti-

tute for Research on Poverty of the University of Wisconsin. Between the inception of this project and its completion there emerged a vigorous scholarly community devoted to the study of law in its social context. I have enjoyed the privilege of working at two foci of this community—at Buffalo and Madison. My involvement in this “law and society” community both delayed and enriched this book.

At various times the International. Legal Studies program of Stanford

University, the Committee on Southern Asian Studies, the College, and

the Law School of thé University of Chicago, the Research Committee of the Graduate School of the University of Wisconsin and the Ford Foundation have provided support enabling this work to reach fruition. Over the years a number of colleagues were kind enough to read various parts of the manuscript. Their comments have enriched the book and have helped eliminate some of its flaws (not without making

xxii

Acknowledgments

me aware of others not so readily eradicable). For this I am grateful to Phyllis Rolnick Arora, Upendra Baxi, Boris I. Bittker, Rajeev Dhavan,

Clarence Dias, George H. Gadbois, Jr., John Mansfield, S. P. Sathe, R. A. Schermerhorn, and especially to M. G. Narasimha Swamy.

I have been the beneficiary of many forms of kindness and fellowship from colleagues who helped me to locate materials, shared their work in progress, and gave generously of their time and intelligence to educate me. Such a list is necessarily incomplete, but I should like to take this occasion to express my thanks to Partap Aggarwal, Vimal Chandra, Suma Chitnis, P. A. Choudhary, Bhagwan Das, J. D. M. Derrett, V. V. Deshpande, Lelah Dushkin, Joseph Elder, Marcus Franda, Henry C. Hart, L. G. Havanur, B. R. L. Iyengar, S.K. Kaul, P. R. Mehendirotta,

N. R. Madhava Menon, N. G. Nag, N. P. Rege, Vimal Shah, G. S.

Sharma, Parmanand Singh, M. N. Srinivas, Gregory Stanton, Richard Taub, and Eleanor Zelliott. The late Professor Max Rheinstein pro-

vided guidance and encouragement as well as his towering example in the comparative study of law. I was particularly fortunate in having access to the South Asian collections of the University of Chicago and the University of Wisconsin, and the assistance and support of Maureen L. P. Patterson and Jack Wells, who preside over those collections. At various times in the life of this project invaluable help in preparing

the manuscript was provided by Betti Arnett Johnson, Doreen Blank, Cynthia Burns, Judith Frank, Eve Galanter, and Mary L. Galanter.

Marge Eiseman did the final round. with great skill and verve. Seth Galanter conscientiously numbered the manuscript sheets. Janet Guthrie, Catherine S. Meschievitz, and Hugh A. Ross ably provided

crucial assistance. Robert Hayden was a source of encouragement and helpful criticism as he worked with me in reviving and revising the manuscript. In the arduous and seemingly endless task of reading the proofs and preparing the Index, I was fortunate to have the capable help of Susan Bissegger, Michael Macy, and Sylvan Sobel. The title of the concluding chapter is borrowed from Upendra Baxi’s 1967 article. I want to thank Jean Lettofsky for the Hebrew calligraphy. My wife, Eve, and my children, Seth, Rachel and Sarah, graciously

abided the protracted intrusion of this incubus. The views set forth here—even when they contradict one another— are my own and should not be attributed to my benefactors.

Note on Citations and References

Law REPORT CITATIONS of the judgments of courts are given in standard Indian legal form. Thus citations of official reports give the names of the first party on each side, the notation I.L.R. (indicating Indian Law Reports), the volume number, the title of the

series (indicating the court), the page on which the case begins, and, finally the date of publication; for example Gopal v. Hanmant, 1.L.R. 3 Bom. 273 (1879). Citations to the All-India Reporter, the most popular series of unofficial reports, follow a slightly different pattern; for example, Sunder Devi v. Jheboo Lal, A.1.R. 1957 All. 215 (names, A.I.R. series,

year of publication, name of court, page on which case begins). There is a separate A.I.R. volume published each year for each court. The coverage of A.I.R. and I.L.R. is overlapping but not identical, Occasionally, cases not found in either of these series are reported in a local or specialized series. The citation of these follows a similar pattern. Articles, books, monographs, etc., cited by author and year in the notes, are listed in the bibliography. A similar style is used for government

decuments (except Census of India publications, which are cited in standard style in the footnotes). . Pre-Independence government documents are cited by the name of the

committee, commission, or other government body from which they issued (identified in some cases by the name of the chairman). PostIndependence government documents, reports of commissions, study teams, committees, and other non-serial documents are cited by the name of the body (identified in some cases by the name of the chairman) and the year of publication. In some cases where such a body operated under the auspices of a Ministry, the name of the body has been added to the

author listing. Other government documents are cited by the Ministry or other government body under whose auspices they are issued and the year of publication. Where a series is listed in the references under a specific agency, a citation of that body and a specific year means the report for that year. Government publications with named authors are cited by author. Some reports and series that are cited frequently are indicated by abbreviations found in the list below.

Note on State Names

IN ADDITION To the three presidencies (Bengal, Bombay, and Madras) and the other provinces of British India, there were hundreds of princely or native states which recognized the paramountcy of the British Crown. Most were small, but a few were the magnitude of provinces (e.g., Mysore, Hyderabad). When, at Independence, British India was partitioned into the new nations of India and Pakistan, various princely states joined the former provinces as units of the Indian federation, the smaller states being combined for this purpose. The resulting array of states included units quite disparate in scale. A surge of linguistic sentiment led to the separation of Andhra from

Madras in 1953. In 1956 there was a general reorganization of states;

boundaries were redrawn largely on linguistic lines, and the vestiges of the princely states were largely effaced. For example, old (“Princely”) Mysore was combined with the Kannada-speaking parts of Madras, Bombay, and Hyderabad to produce a new and large Mysore; the composite (former Princely) State of Travencore-Cochin was combined with the Malabar District of Madras to form the single Malayalamspeaking State of Kerala. The 1956 changes produced the basic map of India as it is today. Subsequently, two remaining bilingual states bifurcated: Bombay divided into Maharashtra and Gujarat in 1960; Punjab divided into Punjab and Haryana in 1966. A series of small tribal states have since been established in the northeast.

Two name changes should be kept in mind: Madras changed its name to Tamil Nadu in 1969; Mysore became Karnataka in 1973.

List of Abbreviations

A.LR. BCC CAD CWSCST

All-India Reporter Backward Classes.

Report of the Backward Classes Commission (1955). Constituent Assembly Debates (1947-50).

Reports of the Parliamentary Committee on the Welfare of the Scheduled Castes and Scheduled Tribes. The Reports are designated by the number of the Lok Sabha, the number of the report (page no. ifany after colon), date and the year issued. E.g.,

G.O. LSD

[NIC] OBC RCSCST

RSD SASTC ST STTDP

Crore

CWSCST (4th) 18: 27 (1970). Government Order

Lok Sabha Debates. Materials from these reports is cited by series, volume, number, column, and date. E.g., LSD (4th Series) Vol. 86, No. 3, Cols. 142-43 (12 June 1963).

Cases so noted are not included in the computations in chapter 14 which form the basis of aggregate statements throughout the text. Other Backward Classes. Reports of the Commissioner for Scheduled Castes and Scheduled Tribes. These reports are designated by the period covered by the report, rather than by the report number or by date of publication. E.g., RCSCST 1957-58:1, 35.

Rajya Sabha Debates. Material from these reports is cited volume, number, column, and date. E.g., RSD Vol. 102, No. 6, Cols.

118-19 (14 Aug. 1970). Reports of the Scheduled Areas and Scheduled Tribes Commission (1961). Scheduled Caste(s).

Scheduled Tribe(s).

Report of the Study Team

(1970).

on Tribal Development Programs

A unit of one hundred thousand (100,000). A unit of ten million (10,000,000).

Introduction

Inpia’s sysTem oF preferential treatment for historically disadvantaged sections of the population is unprecedented in scope and extent. India embraced equality as a cardinal value against a . background of elaborate, valued, and clearly perceived inequalities. Her constitutional policies to offset these proceeded from an awareness of the entrenched and cumulative nature of group inequalities. The

result has been an array of programs that I call, collectively, a policy of

compensatory discrimination. If one reflects on the propensity of nations to neglect the claims of those at the bottom, I think it is fair to say that this policy of compensatory discrimination has been pursued with remarkable persistence and generosity (if not always with vigor and effectiveness) for for the past thirty years. These compensatory discrimination policies entail systematic departures from norms of equality (such as merit, evenhandedness, and indifference to ascriptive characteristics). These departures are

justified in several ways: First, preferential treatment may be viewed as needed assurance of personal fairness, a guarantee against the persist-

ence of discrimination

in subtle and

indirect forms. Second,

such

policies are justified in terms of beneficial results that they will presumably promote: integration, use of neglected talent, more equitable distribution, etc. With these two—- the anti-discrimination theme and

the general welfare theme—is entwined a notion of historical restitution or reparation to offset the systematic and cumulative deprivations

suffered by lower castes in the past. These multiple justifications point

to the complexities of pursuing such a policy and of assessing its

ce.

Independent India firmly and explicitly embraced the compensatory discrimination principle. This book is about the way in which this commitment has been incorporated into a regime of constitutionally guraranteed rights, including notably rights to formal equality, pro-

tected by courts equipped with broad powers of judicial review. It is an

account of both institutional and intellectual accommodation, of gov-

ernment practice, and of legal doctrine. Among the newer (and almost

2

Introduction

all older) nations, India is distinctive in the major policy role played by an independent judiciary. Indian courts have been called upon both to resolve disputes among contending claimants and to provide a way of harmonizing these competing principles. The focus of this study is on the response of the courts to these unprecedented demands. This study is divided into three parts. Part One sketches the social setting and historical background of the compensatory discrimination policy, describes the major programs and their administration, and opens the question of assessing their performance. Part Two takes up questions of who: who are those people whose entitlement to equality requires departures from formal equality? Identifying the beneficiaries turns out to be a complicated and difficult process involving many intellectual perplexities. Once there is a commitment to these groups, we come to the cluster of questions addressed by Part Three: what is the scope of the commitment? what means are allowable? how much is too much? and what institutional arrangements shall be established to strike the balance? A Note on Terminology: Why “Compensatory Discrimination”? It is not obvious what these policies should be called, for names are not

neutral.

(Consider,

for example,

the contrast between

“affirmative

action” and “reverse discrimination” in the American setting.) Indian

discussion

usually

refers

to

“special

treatment,”

“concessions,”

“privileges,” “special provision,” or “preferential treament” for named ups. Scholars have inclined to use “protective discrimination” (which I first encountered in Alexandrowicz’ 1957 book. More recently, scholarly writing has adopted the term “compensatory discrimination.”? Common usages like “special treatment” are accurate but omit any reference to the principles that animate the policies and distinguish them from other distributive schemes. ‘“‘Protective discrimination” has a paternalistic and static quality; it suggests a one-way flow of benefi-

cence and lacks any implication that it is a policy of change. Thus, of the many unsatisfactory candidates, I prefer “compensatory discrimination” for the following reasons. By using ‘“‘discrimination” it does not blink at the fact that some are left out, that we are dealing with something more than a benign process of inclusion. At least where 1. Alexandrowicz

1957. Among

candidates

that have failed to gain currency are

“progressive discrimination” (Rudolph and Rudolph 1967 : 54n.—on analogy to the

progressive income tax) and “preferential opportunity” (M. D. Nandajundaswamy in

Government of Karnataka 1975, Vol. I, Pt. II, p. 5)

2. Parmanand Singh 1976.

Introduction

3

scarce resources are distributed, it employs a principle of selection that is akin to the old discrimination. But the purpose is different: it is not exclusion and relegation but inclusion and recompense both for historic deprivations and to offset present handicaps. And it carries an implication that discrimination will cease when compensatory treatment has remedied these conditions. The drawback of the term—apart from its being ten syllables—is that it draws attention away from the non-discrimination and general welfare themes in these policies. If this name would be inappropriate in a setting where the element of historical compensation is less prominent,? it does seem to fit the Indian situation better than the other

candidates. To minimize repetition, I have freely resorted toa variety of synonyms and abbreviated forms, such as “preference policy” or simply ‘“‘preference.” Discussions of preferences often founder on the problem of the somewhat confusing names of the groups entitled to receive them. In common Indian usage, the generic term for the groups entitled to receive preferences is the “backward classes.” (“Weaker sections” is a rough equivalent for this usage.) This group in turn may be broken down into the Scheduled Castes (“untouchables”), the Scheduled Tribes, and a diverse and less well defined residuum known as the “Other Backward Classes” or sometimes the “backward classes.” To minimize confusion I have attempted to indicate which usage of “backward classes” is meant by omitting capitalization when using the term in its broad and inclusive sense and capitalizing it when it is used in the narrower sense. Thus “backward classes” = Scheduled Castes + Scheduled Tribes +

Other

Classes.

Backward

Classes;

“Backward

Classes”

= Other Backward

3. E.g., to describe “preferential” programs for “Oriental Jews” in Israel, where the

programs are justified in terms of integrating groups with little previous contact; the more advanced are not deemed responsible for the deprivations of the less fortunate groups.

Part One COMPENSATORY DISCRIMINATION: THE SETTING AND THE POLICIES

This is a book about a set of policies and about the role of the Indian judiciary in developing, elaborating, criticizing, effectuating, and legitimizing them. This Part begins with a sketch of the social setting that frames the policy of compensatory discrimination and is the ulti-

mate object of it. Chapter 2 traces the political developments that link

this setting to India’s modern constitutional order. These broad introductory chapters are intended to make the subject accessible to those

who are not specialists on modern Indian society and politics, as well as

to introduce themes that recur in later detailed analysis of these policies. Chapter 3 provides a brief overview of the compensatory programs and their administration. Chapter 4 examines in more depth the most prominent of these programs (reservation of government jobs). Chapters 3 and 4 provide a sense of the scope, design, and implementation of these and introduce the problems of assessing their performance and of estimating their wider effects.

]

The Compartmental Society INDIAN

soctETy

has

been

described

as

a

‘“compart-

mental” society; within it a vast number of groups maintain distinct and diverse styles of life. The system by which these groups are related and mutually accommodated is so complex as to defy general descrip-

tion. However, to place in their setting the legal developments we shall

discuss, it is necessary to sketch some of the principal features of India’s social order and its recent history. One of the distinctive and pervasive features of Indian society is the division into castes.1 Caste in the narrower sense applies primarily to the Hindus, who make up 85% of India’s population. Many features of caste are also found among non-Hindu groups, and the term ‘‘caste”’ as used here applies to these groups as well as to Hindus. There are exceptions to almost any statement that can be made about caste, but only the most prominent

exceptions will be noted here. The

term

“caste” itself has a variety of meanings. It takes on different shadings in

the context of village, locality, region, and nation.? It is im effect a set of 1. The word “caste” 1s not an indigenous Indian term but a graft via English (from the

Portuguese casta). It is used to correspond to several Indian terms, but has no exact

equivalent in Indian languages. Serious students of Indian society regard it as suspect on grounds of both its foreignness (Marriott and Inden 1974:983) and its ambiguity. Thus

Mandelbaum (1970: 29) says: “It [the term ‘caste’] has been pinned to too many social

entities— endogamous

group, a category of such groups, a system of social organisation.

It seems best to use the term, apart from direct quotation, as an adjective that refers to

the highly stratified social system.” But in the inquiry that we are embarking on there is no way to avoid it, since caste is a key element of contemporary public discourse about

the past, present, and future shape of Indian society. Awareness of the ambiguities and multiple referents of the caste notion will illuminate that discourse. Its theoretical insufficiencies do not allow us to dispense with it if we are to know how Indians (especially those educated in English) see their own society—any more than we could

eliminate equality from a discussion of contemporary American society on grounds of

multiple and contradictory meanings attached to it by the natives.

2. And, of course, it has special meanings in the setting of social scientific discourse

where it is generally used to describe a system of social organization in which society

comprises ranked hereditary kinship groups associated with a division of labor and ized into a unified and integrated whole. The literature is vast. Whether the caste notion is usefully employed outside the South Asian setting is a matter of controversy.

8

Setting and Policies

shifting idioms for talking about affiliation and status.> To start with, a caste may be taken to mean a jati*—an endogamous group bearing a common name and claiming a common origin, membership in which is

hereditary, linked to one or more traditional occupations, imposing on

its members certain obligations and restrictions in matters of social intercourse, and having a more or less determinate position in a hierarchical scale of ranks.5 It has been estimated that there are 2,000 or 3,000 such castes (or subcastes, as they are sometimes called) in present-day India.® The population of these endogamous caste groups may range from hundreds to millions. An estimate places the median

size of such groups at between 5,000 and 15,000.”

Membership in the caste group is ordinarily conferred by birth and is unalterable, although it may be lost by expulsion from the group. Members normally marry within the group; most social intimacy occurs within it. Ordinarily, one does not dine with those who are outside the caste. Members of the caste follow a common round of life. They

have a characteristic diet, domestic routine, style of dress, and round of

religious observance. The caste may have distinct forms of religious expression: a preference for certain gods, shrines, and modes of wor-

ship, its own fasts and feasts. Some cultural characteristics may be

shared by all or most castes over a wide geographical area, but others may be distinctive to one or a few castes. Although most castes have a traditional occupation, in only a few— especially the skilled artisans—is it followed by most members today. Agriculture is in effect open to all groups and is engaged in widely by persons belonging to all groups. In many cases, caste tradition precludes entry into certain occupations, while the network of caste ties

provides access to others.

The caste is held together by ties of kinship, by a cycle of group

observances, by bonds of mutual assistance and support, and finally by

the power of the group to exact obedience to its rules. Most castes formerly had (and many today have) councils, or panchayats, formal or informal, which adjudicate disputes among caste members and enforce 3. The multiplicity of usages not only is apparent to the observer but is internalized in the users who can shift categories (e.g., from gross to fine or local to regional) as the occasion requires. Cf. Mandelbaum 1970 : 22. 4, Jati is a term of wide application meaning kind or genus. Jatis are seen as natural units of society, each manifesting a specific kind of human possibility. 5. The notion that hierarchic ordering is the central organizing principle of Indian civilization is expressed in the title of Dumont’s Homo Hierarchicus (1970). Kindred notions

are shared by scholars who hold otherwise disparate views of Indian society-e.g., Mandelbaum 1970 and Marriott 1955. 6. Ghurye 1969 : 27 (2,000); Hutton 1961 : 2 (3,000). The latter figure is frequently

cited.

7. Marriott and Inden 1974 : 983.

The Compartmental Society

9

the received etiquette of relations with other castes. (For example, they take action against those who offend against taboos on eating, drinking, smoking, or sexual intercouse with members of other castes.) Their decisions, backed by the ultimate sanction of excommunication, may

involve penalties ranging from fines and expiations to one or another degree of boycott, temporary or permanent.® While formal caste sanctions have declined in recent years, especially among the educated, caste opinion and caste loyalty remain powerful cohesive and regulatory forces. Our compartmental image requires correction, for castes are not self-sufficient but interdependent; caste unites while it divides. Although there is a greater admixture of conflict and competition than the harmonious articulation of functions depicted in idealizations of caste, their occupational and ritual specialization does make caste groups interdependent in practice. Caste groups live not in isolation but in the midst of a set of other groups with which they have fixed and customary

relationships— economic,

political, social, and religious. In much of

India the village was organized into a network of hereditary patronclient relations known as the jajmani system.? The various nonlandholding castes (priests, washermen, carpenters, field laborers) provided customary services to landowners in return for fixed portions of agricultural produce. The relationship was not a mere economic exchange, but implied general duties of support and prerogatives of control. The relationships between castes are hierarchic; transactions among groups express principles of precedence, obedience, and sub-

servience. Each group has its duties and disabilities; each, even the lowest, has its special prerogatives and privileges. The number of caste groups in a given locality may vary from a few to more than thirty—from five to twenty would be typical.?° There are

no all-India castes and there is no nationwide hierarchy of castes. In 8. On the mechanism of control within castes, see Mandelbaum 1970, chaps. 16, 17. Although castes have characteristically enjoyed a wide autonomy in making and applying rules for conduct of their own social life, the ruler in classical Hindu theory (and

intermittently in practice until recent times) exercised a supervisory control over changes in custom and over matters of expulsion from and re-admittance to caste. On medieval practice, see Derrett 1968 : 171-224. The requirement of royal approval of re-admittance, retained by the Moghul government and in many princely states until this century, was relinquished by the British in the late 18th century (O'Malley 1932 : 57-61; Hutton 1961 : 93-97). On British control over the internal affairs of castes, see Galanter 1968 and 1972. 9. On the Jajmani system, see Mandelbaum

1970, chap. 9; Wiser 1958; Beidelman

1959. The term jajmani was introduced into the scholarly literature by Wiser. Many other terms are current to describe regional varieties of the system—e.g., the balula, or shares from village lands given hereditary village servants in Maharashtra.

10. Mandelbaum 1970: 15.

10

Setting and Policies

any local area, the relative standing of the various castes is more or less fixed, though subject to slow change, particularly in the middle ranges of the scale. Castes commonly lay claim to higher rank. Thus Marriott and Inden refer to the “mild but ubiquitous inclination of castes to improve their ranks [that] can be estimated from the finding that the

average villager believes the correct rank of his caste to be about halfa rank higher than what non-members of this caste are willing to concede.” There is local variation both in the number and kind of castes and in the rigidity with which the distinctions between them are maintained.

Closely related castes may vary in standing from one region to another,

or even within the same area. The top and bottom of the scale tend to be more easily identifiable and generally, if not invariably, accepted. Brahmins are at the top, and the “untouchables” are at the bottom,

although there may be disagreement about the relative rank of various

Brahmin and untouchable castes. Highness and lowness in the local hierarchy correspond roughly to the ritual purity of the several castes. However, a caste’s standing also expresses its political and economic power. To the extent that these fail to correspond, there will be ambiguity in the ranking of castes, and changing economic and political power may be followed by an improvement in ritual standing. Although there are local differences, highness and lowness are expressed in a variety of interactions. Standing is commonly manifested in the willingness with which food and water are

accepted by other castes, by reluctance to accept food and water from other castes, in the giving and receiving of ceremonial gifts, and in reciprocal patterns of precedence and deference. Relative positions in the local caste hierarchy are often conceptualized in terms of the theoretical division of Hindu society into four classes, or varnas. According to this theory, society is divided into

Brahmins, the priestly and scholarly class; Kshatriyas, rulers and soldiers; Vaishyas, merchants and agriculturists; and Sudras, the me-

nial and servant class. 12 The power of the vama theory is shown by the

common tendency to describe the caste system as made up of four “castes” of which the endogamous caste groups are merely subdivisions. '3 (Hence the usage “sub-caste”.) Groups seeking to better their position in the local hierarchy often phrase their claims in terms ofa more prestigious varna identification by adopting names and practices associated with one of the three higher vamas.™* Sudras are clearly 11. Marriott and Inden 1974:985.

12. There has not always been agreement about the relative standing of the three higher varnas. 13, The classical theory attributes multiplicity to miscegenous mixing of castes. 14. Except for Brahmins, the historical association of caste groups with particular varnas

The Compartmental Socety

11

accounted as inferior to the three “‘twice-born” vamas, but it is misleading to think of the latter as ranked on a single ascending scale of precedence. Instead, they represent a set of divergent and incompatible social ideals around which claims for standing may be organized.45 A caste group may cultivate its identification with one or the other, but does not progress from one to the other. Nor are all groups belonging to

the same varna rated as equals. Between the local caste (i.e., jati) hierarchies and the pan-[ndia varna ranking lie a bewildering and irregular set of regional and subregional groupings of castes and conceptualizations of the caste hierarchy.1¢ There are “caste clusters,” ‘composed of several endogamous groups,” spread over a region or sub-region, sharing “a common name and func-

tions . . . often [with] stylized interaction in commensality, hypergamy

and differential status between their constituent jafi groups.”?7 There are what Fox calls “sub-regional vara categories”—i.e., categories embracing several otherwise divergent jatis, or clusters, who share a claim to a common

status. And there are regional versions of com-

prehensive varna schemes—e.g., the “tripartite division into Brahmin, non-Brahmin and untouchables” in much of South India.1® Local, regional, and pan-Indian schemes all draw on a common ideological fund. The notion of graded inequality is explained and justified by the traditional Hindu notions of dharma and karma. Each caste group has its own dharma— the path which each of its members should follow in accordance with his nature and his station in life. The inequality of stations and the more onerous duties incumbent on some is explained and justified by the theory of karma and rebirth— that is, every human

action has a positive or negative worth, and the moral balance of an individual’s actions in previous lives is manifested in the station into which he is reborn. Inequality is thus explained and justified as the reflection of relative spiritual merit accumulated in past existences. Hope for the future is provided, for one may progress through successive rebirths to higher positions and to the ultimate equality of absorpion or reunion in Brahman, the unindividuated cosmic reality. Such

spiritual betterment is the result of doing properly the things tradition-

ally done by persons of one’s station—by fulfilling the dharma determined by one’s caste, stage of life, and family position. Of course, this is often doubtful. In many places there are few representatives of the Kshatriya and

Vaishya varnas, while in ‘others Kshatriyahood has tended to be conferred on these who

hold temporal power. Since vamas are taxonomic categories rather than functional groups, attempts to enumerate their members have been unavailing. The 1931 Census did account

as

Brahmins

4.5%

(Schwartzberg 1978: 106) 15. Marriott 1968 :110. 17, Fox 1969 : 32.

of the

total

(and

6.4%

of the

Hindu)

population

16. Based on Fox 1969 and sources cited there. 18. Id., at 33.

12

Setting and Policies

ideology is not fully accepted by all.!? Nor does acceptance of these

general notions imply agreement that one’s caste has been accorded its

rightful station.2° So the ideology of hierarchy justifies claims of superiority and rationalizes resignation to inferior status; but it also provides leverage for (and incitement to) claims to higher status. In spite of the idealization of stability and fixity of social position, the compartmental society has not been static: caste groups do change their

relative positions. Caste as a principle of social order has persisted over

millennia, if much more disorderly and asymmetrical in practice than

classical Hindu socio-legal theory depicted it.21 Something very like the compartmental society of the recent past existed a thousand years ago.?? Political, religious, and economic changes have all affected the caste order. The principle of a compartmental society endured while the identity of the compartments underwent many changes. Conquest, migration, emulation, isolation and segregation, occupational specialization, conversion and sectarianism, the incorporation of tribal groups—all led to the additon, fission, and fusion of castes and to changes in their relative standing. Although it stresséd ideals of stability and fixity, the caste society did have mechanisms for incorporating new

groups, accommodating changing practices, legitimating changes in

group ritual vidual ment

standing to accord with changing political economic and socioattainments. The compartmental society permitted some indimobility within these communities as well as the slower moveof whole communities.2> But what endured was a pattern of

19. Berreman 1963 : 253-54. 20. Berreman 1963:223-24. Berreman observes the dilemma of the low caste person whose religion enjoins acquiescence in his low place but whose impulse is to improve that position: “In Sirkanda and, I believe, throughout India, this dilemma is often neatly

resolved by exactly the mechanisms described above: accepting caste dharma but denying

that one’s apparent dharma is his real dharma. That is, the individual takes the position that a person should behave according to his caste status but that he or his kin group or jati

is not really of the caste or status to which others ascribe him (or them). He is of a higher

caste. He aspires to live as a member of his ‘true’ caste in order to fulfill his true Dharma.

Therefore his mobility aspirations are legitimized and even made imperative. He rationalizes his own mobility aspirations without challenging the system.”

21. In classical theory the function of political authority was to hold the balance between the groups in the compartmental society. It was the duty of the ruler to uphold

the caste order. The king was enjoined to make the four vanas fulfill their particular duties

and to preserve society from the evil of “confusion of castes.” The confirmation (and

conferring?) of vama

status and control over the order of precedence among castes by

Indian rulers continued until the recent past (Hutton 1961 : 49; Srinivas 1966 : 41). The

British eschewed such prescriptive assignments; but attempts to obtain governmental confirmation of status were channelled into the nominally descriptive arenas

of the courts and the census (Hutton 1961 : 97; Srinivas 1966:16; O’Malley 1932 : 63). 22. Karve 1961, chap. 2. Nothing like jatis are described in the most ancient classical

literature.

23. Silverberg 1967.

The Compartmental Society

13

graded inequality of corporate groups in which differential access to life chances corresponded in large measure to membership in different communities.

At the bottom of the caste hierarchies are found the “untouchables” — castes with the lowest ritual standing and usually the lowest economic position and traditionally subject to onerous social and civic

disabilities. Discussion of untouchability propels us into problems of

nomenclature which condense intense controversies over the identity of

these groups and over the characterization of their relation to the bulk of

the Hindu population.4 We shall trace the recent history of their

identification in chapter 2 and take up some current controversies in chapter 6. For the moment it suffices to note that there is no other general term for the practice of setting apart a segment at a lower end of the caste order and that “‘untouchability” has been adopted as a salient category for public policy: the Constitution declares its abolition, and successive governments have promised to end it summarily. The origins of untouchability as practiced in recent times are unk-

nown. Something like it has existed for at least a millennium,?5 and many of its features were present more than a millennium earlier than that.?6 Untouchability as such is not dealt with in classical Hindu learning. However, the low state of untouchables is reflected in and supported by the classical view of society as consisting of four varnas, or classes, with different natures, different functions, and correspondingly

different stations in life. Rights and duties, procedures and penalties, were graded accordingly. The three higher vamnas were “‘twice-born” by virtue of investiture with the sacred thread, while the Sudras were

subject to numerous disabilities. They were restricted to certain occu-

pations, forbidden to learn the veda, and enjoined to be deferential and subservient to their betters. In recent times, “untouchables”

have been identified as “unclean

Sudras” in some parts of India; elsewhere they are regarded as outside 24. Current terminology includes the official term “Scheduled Castes”; also used

officially is the Gandhian euphemism “Harijans” (lit., “people of god”), favored by politicians and resented as condescending by many so designated. “Dalits” (i.c., oppres-

sed people) is favored by militants and eschewed by politicians and officials. “Backward classes” has a wider connotation (see chap. 6 below), as does “weaker sections.” “‘Depres-

sed classes,” no longer in vogue, was the predecessor official term, replaced by Scheduled Castes in the mid-1930s. “Untouchables” is regarded as callous or insulting. Isaacs’ (1965) usage, “‘ex-untouchables,” has not achieved currency. On origins of some of these

terms, see notes 21 and 46 in chap. 2 below.

25. Alberuni (c.A.D. 1020) noted certain endogamous occupational groups who lived outside villages and had limited intercourse with others (Ghurye 1969 : 316). He mentioned Dom and Chandala as “two groups not reckoned among any caste” (Ghurye 1969 : 313). A 12th-century grammarian reports that Chandalas warn people to avoid their touch (id.). 26. R. S. Sharma 1958 : 125.

14

Setting and Policies

(beneath) the four varnas.2” Hence untouchables are often referred to

as “outcastes,” and there is a persistent tendency for foreign observers

to describe them as people who are not members of any caste.?® But they

are members of castes—those which have the lowest ritual standing and

often the most depressed economic condition. Generally, they are also lowest in income, health, education, and cultural resources. In the 1971

census those officially classed as untouchables (Scheduled Castes) by the government constituted nearly 80 million, or about 14.6% of India’s total population of 548 million (or over 17% of India’s 453 million Hindus). Untouchables are often associated with the most unclean and

degrading occupations—sweepers and scavengers, attendants at cremation grounds, hide and leather workers—but include other artisan groups such as weavers and toddy-tappers. In most regions a single large caste of field servants

(like the Mahars

in Maharashtra,

the

Paraiyans in Tamil Nadu, and the Chamars in much of North India) makes up a large portion of the total untouchable population, with the remainder divided among various untouchable artisan castes.?° The groups considered untouchables differ from locality to locality. A caste such as washermen or weavers may be considered untouchable in one

part of India but not in another. Actually, only a small minority of today’s untouchables practice these stigmatized occupations; most are landless agricultural laborers, often ensnared in debt. They are usually, but not invariably, associated with certain habits whch are unclean and loathsome to higher caste Hindus—such as eating beef and other forbidden food and in some cases eating carrion. The relations between castes throughout the hierarchy are expressed

in terms of relative social distance and exclusiveness corresponding to the necessity of avoiding pollution (i.e., ritual uncleanliness). Pollution might

derive from association with death, menstruation,

childbirth,

impure food, etc. But while everyone is subject to and causes temporary and expiable pollution, “untouchables” are permanently and inexorably polluted and polluting. Pollution is transmitted by contact with their persons, by food, and in some places in the south, by proximity (and historically in few cases, sight) as well. Different untouchable castes might be polluting to different degrees and subject to disabilities to varying extents. They are not a unified group, but recognize such distinctions among themselves and sometimes impose similar disabilities on each other. 27. R. S. Sharma (1958 : 208, 261, 279) detects comparable ambivalence in the relation

of untouchables to the Sudra vama two thousand years ago. 28. “‘Outcaste” also means expelled from one’s caste. Its use to describe untouchables flows from the identification of castes and vamas; i.e., not being included in a vama is equated with not having a caste. 29, Patankar and Omvedt 1979 : 410.

The Compartmental Society

15

Untouchable groups are set off more or less distinctly below all of the

other caste groups.0.29 VWhile each caste enjoys certain precedence and is subject to certain disabilities, the “untouchable” castes are subject to more extensive and severe disabilities and enjoy few privileges. Disabilities vary from place to place and in their incidence on specific castes. The imposition of disabilities is not unique to untouchables. It exists to some extent—particularly in regard to religious facilities, commensality, and intermarriage—throughout the system of castes. But it is found in a more exaggerated form for untouchables and to a lesser degree for other groups; some disabilities are peculiar to untouchables or more intense in regard to them. The following list of disabilities applied to untouchables is by no means a description of the practices in any single locality, nor is it an exhaustive listing of practices throughout

India. It is instead a catalog of the disabilities which have in one place

or another been typically associated with untouchability:

Denial or restriction of access to public facilities, such as wells, schools, roads, post offices, and courts.

Denial or restriction of access to temples where their presence might pollute the deity as well as the higher-caste worshippers, and from resthouses, tanks, and shrines connected to temples. Untouchables and Sudras

were ineligible to become sanyasis (holy men) and forbidden to learn the Vedas (the earliest and most sacred books of orthodox Hinduism). Exclusion from any honorable, and most profitable, employment and relegation to dirty or menial occupations. Residential segregation, typically in a more extreme form than the segregation of other groups, by requiring them to remain outside the village. Denial of access to services such as those provided by barbers, laundry-

men,

restaurants, shops, and

theaters or requiring the use of separate

utensils and facilities within such places. Restrictions on style of life, especially in the use of goods indicating comfort or luxury. Riding on horseback, use of bicycles, umbrellas, footwear, the wearing of gold and silver ornaments, the use of palanquins to carry bridegrooms—all of these were forbidden in many areas. Requirements of deference in forms of address, language, sitting and

standing in presence of higher castes.

Restrictions on movement. Untouchables might not be allowed on roads and streets within prescribed distance of the houses or persons of higher castes.

Liability to unremunerated labor for the higher castes and to the perform-

ance of menial services for them.

30. Although untouchables might readily be identified in a given locality, considerable flificulty was experienced in devising pan- - Indian criteria to distinguish them. See chap. 5

16

Setting and Policies

Restrictions were severe and ubiquitous, but custom did not always prescribe complete exclusion. Instead, there was a complex pattern of exclusion, separation, and restricted use. For example, wells might be entirely closed or might be used with separate buckets or only througha higher caste intermediary. Schools might be entirely closed, or untouchable children might be required to sit either outside the classroom or inside but away from other children. Untouchables might be excluded from temples, or they might be permitted to enter but required to remain a specified distance from the idols. The opportunities for group mobility within the caste system were more or less closed for them.31 Systematically degraded, subjected to paternalistic control where not harshly victimized, the untouchables endured a kind of perpetual slavery with little hope of betterment in this world. Caste is not the only basis of affiliation in this compartmental society. Sometimes enveloping and sometimes cutting across caste are many other sorts of groupings. Indian society was and is divided into a multitude of religions, each containing its own orders, denominations, sects, and movements. The great family or cluster of religious assemblages known as Hinduism (containing some 85% of the present population of India) includes numerous “unorthodox” and “reform” sects as well as various “orthodox” Vedic and Brahmanic communions, em-

bracing an amazing diversity of beliefs and practices. The contours of Hinduism are blurred in many areas. The line between Hinduism and tribal religions is indistinct. Major groups like Lingayats, Jains, Sikhs,

and

Buddhists are accounted

Hindus

for some

purposes, but are in

other respects regarded as non-Hindus. The adherents of the religions that

originated

outside

India— Muslims,

Christians,

Parsis,

and

Jews—are seen as distinctly outside the Hindu fold, but here too there are instances of syncretism and overlap. There are caste groupings that cut across religious lines to include adherents of several religions—Jains and Hindus, Sikhs and Hindus, Christians and Hindus. The movement

into non-Hindu religions of members of low-ranked castes, a recurrent recourse of the oppressed, has been accompanied by the development of communal distinctions in the new fold. Similarly, sects within Hinduism opposed to caste distinctions have been regularly absorbed into the compartmental order. Even where they are distinct from Hinduism, all of these religious groups form “communities’”3?—

31. Srinivas 1966. But even this is not without exception. The Nadars of southern Tamil Nadu and the Ezhuvas of Kerala provide two examples of groups that were regarded as untouchables early in this century and managed by strenuous efforts to raise their social standing as well as their political and economic position (Hardgrave 1969, chap. 5;

Aiyappan 1965 : 133-46.)

32. “Community” in recent Indian usage is used to refer to a section of the population

The Compartmental Society

17

components of the compartmental society. They either constitute or contain endogamous castes or caste-like groupings. Outside of and overlapping with both caste and religious groupings are the tribal communities, concentrated in isolated hilly regions. Some

follow tribal religions, most are accounted as Hindus, and a significant minority are Christians. Thus, although the endogamous Hindu caste (jati) is the paradigmatic unit of the compartmental mosaic, in practice it includes communities which are neither Hindu nor castes. The compartmental structure of communities is crosscut by other bases of cleavage and affinity—village and regional loyalties, occupational and class ties, factional and political alliances. In recent generations all of these have been overlaid by differences in exposure to and acceptance of “modern” or “Western” cultural influences. Our picture of the compartmental society is static and understates the impact of recent changes. Important qualifications must be made for settings where communal identity has declined in salience—as among educated urban elites and where class formations may have eclipsed community as a basis of group mobilization. The compartmental society’s notions of prestige and models of affiliation have been both en-

larged and effaced by “Western” models.*3 These and other elabora-

tions will be introduced as we proceed. For the moment, a sense of a society comprising ranked communal compartments provides a useful base line to understand the policies whose development is described in this book. differentiated by religion or caste or both (e.g., the Muslim community, the Brahmin community, the Agarwal Jain community). The meaning shifts according to the context

(e.g., Hindu community comprises both Brahmin and non-Brahmin communities and

these in turn may be divided into communities). “Communal” does not mean things held

in common, but refers to any arrangement or view which emphasizes religious or caste

groupings as social or political units. ‘“Communalist” and “communalism” refer to persons or ideologies that support such arrangements, and they are ordinarily pejorative. 33. Thus in a seminal series of works, M. N. Srinivas contrasted “westernization” and

“sanskritization” as methods for achieving higher standing. See Srinivas 1966, chaps. 1 and 2.

2

Reform, Mobility, and Politics under British Rule

EARLIER

INVADERS

and

conquerors

had

accom-

modated and eventually been absorbed into the compartmental social order. The British came to India as traders and conquerors, not as social reformers. By the early 19th century, when they had fastened their rule over most of India, British policy was to recognize and accommodate the Indian social order—a policy they termed one of non-interference. But the impact of British rule was to alter the social order profoundly, if inadvertently. Conditions of peace, new communications, new economic activities, new kinds of employment, a new legal

system, a new system of property relations, and new ideas brought in

their train new opportunities and new modes of mobility. Old powers and prerogatives were abolished; occupations and learning were rendered obsolete or marginal; new opportunities for gain and advancement were introduced; power and access to it were redistributed. Groups rose by association with the British Government, by trade with the British, by utilizing the new educational opportunities that the British introduced, by manipulation of the new legal facilities. Advantages gained in these ways tended to be cumulative, for utilization of these new opportunities was biased by the existing distribution of resources and by the network of kinship and community which offered access to them. 1. In addition to works specifically cited, the account in this chapter relies generally

on the following sources: on social reforms, Heimsath 1964 and S. Natarajan 1959; on

changes in Indian society, Ghurye 1957, Srinivas 1966, and Rudolph and Rudolph 1967; on the untouchables and their movements, Zelliot 1969 and Dushkin 1957; on the non-Brahmin movements, Irschick 1969 (South India), Omvedt 1976 (Bombay), and Dushkin 1974 (Mysore); on Gandhi's views and activites, Gandhi 1954, Dalton 1967, and Pyarelal (Nair) 1932; on Dr. Ambedkar, Keer 1962, Zelliot 1969, and Ambedkar

1945, 1946, and 1968; on legal developments regarding caste generally, Galanter 1968, and specifically on legal enforcement of and abolition of social disabilities, Galanter 1972a; on the drafting of the Constitution, Austin 1966 and the Constituent Assembly Debates.

British Rule

19

Non-interference proved elusive, for whatever the British did or refrained from doing was bound to have an impact on the social order. The paradoxical aspect of the policy of non-interference is plainly revealed if we consider the disposition of the legal system to the compartmental social order. Non-interference implied doing what rulers in India had always done—actively upholding and supporting the caste order.? But it also implied an aloof neutrality. The outcome was a changing patchwork of support and aloofness, threaded by impulses of reform given intermittent and reluctant expression, all bent and reshaped in unplanned ways by forces released by the British presence. The establishment of a nationwide legal system brought a general movement of disputes from caste and village tribunals, responsive to the locally powerful, into the government's courts and spread a consciousness of rights which might be vindicated independently of local opinion. The government's courts espoused a norm of equality before the law. With few exceptions, the same rules applied to all. Outside the “personal law” fields, sastric and customary law was supplanted by a general law applicable to all. The British undertook to apply separate bodies of law in matters of family, property, and religion according to the religious affiliation of the parties. Application of these personal laws in the government's courts, with common-law procedure and commonlaw trained judges, introduced inadvertent changes. In the formal setting of the British-style court, the classical textual element of the law was elevated to a supremacy over the customary element greater than it had enjoyed earlier.? (In Hindu law, for example, the salience of the varna categories was greater in the late 19th century than earlier.) The general features of the legal system were not articulated to a system of graded inequality. Legal enforcement of slavery-was withdrawn.‘ The use of caste in the general civil, criminal, and commercial

law was sharply restricted and soon abandoned. The British law did not recognize or try to maintain the caste order as such. The law recognized

the autonomy of castes as groups. But, after an initial flirtation, the

British were reluctant to provide affirmative support for their internal powers of governance. Affirmative support was confined to a narrower ambit than in the past, when Hindu rulers gave redress for grievances which the Anglo-Indian courts excluded as involving only social pri2. Derrett 1968, chap. 7; Jackson 1907; O’Malley 1932. 3. Derrett 1968, chaps. 8. 9; McCormack 1966; Rudolph and Rudolph 1967, pt. 3;

Galanter 1968.

4. Legal enforcement of slavery was outlawed by the Indian Slavery Act (Act V of 1843); possession of slaves was made a criminal offense by the Indian Penal Code (Act

XLV of 1860) §370. On the way that agrestic slavery in southern India compounded the

diabilities of the lowest castes, see Adam

1840: 171 ff.; Banaji ca. 1933:85; Kumar 1965.

20

Setting and Policies

vileges. Yet the British courts were more ubiquitous and accessible, more efficient and relentless in operation; even their attenuated and indirect support had great impact.5 . The British were less willing than earlier rulers to lend the civil power to uphold prerogatives claimed in inter-group relations. The assertion of precedence and the imposition of disabilities received limited and for the mest part indirect support from the law.® Caste groups enjoyed the active support of the courts in upholding their claims for precedence and exclusiveness respecting the use of religious premises. Exclusionary practices did not enjoy the same active judicial support—at least not from the higher courts— in regard to “secular” public facilities such as schools, wells, and roads.

However, where members of higher castes themselves undertook to “enforce” their prerogatives against lower castes, the courts were reluctant to provide remedies against such “self-help” — including boycott and secondary boycott. In their respect for autonomy of castes, the courts were reluctant to interfere with the disciplinary powers of castes against violators of received etiquette of inter-caste relations. While at the upper levels of the legal system, support for high caste dominance was mainly indirect and passive, there is some indication that at the lower levels of the legal system there was often active governmental

support for the claims of the higher castes.

Even where lower castes enjoyed rights that were formally enforce-

able, they could be vindicated only by engaging in the expensive and uncertain process of litigation, a process only partly insulated from other sources of advantage. Surveying a range of studies, Cohn concludes:

The lower castes have generally been unsuccessful when, through the use of police or of the urban courts, they have sought to redress what they believe to be the corporate wrongs done to them by upper castes. The upper castes maintain their economic position, their knowledge of the courts and the intricacies of the law and better access to officials have thwarted attempts to change the position of the lower castes in the village society and economy.”

If in practice legal institutions often adapted themselves to prevailing patterns of disability, the lowest castes did enjoy a new equality in the 5.Galanter 1968. 6. Anandan 191] : 529 observes that the famous proclamation of 1858 in which Queen

Victoria assured Indians that none would be favored, molested, etc., by reason of their

religion or observances was “tantamount to saying that the state will not interfere in the most provoking inter-caste observances and the consequent ill-treatment of the depressed classes. In other words, the state cares to respect only the feelings of the high caste ple... . 7. Cohn 1965: 108.

British Rule

21

eyes of the law and had access to it, at least formally. The law opened

some possibilities for advancement and change to the lower castes, as it did to others. But it did not provide any special leverage for the lowly to use these opportunities, so use of them tended to correspond to the existing distribution of resources. Just as, for example, new opportunities for advancement through education and government service were grasped largely by Brahmins and other high castes with a literate tradition, so those who were already supplied with other advantages tended to make use of the new opportunities afforded by the legal system. The law did not provide the higher castes with an instrument for aggressively suppressing their inferiors (as, e.g., Jim Crow laws did in the post-Reconstruction southern United States), but the

law provided another resource which higher castes could use to protect their claims to precedence (and in some instances perhaps even tighten their hold on valued resources to the exclusion of lower castes).®

Most of the new opportunities in education and government services were grasped by. those already supplied with advantages,? but some members of the lower castes did find new opportunities as menial government servants, in army service, and in trade and contracting (sometimes protected by their monopoly in a degraded occupation like leatherwork or sweeping). British courts declined to enforce customary service relationships. But while this meant that weaker parties could not invoke the assistance of the courts when traditional expectations were disappointed, such assistance was hardly required by dominant groups. Many artisans and village servants were displaced from their traditional callings and prerogatives by new technologies and new structures of government.?° The problems

faced by the lowest castes in utilizing these new

opportunities are exemplified in their struggle to obtain schooling. In 1856 a Mahar boy, refused admission to a government school, appealed to the Bombay Education Department, but was rebuffed on the ground

that

it would not be right for the sake ofa single individual, the only Mahar who had ever yet come forward to beg for admission into a school attended only by pupils of caste, to force him into association with them, at the probable risk of making the institution practically useless to the great mass of natives.14 8. Chidambaram Pillai (1933) argues that British law recognized and promoted more restrictive claims of exclusiveness regarding temples in South India than had prevailed earlier. 9. Ghurye (1957 : 190) notes that government service was staffed by higher castes and that their traditional attitude to lower castes carried over into government work. 10. Zelliott 1969 : 53; Galanter 1972a : 233 and n. 41.

11. Quoted at Zelliott 1969 : 47.

22

Setting and Policies

In 1858 the Bombay Government announced that it reserved the “full right of refusing . . .support . . .toany . . .school in which the benefits of education are withheld from any class of persons on account of caste

or race,” and further resolved ‘“‘that all schools maintained at the sole

cost of Government shall be open to all classes of its subjects without distinction.”!2 Nevertheless, reported the Department of Public Instruction, ‘“‘the opposition of the higher castes to the admission of [Untouchable] boys to a public school was often so strong that, even with the best will in the world, the Department could do very little in the

matter.””3

A few special schools were opened by reformers, by missionaries, and by the government. In regular schools, the few “untouchable” children were subject to invidious segregation—being forced to sit away from others on the veranda was a typical experience. Some concessions in primary school fees and special scholarships in secondary schools and colleges were made available in the late 19th century. But not until the 1920s was there more than a trickle of untouchable students. Thus, non-interference combined with rising aspirations and expanding opportunities to produce a complex and uneven movement; it both liberated the lower castes from official imposition of disabilities and enabled a fortunate fraction to raise themselves, while it added

richly to the array of advantages of the higher groups. The law, educa-

tion, government service, and the new economy carved out and defined

a sphere of public life in which caste distinctions were of little import; but the thrust of British policy was to give the higher castes more resources and opportunities to enter and succeed in that sphere. The growth of a class of Indians educated on Western lines, the spread of British political and social ideas, and the insistent criticism of missionaries all contributed to an inteliectual ferment which questioned many aspects of Indian life. Practices which had appeared natural and inevitable now seemed problematic. Religious reformers responded to

the humanist and rationalist critique of Indian institutions with pro-

grams to reconsturct and purify Hindu society. Groups of educated Hindus devoted themselves to refashioning social institutions to accord with new notions of humaneness, justice, and equality. The great reform issues throughout most of the 19th century were the abolition of sat, remarriage of widows, education for women, and opposition to child marriage—issues which reflected higher caste practices and options. Caste restrictions (like ostracism after sea voyages) were an issue, but not the system of caste itself. Reformers were not indifferent to the strictures of caste, of which they were painfully 12. Ghurye 1957 : 189. 13. Quoted at Zelliot 1969 : 48,

British Rule 23 aware,since the strictures often bore with fearsome force on those who

outraged conservative opinion by putting into practice such reforms as widow remarriage.

The currents of reform stirred others than emancipated intellectuals. The changes under British power gave new scope to the impulse to self-improvement and advancement in many groups. In the late 19th century, many castes organized, held conferences, published newspapers, and undertook internal reform efforts. At first they tended to

seek advancement in traditional terms of ritual status, vying for pre-

stigious entries on the census, putting forth claims to exalted varna

position, and upgrading their practices in terms of received notions of respectability. (This sometimes involved changes like adopting child marriages and banning widow remarriage—the very practices opposed by reformers in the highest castes!) The interest of caste organizations turned increasingly to secular betterment: they founded hospitals, hostels, and loan funds; they sought representation, government jobs, and scholarships. Caste organization brought with it two important and related changes in the nature of castes. The salient groups grew in size from the endogamous jatis into region-wide alliances. Concomitantly, the traditional patterns of organization and leadership in the village setting were

displaced by voluntary associations with officials whose delimited authority derived from elections. New forms of organization were taken up by some castes at the lower end of the spectrum.'* Some embraced the opportunities afforded by

missionaries; others were influenced by Hindu reform movements; still others developed their own organizations. The first militant action by low castes was the campaign of the Shanars (later Nadars) to enter

temples in the Tinnevelly District in Madras, a demand connected to their claim to be Kshatriyas. The Ezhuvas (Iravas) of Kerala took a different tack: by the early years of the 20th century, they had developed their own system of temples and schools. The first efforts to assist the low castes had come at mid-century

under the auspices of a Poona reformer, Jyotirao Phule, who united a

program of education and self-help with a thoroughgoing attack on caste,!5 from missionaries who proffered conversion as a solution,'®

and from Hindu religious reformers like Dayananda Saraswati, founder of the Arya Samaj, who preached a purified varna and attacked caste exclusiveness (in its existing degenerate form).!7 Although anti-caste associations had earlier voiced opposition to 14. 15. 16. 17.

Lynch 1969 : 67-69; Hardgrave 1969 : 130-32. Keer 1964; Ghurye 1957 : 200-1; Omvedt 1976: 106-23. Heimsath 1964; 53-54. Heimsath 1964 : 321, 330, S. Natarajan 1959 : 70; Ghurye 1957 : 258

24

Setting and Policies

caste distinctions, it was not until the end of the century that mainstream reformers saw caste hierarchy and inequality as problems in their own right. Now reform of the caste system, rather than of the practices of individual castes, became a major plank of the reformers. (The issue derived dramatic impact from its juxtaposition with the discriminations suffered by Indians in South Africa, which had become

a public issue in India in the 1880s.) In 1895, Justice M.G.Ranade of

the Bombay High Court circularized social reformers regarding the claims of aboriginals and untouchables on Hindu society. In the next dozen years, missions, schools, and scholarships for Depressed Classes were founded. Not only did amelioration of the lowest castes become one of the most prominent items of reform, but it was accompanied by a shift in attitude toward the caste system. Earlier reformers sought improvement from within. At first they had concerned themselves with sastras, trying to provide textual authority for such reforms as widow remarriage and abolition of child marriage. Gradually, they had come to realize that sastric learning provided little leverage to change current custom. It was the adherence of the caste

group that upheld custom, not sastric learning. And, improvement from within by caste associations engaging in self-reform, it was discovered, cut both ways, for a caste might embrace orthodox usage as well as depart from it. This kind of piecemeal approach came to seem less inevitable as the image of Indian society changed. The notion of reforms proceeding from and directed to a single national community transcending divisions of religion, caste, and sect was not self-evident in the late 19th century. Reform of social practice was considered appropriate for each constituent group within the society.

Thus, in his presidential address to the second meeting of the Indian

National Congress in 1886, Dadabhai Naoroji explained his objections to making social reform part of the Congress program:

How can this gathering of all classes discuss the social reforms needed in

each individual class? What do any of us know of the internal home life, of

the customs, traditions, feelings, prejudices of any class but our own? How

could a. . .cosmopolitan gathering like this, discuss to any purpose the reform

needed

in any one class? Only

the members

of that class can

effectively deal with the reforms therein needed. A National Congress must confine itself to questions in which the entire nation has a direct participation, and it must leave the adjustment of social reforms and other class questions to class Congresses. *®

But as nationalism grew, this division into “classes” seemed less natural and commendable. The caste institution itself fell into bad 18. Naoroji n.d., p. 8

British Rule 25 repute, blained for disunity, weakness, and lack of civic spirit. Increasingly, reformers saw the caste system itself as the enemy. In the early years of the century, pleading for the relaxation of restrictions among “sub-castes” was replaced by opposition to caste itself. Among the most radical, the aim was not merely to raise the conditions of the lowest castes, but to abolish the system of caste altogether.’9 The ferment within Hindu society has to be seen in the context of changing Hindu-Muslim relations. British rule shattered the earlier patterns of dominance and accommodation, permitting, ifnot fostering, open rivalry that found outlets in new political arenas.2° The concern of Muslims over the loss of their former preeminence was compounded by the feeling that they were falling behind the Hindus in modern callings and opportunities and were being reduced to a subordinate position. As the movement toward greater Indian participation in government gathered momentum in the late 19th century, it threatened to accentuate this imbalance, compounding Muslim anxiety about being dominated. In 1906 a deputation of Muslim notables petitioned the Viceroy to safeguard Muslim interests in the forthcoming electoral regime, reserving seats to be chosen by separate Muslim electorates. The British were responsive to Muslim apprehensions. In 1909 when the Minto-Morley reforms were promulgated, separate electorates were provided for Muslims. At this point, the condition of the untouchables—as they were com-

ing to be called?1— which had become a major item of reform interest,

was propelled from the realm of philanthropy and social uplift into the political arena. Rising concern about the untouchables was augmented by the first appreciation of their political significance. Eleanor Zelliot

e3:

19. K. Natarajan, a prominent Bombay reformer, first formulated the notion that untouchability was an inevitable concomitant of the caste system; to remove it, caste must itself be removed (S. Natarajan 1959: 120). 20. Barrier 1968; Jones 1968.

21. In 1909 the problem of the lowest castes was for first time conceptualized under the rubric of “untouchability”—a general term which opened the possibility of visualizing the problem not as that of a congeries of depressed groups, but as a stratum of all-India dimension with shared characteristics. The vicissitudes of the search for that

stratum are detailed in chap. 5, §A, below. Use of the English term “untouchable” in

print can be traced with fair precision to the year 1909. The Maharaja of Baroda in his

remarks to the Depressed Classes Mission of Bombay on October 18, 1909, uses the

term and provides an explanation to his audience (Sayaji Rao 1928 : 244-45). The abstract form “untouchability”’ is used by Sridhar V. Ketkar (1909 : 86) in a footnote to his study of the caste system; the preface of the volume is dated September 1909. Internal evidence suggests that the author did not have this term at his disposal when he wrote the text (cf. pp. 99 and 121-22). The Maharaja of Baroda was the patron of Mr. Ketkar and supported him during the years at Cornell University when he wrote this book. (See Ketkar 1911: xxvi). Weitbrecht-Stanton (1920: 173) attributes the term’s prominence to the Gaikwad [Maharaja].

26

Setting and Policies The granting of an electorate for Muslims in which they alone would vote brought the idea of communal electorates for minorities to the forefront in

the minds of all communities which feared for their submersion in a government run by the dominant caste Hindu community. The granting of special electorates to the Muslim community also made numbers important.

Whether the -vast numbers of Untouchables were truly Hindu and to be counted as such, or not, became an important question for the first time.??

Concern about Hindu numbers was accentuated when the Census

Commissioner (for 1911) bruited the suggestion that the untouchables

(then known as Depressed Classes) should be enumerated as a group

Separate from

the Hindus,

a suggestion promptly endorsed

by the

Muslim League.” Suddenly there was quickened interest in the Depressed Classes from previously indifferent Hindu leaders who were now insistent that Depressed Classes were within the Hindu fold.% Apprehensions about a continued Hindu majority must have been seen in the light of then current estimates that the Depressed Classes numbered 50 to 60 million**—from a sixth to a fifth of the total population and roughly themagnitude of the plurality of Hindus in the population.”© “Untouchability” was thus propelled into a prominent (but not central) place on the Indian political stage, which it was to occupy until after Independence. This new position was soon reflected by the reversal of the Indian National Congress’s long-standing policy of excluding “social reform” from its program. At its 1917 annual meeting, the Congress for the first time passed a mild and hesitant anti-disabilities

resolution.2”? As Congress had avoided social reform because of its

divisive potential, it now embraced it (in part) because of its unifying 22. Zelliot 23. “Since of 1909] the outcastes are

1969: 141. the publication of the recent Reform Measures [the Minto-Morley reforms Mahomedans have been loud in declaring that, properly speaking, the beyond the pale of Hinduism, and therefore their strength should not go to

swell the numerical force of the Hindus. . . .” (Saint Nihal Singh [in Jndian Review,

September 1910, reprinted] in Anonymous ca. 1913 : 89.) 24. A contemporary observer reports that “‘a question that is agitating Hinduism at

the present moment is as to whether these [untouchable] classes should be counted as

Hindus or not. Ten years ago the answer would have been emphatically in the negative. Even now the conservative feeling of the country is for their exclusion. But the conscience of the more advanced section of the educated Hindus is a little more sensitive on

this point” (Holderness 1911:101-2). Cf. Tinker 1962:131, who reports that “‘as late as

1910, upper caste politicians were demanding that [untouchables] should be excluded from the Hindu fold in the Census.” 25. On these estimates, see chap. 5, §A, below. 26. The Hindu percentage of the population had been falling slowly but steadily for decades. See n.45 below. 27. “The Congress urges upon the people of India the necessity, justice and righteousness of removing all disabilities imposed by custom upon the Depressed Classes, the

British Rule

27

potential. From this point on, the social question was indissolubly linked to the political. At the time of Congress’s first gesture of interest in the untouchables, there was only a faint scatter of political activity among untouchables

themselves. There was, however, more noticeable stirring among the

castes above them but below the social peaks—i.e., in the middle ranges of Hindu society. Movements of the middle castes against the cultural preeminence of Brahmins had emerged in the 19th century; now, focussing on Brahmin predominance in education and government service, they took on political form. The Justice Party (South Indian Liberal Federation), the organ of the Non-Brahmin movement in Madras,

was founded in 1916; the organization of the Non-Brahmin

party in Bombay got under way in 1917. The Princely State of Mysore instituted a system in which all communities other than Brahmins were

denominated “Backward Classes;” from 1918, places were reserved for

them in colleges and state services and thus was inaugurated the first modern regime of communal quotas. Non-Brahmin demands for a greater share of political representation were quickly recognized by the British. The promised post-war reforms (the Government of India Act,

1919) accorded them special representation. During the 1920s, the nonBrahmin parties flourished in the Madras and Bombay presidencies,

erecting a system of preferences in educational admissions and recruitment to government posts. At first, these non-Brahmin, or Backward

Classes, movements undertook to speak for the lowest strata as well. but as they prospered, their voices increasingly reflected their higher echelon constituents. The 1919 reforms provided only a few nominated seats for the Depressed Classes,”® but for the first time some untouchable spokesmen were heard in political assemblies. As untouchables increased their political activity and as concern for their welfare became a plank of Congress policy, legislatures for the first time took an interest in their problems. There was a scatter of undertakings by government to provide land,

housing, schooling, and government posts to the Depressed Classes. The illegality of denying access to schools, wells, and roads was de-

clared in legislative resolutions and administrative orders— which

disabilities being ofa most vexatious and oppressive character, subjecting those classes

to considerable hardship and inconvenience’* (quoted in S. Natarajan 1969 : 144). The

Congress resolution omitted a crucial passage of the original, as submitted by the

Depressed Classes Mission Society, which specified ‘disabilities imposed by religion and

custom ”’ (italics supplied). 28. On the isolation of untouchables from non-Brahmin movements, see Irschick

1969: 188fF Zelliot 1969: 143 ff; Dushkin 1974:252 ff; Ramaswamy 1978:296. 29. Depressed Classes representation was increased (and extended to include a

nominated member in the Central Legislative Council) following the findings of the Muddiman Committee in 1925 (Zelliot 1969: 158).

28

Setting and Policies

were honored largely in the breach.3° Special government officers were appointed to look after their welfare. There were fee concessions and scholarships and some pressure on schools to be accessible. School enrolments increased, and a few untouchables entered public services.31 By the close of the 1920s, as Eleanor Zelliot observed, “the

principle of special attention was firmly established.”3?

Untouchability (and the excesses of caste hierarchy) presented a problem whose solution was required to unlock India’s national destiny. The question remained how this was to be done. The various answers might roughly be classified by their resemblance to two polar types that we may call the “evangelical” and the “secular.” The “evangelical” approach stresses the uplift of untouchables to higher Hindu standards and the penance of caste Hindus for the injustices of untouchability, which is seen not as an integral part of Hinduism but as some external impurity. Uplifted untouchables and repentent Hindus will join together in a purified and redeemed Hinduism. At the other extreme, the secular approach stresses the denial of civil rights and economic opportunities and seeks to combat them by vigorous government intervention, and political action to obtain it. Untouchability is seen as an integral part of Hinduism. The secular approach does not seek to restore Hinduism to some original purity, but seeks instead to liberate all by dissolving the graded inequalities of caste into the common bond and common rights of citizenship in a secular state. In the “evangelical” approach, the emphasis is on moral regeneration and philanthropic uplift; in the secular approach, it is on civic and economic improvement under government auspices. These divergent approaches were present from the first stirrings of concern about the untouchables. In the 19th century, the evangelical 30. E.g., in 1923 the Bombay Legislative Council resolved that untouchables be allowed to use all public watering places, wells, schools, dispensaries, etc. The Provincial Government did not, however, undertake direct responsibility for enforcement as to

local facilities, but requested “collectors . . .to advise the local bodies in their jurisdiction to consider the desirability of accepting the recommendations made in the Resolution so far as it relates to them” (Bombay 1930 [Starte Committee]: 52). The flavor of

enforcement activity is conveyed by the action of a District Board which, seven years later, resolved to post notices that facilities were open to untouchables “‘at only those villages in the district where the public opinion is favorable for such action” (quoted at Sanjana 1946 : 237). At that time, the Starte Committee concluded that the policy was a “complete failure” when it could not “find a single instance where Depressed Classes are continuously using the same public well as the higher classes. There may be such wells, but if'so, they must form an infinitesimal proportion of the hundreds of thousands

of public wells in the Presidency” (Bombay 1930 [Starte Committee] : 52).

31. Ghurye 1957 : 260; Hutton 1961:208. The first untouchable gazetted officer was appointed in 1927 (Zelliot 1969: 162). 32. Zelliot 1969: 166.

British Rule

29

approach was exemplified in the activities of the Arya Samaj; the secular, by Jyotirao Phule. During the 1920s, each acquired a champion of immense stature. Their individual struggles illuminate the problems inherent in these approaches; the wrestling between them cast the mold of policy about untouchability for independent India. In 1916 Mohandas K. Gandhi returned to India from South Africa and assumed leadership of the Indian National Congress. In his autobiography, Gandhi tells us that from the age of twelve he was convinced that untouchability could not be sanctioned by Hinduism. From his first public statements on the subject, he describes untouchability as a sinful excrescence upon Hinduism. He calls upon caste Hindus to do penance, to purify themselves and Hinduism, while urging the untouchables to observe cleaner habits. The problem is one for voluntary private action, not for legal coercion. While rejecting caste in its modern manifestations, Gandhi accepted varnashrama dharma as a healthy division of labor by birth. Untouchability should be removed by abolishing

the unwarranted fifth division and restoring the untouchables to their rightful estate as Sudras in a purified varna expressing not inequality but a horizontal and non-invidious division based on hereditary qualities. * The notion that some are by birth polluting must be abolished, and those self-imposed restrictions on intercourse that are conducive to spiritual development should be retained. If Gandhi personified the evangelical appreach, the secular approach was personified by the most gifted and wide-ranging spokesman to appear among the untouchables, Dr. Bhimrao Ramji Ambedkar.33

Dr. Ambedkar

(born

1890) was younger and less ex-

perienced than Gandhi when he stepped into the political stage in the early 1920s.>4 While Gandhi’s appeals are directed principally to the 33. These giants throw long shadows. Using them to illuminate the contlict over untouchability should not obscure the fact that each had rivals and critics within his own camp. The untouchables were (and are) a vast and heterogeneous collection. Dr. Ambedkar, throughout his career, received the most intense and unswerving support from his own

Mahar

community.

Notwithstanding his subsequent veneration, his

command of the allegiance of other groups of untouchables varied considerably, and there were intermittent attacks on his capacity to represent them. On other untouchable leaders and movements, see Zelliot 1969 : 183-85, 193-94, 206-7; Patankar and

Omvedt 1979 : 415. 34. These two men shared a number of qualities but displayed a number of striking contrasts. Both hailed from the hinterlands of Bombay: Gandhi from Gujarat, Ambedkar from Maharashtra. Gandhi was a bania (merchant) by caste, while Ambedkar

was a Mahar. The Mahars are a populous caste of field laborers and village servants in western India, which had a military tradition, having formed many units in the East

India Company’s Bombay army. Both came from families with a tradition of government service. Gandhi's elders had traditionally served as Diwans of princely states in Kathiawad, Ambedkar’s as non-commissioned officers in the Indian army. Both grew

30

Setting and Policies

caste Hindus, Ambedkar addresses the untouchables and directs their

attention to the potentialities of action to be taken by government. His programs express conflicting impulses between assimilation into Hindu society and separation from it, a pattern that recurs throughout his career. These conflicting impulses are displayed in two successive conferences over which he presided in 1929. At the first, the conferees were invested with the sacred thread by a Brahmin priest to the

chanting of vedic hymns, in a classical instance of elevation in “sanskritic” terms. At the second, the conferees resolved to embrace

“any other religion” if the Hindus continued obdurate in their mistreatment.>5 The next major round of bargaining over the future of India began in 1928. For the first time the untouchables were full-fledged participants. The negotiations began with the visit of the Simon Commission, which was empowered to make recommendations for a new round of constitutional reforms. Dr. Ambedkar, by this time recognized as an important spokesman, appeared before the Commission to demand reserved seats for the untouchables in legislative bodies, special educational concessions, and recruitment to government posts, recommendations substantially accepted in the Commission’s Report.3¢ When this Report proved unacceptable to the major contenders, the British Government convened a Round Table Conference in London to

up in an atmosphere of devotional religion: Vaishnavism in Gandhi's case, while

Ambedkar’s family were adherents of the 15th-century poet and religious reformer Kabir who had Muslim as well as Hindu followers. Both were married young to younger

and less educated wives, and both were sent overseas to complete their education.

Gandhi went to London, though the sea journey meant ostracism from his caste, to join the Bar. Ambedkar, under the patronage of the Gaikwad of Baroda, was sent to

America—the first prominent Indian leader to be educated there. Both returned to

Bombay and joined the Bar. When Ambedkar returned to India, he joined the Baroda civil service in a responsible job under his patron, but lack of acceptance of the untouchable by his fellow officers, clerks, and landlords drove him back to Bombay.

Under the patronage of another Hindu prince, the Maharajah of Kolhapur, he entered litics.

Po ambedkar wasa scholar, who managed to earn several doctorates, and a bibliophile.

Unlike Gandhi, he was an aloof, solitary, and domineering man who found sustained

organizational efforts trying. His impatience with organizational detail and his unwil-

lingness to delegate authority made each of the many organizations he founded a

personal organ, run along paternalistic lines and lacking in solid leadership in the

middle ranks. Unlike Gandhi, whose devotions never ended, Ambedkar was for the

most of his life a mild theist or indifferent to religion. Unlike the mild Gandhi, he was given to outspoken and caustic attacks on his opponents. Gandhi’s Western heroes were

Ruskin, Tolstoi, and Thoreau; Ambedkar's favorite historical figure, according to his

biographer (Keer 1954 : 214), was Napoleon. 35. Keer 1954: 128-30.

36. Reported in Zelliot 1969 : 170-72.

British Rule

31

which it invited delegates from all parties and prominent interest groups. The Congress, claiming that it alone represented Indian opinion, refused to participate. In the course of the Conference, Dr. Ambedkar abandoned his earlier opposition to separate electorates.°7 His demands for safeguards in a future constitution now included separate electorates as well as recruitment to government services, laws against discrimination, and a special department to look after the Depressed Classes. The Civil Disobedience Campaign, which had kept Congress aloof from the first Round Table, ended with an arrangement for a Second Round Table Conference. Gandhi, representing Congress, attacked all the other delegates as unrepresentative. In particular, he attacked Ambedkar’s claim to represent the untouchables. “I claim myselfin my own person to represent the vast mass of the untouchables.” Opposed to separate electorates for any group, he grudgingly conceded them to

Muslims, Christians, Sikhs, and Anglo-Indians, but the demand of the untouchables for separate electorates he termed “the unkindest cut of

all” and vowed to resist it with his life.36 When the Conference’s Franchise Committee was unable to reach agreement on representation for minorities, the delegates authorized

Prime Minister Ramsay MacDonald to make an award which would be

binding on all parties. The Communal Award, announced on August 17, 1932, gave the untouchables regular votes in the general electorate and granted their demand for separate electorates in areas where they were concentrated in addition to the regular votes they would cast as members of the general electorate.39 Gandhi, who was imprisoned as a result of his civil disobedience campaign, now undertook to carry out his vow by fasting until death unless the separate electorates for untouchables were revoked. His objection was that separate electorates would signify a perpetual split in Hinduism, perpetuating the stigma of untouchability and making impossible the assimilation of the untouchables into the Hindu community. It is evident that Congress opposition, if not Gandhi’s personally, was inspired by fear that the great Hindu base of Congress support

would be weakened and that candidates returned from separate electo-

rates would be advocates for group interests with little concern for the Congress program. Finally, many Congressmen, including Gandhi, 37. Dr. Ambedkar’s oppsition to separate communal electorates was shaken by the

stubborn attachment of other communites to this device, which was also favored by

most untouchable activists. His sense of obligation to represent his constituency compounded his fears that the claims of the Depressed Classes would be submerged without their own representatives (Zelliot 1969 : 176-77; Patankar and Omvedt 1979 : 419). 38. Zelliot 1969: 181. 39. Zelliot 1969: 185.

32

Setting and Policies

did not have a very high regard for the political capacities of the untouchables.*° The British Government declined to alter the Award without the consent of the affected groups—i.e., without the consent of Ambedkar as the representative of the Depressed Classes. During the twenty-one days of Gandhi’s fast, Ambedkar was subjected to immense pressures, not least the sense of the cataclysm that might engulf the untouchables if Gandhi were to die. Wresting as much as he could from the Congress leaders who negotiated with him, Ambedkar relinquished the separate electorates in return for a system of reserved seats to be elected from “joint” or general electorates. The final agreement reached on September 28, 1932, known as the Poona (or Yeravda) Pact,

set the subsequent pattern of representation for untouchables.*1 The Poona Pact provided 148 seats for untouchables, instead of the 78 separately elected members given by the Communal Award. It also provided a system of primary elections for those reserved seats; a panel of four candidates was to be chosen by electors from the Depressed Classes. In the light of the agreement, it is clearer just what was really at stake. Gandhi argued for joint electorates so that caste Hindus would be forced to seek votes from untouchables. But, of course, this was the case under the Communal Award, since untouchables also had votes in the

general constituencies. Gandhi’s principal objection had been to having separate rolls of untouchable voters, but this was preserved by the Poona Pact’s arrangements for nomination of the candidates.*? Whatever the motives of the fast, the political effect was to dramatize the issue of untouchability while preventing arrangements that would foster an independent political movement sufficiently strong to belie the Con40. After meeting Ambedkar for the first time in 1931, Gandhi told his secretary, “I

did not know that he was a Harijan. I thought he was some Brahman who took deep interest in Harijans and therefore talked intemperately” (quoted by Zelliot 1969 : 178).

Since Ambedkar had been a prominent spokesman for several years, this suggests inattention to untouchable movements as well as an assumption that untouchables were incapable of providing their own leadership. Cf. Patankar and Omvedt 1979: 419.

41. For Ambedkar’s account, see Ambedkar 1946 : 322-28; Zelliot 1969 : 182-88. For a Gandhian account, see Pyarelal (Nair) 1932: 165~69, Gandhi 1954 b: f: 177, 65-67;

also Dushkin 1957, chap. 4.

42. The Poona Pact provided a system of primary elections in which eligible voters of the Depressed Classes would “form an electoral college” to elect “‘a panel of four candidates belonging to the Depressed Classes” to stand for each reserved seat. This system was to end after ten years unless terminated sooner by mutual agreement (Poona Pact, reprinted at Ambedkar 1946 : 88-89). Subsequent disagreement over this clause

was resolved when the Hammond Committee (on construction of the constituencies)

adopted Ambedkar’s interpretation that “four” meant “no more than four” and that voting in multi-member constituencies was to be cumulative rather than distributive (Ambedkar 1946 : 91-94). In the 1937 elections, primaries were held in only 43 of the 151 reserved constituencies. Ambedkar complained that the primaries were little used

British Rule

33

gress claim to represent all of India.“ Ironically, this fast is often referred to as Gandhi’s fast against untouchability. In a setting of Hindu-Muslim divergence, this sequence of events— the Communal Award, the fast, and the Poona Pact—raised concern

with untouchability to new heights.“ It was recognized as a problem that demanded a solution as part of the question of Indian nationhood.

The new salience of untouchability was not unconnected with concern

about political alignments.*S As one Congress sympathizer put it:

because the double election was prohibitively expensive for Scheduled Caste (as they

had come to be called) candidates. And, analyzing the elections, he argued that when

primaries were held, they did not insure that those successful in the general election were genuine representatives of the Scheduled Castes (Ambedkar 1947 : 44). This

displeasure was shared by other untouchable leaders. Ambedkar had earlier proposed modifying the scheme to require that candidates could stand in the general election only if they had received 25% of the Scheduled Caste vote in the primary, a suggestion rejected by Gandhi (Zelliot 1969 : 188). A proposal to continue the primary system with a 20% requirement may be found in a report to the 1945 Sapru Committee (Sapru 1945 : Lxviii), but after partition and the adoption of Article 325 barring electoral rolls on the basis of caste, the possibilities for such primaries faded from sight as well as from memory. 43. Lelah Dushkin (1957 : 74) observes that “the basic solution (except for the panel

system) was what Ambedkar had advocated in 1929, what Sapru had advised at the second R[ound] T[able] C[onference], and what Rajah and Moonje had proposed in their pact. The whole affair would not have occurred if Gandhi had not treated the D{epressed] C[lasses] delegates with such scant courtesy at the second R.T.C.” 44. A conference of Hindu leaders, convened in Bombay on September 25, 1932, to

ratify the Poona Pact, unanimously adopted the following resolution : “This Conference resolves henceforth, amongst Hindus no one shall be regarded as an untouchable by reason of his birth, and that those who have been so regarded hitherto will have the

same right as other Hindus in regard to the use of public wells, public schools, public roads and all other public institutions. This right shall have statutory recognition at the first opportunity and shall be one of the earliest Acts of the Swaraj Parliament, if it has not received such recognition before that time. It is further agreed that it shall be the duty of all Hindu leaders to secure, by every legitimate and peaceful means, an early removal of all social disabilities now imposed by custom upon the so-called untouchable classes, including the bar in respect of admission to temples” (quoted in Rajagopalachari 1933 : 1). 45. Anxiety about the departure of the untouchables should be seen against a background of the much-discussed, long-term decline in the Hindu proportion of the total population of India, roughly summarized in the following table : Date Percentage Source 1881 1891 1901 1911 1921

1931 1941

74.3 72.3 70.3 69.3 68.4

68.2 65.9

a a b b b

b b

Sources : a : Census of India, 1911; b : Census of India, 1941: Vol. I, Part I, 102 ff.

34

Setting and Policies With the Harijans within Hindus is safe. Without the uncertain. . . .We do want struggle, and even if it were

the Hindu political group, the future of the Harijans, the destiny of the Hindus will remain to have the Harijans on our side in the political for this purpose alone, the Caste-Hindus must

raise this curse of untouchability and bridge the gulf that divides them from

their co-religionists. There is a real crisis in the political life of the Hindu community. . . .47

Efforts to remove untouchability and help its victims proliferated in the wake of the Poona Pact.** Congress had earlier committed itself toa program of fundamental rights in the future republican India in which there would be no recognition of civil and social disabilities.*? But only after Gandhi’s fast were Congress leaders willing to countenance the affirmative use of the law to abolish disabilities—and in particular to obtain by law the admission of untouchables to Hindu temples, thereby symbolizing their inclusion within the Hindu community. While temple entry absorbed advanced Hindu opinion, Dr. Ambedkar turned away from it to emphasize economic and educational advancement. Temple entry, he asserted, was up to the Hindus; it was through political power that the untouchables were to find their salvation, not through acceptance by the Hindus. Both Gandhi and Ambedkar,

Hinduism

of the day.

each in his own way, attacked the

In a charismatic

transvaluation

of Hindu

46. One enduring product of these events was a new terminology. “Untouchable”

was in general use by 1920, but “Depressed Classes” remained the official term. In 1931, Dr. Ambedkar and R. Srinivasan, (the Depressed Classes’ representatives at the

Round Table Conference) observed that the term “Depressed Classes” is “degrading and contemptuous” and proposed as alternatives “‘Non-caste Hindus,” “Protestant

Hindus,” or “‘Non-conformist Hindus” (Indian Round Table Conference 1932 : 76).

But events took a different course. Shortly afterwards, Gandhi began to popularize

“Harijan”

(Gandhi

1954 : 16-17), and then the Government of India Act,

1935,

denoted those who would enjoy the special electoral arrangements for untouchables as “Scheduled Castes,” a term destined to remain the official designation for untouchables

down to the present. Just as “‘Harijan” fossilizes the evengelical fervor of Gandhi's anti-untouchability campaigns, so “Scheduled Castes” summons up the bureaucratic rigidification and torpor that overtook Ambedkar’s measures for secular improvement. 47. Agarwal 1934 : 22. Mr. Agarwal was a Cambridge-educated lawyer and professor in Poona. 48. These included the formation of the Harijan Sevak Sangh, an organization devoted to the uplift of the Harijans. Its board was composed entirely of non-Harijans, giving rise to reproaches from Dr. Ambedkar and others. 49. In 1931, the annual meeting of the Indian National Congress at Karachi propounded a program of fundamental rights for future republican India which included : “(vi) no disability to attach to any citizen by reason of his or her . . .caste .. .in regard to

public employment, office of power or honors, and in the exercise of any trade or calling. (vii)equal rights of all citizens in regard to public roads, wells, schools and other places of public resort” (All-India Congress Committee, n. d., p. 66).

British Rule practice, Gandhi

dramatized

35

his identification with Harijans, and

particularly with the bhangis (sweepers), the lowest and most polluting of castes in traditional terms.S° During his tours, he stayed in the sweepers’ quarters. Calling himselfa Harijan and a Bhangi, he expressed the wish to be reborn as an untouchable. Gandhi advocated a purified vamashrama dharma in which untouchables would be restored to their rightful place as Sudras.51 But to Ambedkar, the four-vamna system is the very root of untouchability. For Hinduism to accommo-

date the untouchables, it must be purged of the vama system. Where Gandhi proposed to rid Hinduism of its imperfections and restore it to its ideals, Ambedkar was skeptical of these ideals themselves. To him,

those who observed untouchability were not bad Hindus, but merely observant ones. The only way to change their conduct was to destroy the authority of the sastras. Thus, reform of Hinduism meant extirpating the offending doctrines and abolishing the hereditary priesthood; in their place a new standard book of doctrine should be promulgated and a licensed, certified priesthood established.5?

Dr. Ambedkar increasingly despaired of the possibilities of reform within Hinduism. He alternated proposals for radical reform of Hinduism with recommendations that untouchables convert to a religion that would guarantee them equality. In 1936 he conducted an extended flirtation with Sikhism.5? Proposals for conversion aroused a storm of opposition. Gandhi charged that the religious feelings of the untouchables were being exploited in a political game. In a characteristic combination of genuine solicitude and uncomprehending condescension, he expressed doubt whether the untouchables could 50. “A Bhangi does for society what a mother does for her baby. A mother washes her

baby of the dirt and insures his health. Even so the Bhangi protects and safeguards the health of the entire community by maintaining sanitation for it. The Brahman’s duty is

to look aiter the sanitation of the soul, the Bhangi’s that of the body of society” (Gandhi

1954b : 215~18 [from Harijan of November 28, 1936]). 51. Gandhi’s program was by no means moderate in its day. C.B. Agarwal is probably closer to popular feeling among the educated when he complains that Gandhi is moving too fast. While favoring civil rights and educational help, Agarwal rejects interdining and inter-marriage and would restrict temple entry to those untouchabales who give up meat-eating.

52. This was a program so calculated to outrage respectable Hindu sentiment that the

organization cancelled their invitation when they saw advance copies of the speech (see Ambedkar 1945). But it was not entirely without precedent; e.g., in 1928, a rural Classes Conference had called for proscription of the Manu Smrti, the 1 Hindu law text (Zelliot 1969: 170).

53. One interestung sidelight of these negotiations is that Ambedkar called in the

Hindu

Mahasabha,

the foremost of the Hindu communalist

parties, to seek their

approval. Apparently the aim of this was to remain Hindu for purposes of the reserved seats awarded by the Poona Pact. The Hindu Mahasabha agreed on the condition that conversion was to Sikhism rather than to Christianity or Islam. Ambedkar himself

36

Setting and Policies

distinguish tne relative merits of the contending religions any more than a cow could.*

The new popular governments which took office in 1937 avowed their

opposition to untouchability. Madras passed the first legislation imposing criminal penalties for denying to untouchables the use of public facilities. Temple entry had been proclaimed earlier in a few progressive princely states. In 1938, for the first time in British India, governments

intervened to secure the opening of temples when Bombay and Madras passed temple-entry acts. Encouragement of education for untouchables increased, and there were measures to secure them some posts in government service. The non-Brahmin movements that had emerged so dramatically a generation earlier now bifurcated, as lower castes insisted on provisions separate from those for the upper non-Brahmin groups.°5 (The former emerge as the “other backward classes,” whose story we take up in chapter 6.) To a large extent, these movements were absorbed into the Congress. Congress governments continued and extended the regime of communal quotas in public services and college admissions that these movements had produced, as did non-Congress governments elsewhere.5® In the central services, the Government of India reserved

quotas for Muslims and other religious minorities. During the Second World War, when the Congress opposed coopera-

tion with the British war effort, Dr. Ambedkar served as a member of

the Viceroy’s executive council, the first untouchable to attain such a lofty station. He secured for the Scheduled Castes (as they were now called) reservations of posts in central government services and various other concessions, including recruitment into the army and the

police.5”

Gandhi, who was in prison along with the Congress leadership for

agreed that it would be preferable to stay within Hindu culture, thereby avoiding Muslim domination. As it happened, after the conversion of a few of his followers, bad

feeling developed between Ambedkar and the Sikhs, and the matter dropped out of sight (Zelliot 1969 : 224-26; Keer 1962 : 288). But the issue of the untouchables enjoying

special treatment after abandoning Hinduism is a persistent one that resurfaced dramatically in the 1950s and has remained unresolved. See chap. 9, §§E and F, below. 54. Gandhi's response to charges that this statement was demeaning was that neither could his wife (Gandhi 1954b : 154 [from Harijan of March 13, 1937]). 55. The Madras Provincial Backward Classes League, consisting of the less forward non-Brahmin communities, was founded in 1934 for the purpose of securing separate treatment from “‘the forward non-Brahmin communities.” See chap. 6 below. On the tension between the upper and lower strata within the non-Brahmin movement, see

Saraswathi 1974, chap. 4, Ramaswamy 1978 : 297 (Andhra districts).

56. On communal quotas in public services and educational institutions under the Unionist government in the Punjab, see Tinker 1962 : 29. 57. On wartime developments, see Zelliot 1969 : 262-65.

British Rule

37

obstructing the war effort, had undergone a radical change in his thinking about untouchability. He was now willing to countenance legal compulsion to permit untouchables to assert their rights, and envisioned a future constitution in which all forms of untouchability

would be abolished by law. He no longer espoused personal exclusiveness, but opposed all caste distinctions. He no longer defended vama while denouncing caste. Untouchability could be eradicated only if Hindu society became casteless. To this end he now encouraged interdining and intermarriage, which he hoped would lead to a situation in which “finally there will be only one caste, known by the beautiful name of Bhangi, that is to say, the reformer or remover of all dirt.”’5®

When Independence came in 1947, caste was in bad odor. It was widely viewed as an impediment to individualism and to broad national loyalties and thus inimical to progress and democracy. The hardships inflicted on the lowest castes inspired humanitarian revulsion. It was widely accepted that caste would have no place in independent India and that efforts to ameliorate the effects of past inequalities were in order. As power passed into Indian hands, the exclusion of untouchables from public facilities and from Hindu temples were made statutory offenses throughout most of India. Reservations for untouchables were established in the central services, and a program of educational assistance was begun. Ambedkar was one of the principal architects of the Constitution that emerged from almost three years of deliberations.5? It is a federal constitution: authority is explicitly apportioned between the center and constituent territorial units. There is a greater subordination of the units to the center than in the United States, but the states have

considerable autonomy and are more than mere administrative divisions. It is a written constitution—with the distinction of being the longest in the world. It had to be written in order to be federal—in order to arrange the division of powers between the center and the states. But also—and this accounts for its length— because there were a number of other matters which the constitution-makers wished to put

beyond the reach of temporary or narrow majorities. These include . safeguards for various minorities and a large part of the fundamental 58. On development of Gandhi’s thinking on caste, see Gandhi 1954b : 72-85; Dalton 1967: 171-74. 59. Ambedkar’s hopes of leading an independent political party in which untouch-

ables played a central role foundered in the 1945 provincial elections as they had in the

1937 elections. Through

the good offices of the Muslim

League,

he became a

member of the Constituent Assembly, which became the governing body of India when the country was divided and given independence on August 15, 1947. In a gesture of reconciliation, Prime Minister Nehru appointed him Law Minister, and he was chosen as chairman of the Drafting Committee of the Constituent Assembly.

38

Setting and Policies

structure of the government apparatus. Thus, detailed provisions regarding the judiciary, the civil service, and the proceedings of legislatures, adopted without fundamental change from the British administration, are all entrenched in the Constitution. There is a written and

enforceable set of Fundamental Rights— modelled in large measure after the United States Bill of Rights. There is a set of Directive Principles of State Policy, which specify the general goals of governmental policy — the right to work, higher living standards, free and compulsory education,

etc.—and

anticipate a welfare state which was

described by one skeptical observer as ‘‘Fabian socialism without the socialism.” Finally, judicial review is explicitly provided: the Constitution is to be interpreted and its provisions enforced by a unitary national judiciary. The chapter on Fundamental Rights contains a number of familiar features: prohibitions of governmental restriction of the freedoms of religion, speech, press, association, and movement; minimal standards of criminal justice; protections for property—all analogous to their American counterparts. The draftsmen, cognizant of the difficulties in interpretation and in application encountered in America and Aus-

tralia, specified exceptions and limitations to each right. The listing of exceptions and the deliberate omission ofa “due process” clause reflect an attempt to reduce the scope for judicial discretion in determining the extent of these rights. In general, the draftsmen wished to confine

judicial review to the application of express principles of law and tried to avoid giving the courts a role of continuous constitution-making. The Fundamental Rights also contain bans on discrimination on grounds of religion, caste, sex, language, etc. The attempt to secure

equality goes beyond guarantees of equal treatment by government:

crucial provisions regulate not only “state action” but also the behavior

of private persons and groups. Caste discrimination by government or by private individuals is banned. But it was recognized that evenhanded treatment would not suffice to overcome the cumulative disadvantages of those at the bottom of the social hierarchy. To promote the advancement

of the untouchables,

tribals, and other

‘‘socially and

educationally backward classes,”’ the Constitution allows protective or

compensatory discrimination in their behalf. The notion of communal quotas and safeguards which had become an accepted and prominent principle of administration during the last forty years of British rule was

decisively rejected as a general principle of governance. There are no

safeguards for religious minorities apart from the guaranteed fundamental rights. Only for the Scheduled Castes and Scheduled Tribes (and to a lesser extent, for other backward classes) are these devices 60. Jennings 1953 : 31.

British Rule 39 countenanced. There is a deliberate departure from formal norms of equality in order to offset the historic inequalities of these groups. Ambedkar’s role in the Constituent Assembly underlines the decisive rejection of Gandhian notions of village self-rule. While Gandhi idealized village India and its handicraft technology, Ambedkar despised Indian villages as “dens of ignorance, narrow-mindedness and communalism’”’*®! and ardently supported machine technology which would provide leisure, cultural advancement, and finally equality. To the same end, he rejected Gandhi's ideal of trusteeship by the rich in favor ofa kind of state socialism which would promote rapid industrialization. And in spite of his suspicion that India might require benevolent autocracy, he was a supporter of centralized parliamentary government rather than of village autonomy. In all of this Ambedkar stood closer to the left wing of the Congress than either stood to Gandhi. The Constitution incorporated much of the secular program he had championed, but Dr. Ambedkar was dissatisfied with the performance of the Congress government. After a bitter political disappointment in the

1952

election,

he disowned

his role as designer of the Con-

stitution.©? Throughout his career, Dr. Ambedkar had written on the historical aspects of the caste system and displayed a strong interest in Buddhism, which he considered a religion of equality and social reform. This was reinforced by his belief that the untouchables were former Buddhists who had been degraded by the Hindus for refusing to give up Buddhism and for continuing to eat beef. Ambedkar’s interest in Buddhism intensified after Independence, and he dedicated himself to the revival of Buddhism in India. After his political disillusionment was complete, he returned to the notion that the problems of untouchability could be solved only by the abandonment of Hinduism. In 1955, a year before his death, he launched a mass movement

for conversion to

Buddhism that inspired millions of untouchables to abandon Hinduism

61. VICAD 39. 62. Dr. Ambedkar’s bitter disappointment is understandable, but in retrospect and from afar, his sense that the constitutional protections which he designed or at least concurred in were a failure seems less than fully warranted. The untouchables, unlike the Muslims, were a scattered minority with no territorial base, no unity among themselves, no rallying point in religion or historical glories, and with little capacity and fewer resources for organization. If ultimately they had little choice but to accept

what they were given by the majority, their leadership did well by them in obtaining

safeguards and preferences out of the independence struggle. Although their constitutional position might have been even stronger, the major shortfall has been not in constitutional protection but in organization and imagination to utilize effectively the protections and opportunities that are available. 63. Ambedkar

1948 : 76, 80-83. The idea that Untouchables or low castes were

former Buddhists had gained popularity among a scattering of low caste groups (Zelliot 1969 : 231 and n. 83). It had been expounded earlier by Colone! Olcott of the Theosophical Society and others.

40

Setting and Policies

and embrace Buddhism. A generation later, the Buddhists were a

significant element of both political militancy and intellectual leadership; and Dr. Ambedkar was the venerated “patron saint” of the untouchables.© The wrestling of Gandhi and Ambedkar displays an ironic reversal. Gandhi, for decades the embodiment of the evangelical approach of repentance and personal growth, acknowledged the need for external controls and eventually adopted all the planks of the secular reform platform championed by Ambedkar. Dr. Ambedkar, on the other hand, lost faith in the efficacy of secular political reform without a cultural transformation with its roots in new (or restored) religious identity. What emerged was a national commitment to reduce invidious distinctions among groups, coupled with a conviction that it was permissible and perhaps necessary to employ these discredited caste notions to

effectuate

chapters.

the

We

equalizing

policies

that we

survey

in the following

shall then return to explore the evolution of these

categories and to trace their career as instruments of the equalizing

policies that we call compensatory discrimination.

64. The number of Buddhists reported by the census increased from 180, 823 in

1951 (of whom only 2,500 were in Maharashtra) to 3,250,327 in 1961. Of these, 2,789,501 were in Maharashtra (Census of India 1961, Paper No. | of 1963). Zelliot (1969: 239) estimates that 80% of the Mahar caste reported themselves as Buddhists in 1961. The 1971 census numbered 3,874,942 Buddhists. These figures represent a consideable understatement: many who regard themselves as Buddhists would not

publicly identify themselves as such for fear of losing educational and other benefits conferred on Scheduled Castes. See chap. 9, §E, below. 65. Although respected as a “founding father,” Ambedkar is the prime symbol of

self-assertion by low castes. The renaming of Marathwada University in his honor in

1978 touched off widespread rioting throughout that region of Maharashtra. Resentment of reservations and concessions for untouchables was given murderous expression; Mahars were especial targets of attack (Morkhandikar 1978; Abraham 1978).

3

Compensatory Discrimination Programs and Their Implementation

INDIA’s POLICY OF compensatory discrimination comprises a wide array of preferential schemes. In this chapter we will survey some of the most prominent programs, consider some general features of their administration, and explore the question of how they should be assessed. These programs are authorized by constitutional provisions that permit departure from formal equality for the purpose of favoring specified groups. We shall later examine closely the meaning of these provisions and the tensions between them and other constitutional commitments. For the present it may be helpful to identify them as the constitutional anchorage of the various policies described here. Article 46, a “Directive Principle of State Policy,” declares:

The State shall promote with special care the educational and economic

interests of the weaker sections of the people, and, in particular, or the Scheduled Castes and the Scheduled Tribes, and shall protect them from social injustice and all forms of exploitation.?

The Constitution’s ban on discrimination in government employment is qualified by Article 16(4), which permits the State to make any provision tor the reservation of appointments or posts in favour of any backward

class of citizens which, in the opinion of the State, is not ade-

quately represented in the services under the State.”

1, The Directive Principles of State Policy, contained in Part IV of the Constitution,

are not themselves justiciable, but the Constitution prescribes it as “the duty of the State to apply these principles in making laws” (Art. 37.) The jurisprudential status of the Directive Principles is discussed below in chap. 11. 2. In addition, Article 335 provides: “The claims of the members of the Scheduled

Castes and the Scheduled Tribes shall be taken into consideration, consistently with the maintenance of efficiency of administration, in the making of appointments to services

and posts in connection with the affairs of the Union or ofa State.” The relation of these provisions is discussed in chap. 11, §B, below.

42

Setting and Policies

The general provisions banning discrimination by government (Art. 15) and banning discrimination in government-aided educational

institutions (Art. 29 [2] ) are similarly qualified by Article 15(4), which provides:

Nothing in Article 15 or Article 29(2) .. .shall prevent the State from making any special provision for the advancement of any socially and educationally backward classes of citizens or for the Scheduled Castes and

the Scheduled Tribes.

There are specific provisions that we shall encounter later. But these, along with specific provisions for reservation of seats in legislative bodies (which will be discussed shortly), are the core of the constitutional commitment to compensatory discrimination. The benefits of “compensatory discrimination” are extended to a wide array of groups. There are three major classes. First, there are those castes designated as Scheduled Castes (SC) on the basis of their “untouchability.” They numbered nearly 80 million (14.6% of the population) according to the 1971 census.* Second, there are the Scheduled Tribes (ST), who are distinguished by their tribal culture and physical isolation and many of whom are residents of specially protected Scheduled Areas. They numbered more than 38 million (6.9% of the population) in 1971.5 Third, there are the “Backward Classes” (or, as they are sometimes called, “Other Backward Classes,”)®a heterogeneous category, varying greatly from state to state, composed for the most part of castes (and some non-Hindu communities) low in

the traditional social hierarchy, but not as low as the SC, Also included among the Other Backward Classes (OBC) are a few tribal and noma-

3. Temporary preferences for yet another group, the Anglo-Indians, were not based on their backwardness but were designed to cushion the loss of the privileged position they had formerly enjoyed. Reserved posts in certain services and special grants to Anglo-Indian educational institutions were guaranteed by Arts. 336(1) and 337. These

preferences were on a diminishing scale and ceased to be operative on January 26, 1960. Reservation for Anglo-Indians in the services has ceased. The states have, with some reductions, continued to make grants to Anglo-Indian schools. Since the constitutional grants were not merely authorized but guaranteed, there was no occasion for this group to be treated under the general provisions for preferences, Arts. 15(4) and 16(4). On the limits of state power to place conditions on the recipients of these grants, see Bombay Education Society v. State of Bombay, A.1.R. 1954 S.C. 561; In re Kerala Education Bill, 1957, A.1.R. 1958 S.C. 956. The Constitution not only provides a formula for determining the amount of grants and reservations, but also defines membership in the Anglo-Indian community (Art. 366[2]). On the status of Anglo-Indians and the background and effects of the Bombay Education Society case, see Grimshaw 1959. 4. On the selection of these groups, see chap. 5, §A, below. 5. On these groups, see chap. 5 §C, below. 6. On the selection of these groups, see chaps. 6, 7, 8, below.

Programs and Their Implementation

43

dic groups, as well as converts to non-Hindu religions from the SC, and in some areas the Denotified Tribes.” It has been estimated that there were approximately 60 million persons under the Other Backward Classes heading in 19618— roughly the magnitude of the SC population at that time (64 millions). (Today the portion of the population designated under this heading is probably larger.) For the most part, preferences have been extended on a communal basis. Members of specified communities are the beneficiaries ofa given

scheme, and all members of the community, however prosperous, are

entitled to the benefits. However, some schemes use a means test to

supplement the communal one—only members of the listed communities with incomes below the specified ceiling are eligible. In a few instances, the communal test has been replaced by an economic one— income or occupation or a combination of the two—and a few schemes use tests neither communal nor economic.? Preferences are of three basic types: First, there are reservations, which alfot or facilitate access to valued positions or resources. The most important instances of this type are reserved seats in legislatures, reservation of posts in government service, and reservation of places in academic institutions (especially the coveted higher technical and professional colleges). To a lesser extent, the reservation device is also used

in the distribution of land allotments, housing, and other scarce re-

sources. Second, there are programs involving expenditure or provision of services—e.g., scholarships, grants, loans, land allotments, health

care, legal aid—to a beneficiary group beyond comparable expenditures for others. Third, there are special protections. These distributive schemes are accompanied by efforts to protect the backward classes from being exploited and victimized. Forced labor is prohibited by the Constitution,?° and in recent years there have been strenuous efforts to release the victims of debt bondage, who are mostly Scheduled Castes and Tribes. Legislation regulating money lending, providing debt re7. “The Denotified Trbes, or Vimukta Jatis, are the former Criminal Tribes. They

became ‘ex-criminal’ when the Criminal Tribes Act of 1924 was repealed in 1952. Their population is estimated at upwards of 40 lakhs [4 million], of whom more than half live

in (Uttar Pradesh.] Although U.P. lists several of them as Scheduled Castes, most

“States count them all as Other Backward Classes. They pose special psychological, social and economic problems, however, and have come to be treated as a separate

fourth category for administrative and budget purposes” (Dushkin 1961 : 1665).

8. Dushkin 1961 : 1665. 9. Economic tests are discussed in chap. 8, §G, below. For an example of a “residence” criterion, see the Home Ministry’s grants for all persons in selected

“undeveloped areas” geographic criteria.

10. Art. 23(1).

(RCSCST

1959-60:1,

323).

See

chap.

8. §H,

below,

on

44

Setting and Policies

lief, and restricting land transfers attempt to protect SC and ST trom the economic oppression of their more sophisticated neighbors. Antiuntouchability propaganda and the Protection of Civil Rights Act

attempt to relieve untouchables from the social disabilities under which

they | have suffered.1! This legislation is not “compensatory discrimina-

tion” in the formal sense of departing from equal treatment to favor these groups; it enjoins equal treatment rather than confers preferential

treatment. But in substance it is a special undertaking to remedy the disadvantaged position of the untouchables. Unlike the measures discussed above, these are regulative rather than distributive; they involve different claimants in different settings; they are implemented by different sets of officials—e.g., by police and magistrates rather than by

administrators. There are important connections and perhaps important parallels. But the “protections” involve a whole world of different problems and richly deserve detailed study in their own right. We shall take up these protections here only as they intersect with compensatory programs of the first two types. To make the present survey manageable we shall concentrate on several of the most prominent programs—reserved seats in legislatures

and preferential treatment in education (including reservations in professional colleges). Other programs and some general features of their administration will be noted briefly, and the problem of assessing their performance will be broached. The following chapter presents a more fully developed “case study” of preferential treatment in the area of government employment. A. LEGISLATIVE RESERVATIONS The most prominent of all preferential policies is the reservation of seats

in elective legislative bodies. The Constitution specifically provides reserved seats in proportion to their numbers for the Scheduled Castes

(SC) and the Scheduled Tribes (ST) in the Lok Sabha (lower house of Parliament)!? and the Vidhan Sabhas (lower houses of the state

legislatures).!3 No seats are reserved in the upper houses, central or

state. There are no reservations in legislatures for the Other Backward

Classes or for any minority groups.'*

11. The Protection of Civil Rights Act is the new name given to the Untouchability Offences Act (Act 22 of 1955) upon its amendment by the Untouchability (Offences) Amendment and Miscellaneous Provisions Act (Act 106 of 1976). For a review of the background of this legislation and its operation into the mid-1960s, see Galanter

(1972a). 12. Article 330. 13. Article 332. 14. The Constituent Assembly definitively rejected political safeguards for religious and other minorities. See Austin 1966 : 146-56. The Constitution does, however.

Programs and Their Implementation

45

All of the other constitutional provisions for preferences are merely authorizations empowering the State to make special provision for disadvantaged groups. But legislative reservations are specifically provided in the Constitution itself.15 Seats are reserved in proportion to the population of SC and ST to the total population of each state. Thus in 1976, 78 (14.4%) of the 542 seats in the Lok Sabha were reserved for SC, and 38 seats (7%) for ST. Of the 3,997 seats in the Vidhan Sabhas, 540 (13.5%) were reserved for SC, and 282 (7%) for ST.*©

These reservations do not involve “separate electorates” —i.e., the

representation of a particular group by legislators chosen by an electo-

rate composed solely of members of that group.?” The seats are “re. served” in the sense that candidates who stand for them must belong to the specified groups, but the entire electorate participates in choosing among candidates so qualified. Separate electorates for Parliament and the state legislatures are specifically outlawed by Article 325, which provides that no person shall be excluded from any electoral roll on grounds of religion, race, or caste.1®

The constitutional provision of reserved seats is complemented by statutory provisions to enhance political participation by SC and ST. Smaller election deposits are required from members of these groups, and candidates for reserved seats enjoy more permissive residence

requirements.!9

provide for nominated representatives of the Anglo-Indian community in tne Lok Sabha (Art. 331) and in the Legislative Assemblies (Art. 333). On the implementation of these provisions, see RCSCST 1975-77: I, 37.

15. The Constitution contains no similar provision for representation in other politi-

cal bodies. However, such political safeguards may lie within the scope of Art. 15(4),

which is apparently broad enough to authorize the reservation of seats in elective bodies ona local level. 16. Election Commission 1976a: 3-6. 17. By separate electorates is meant the representation of religious (or other) minorities by legislators chosen by an electorate composed only of members of that minority. The provision and extent of such representation was an extremely troublesome political issue in India during the forty years preceding Independence. The Government of India Act, 1909, gave separate electorates to Muslims; the Government of India Acts of 1919 and 1935 provided separate electorates for Muslims, Sikhs, Indian Christians, and other groups. Proposals to give separate electorates to the Scheduled Castes under the 1935 Act were withdrawn after adamant resistance by Mr. Gandhi. See chap. 2 above. 18. The Supreme Court has indicated that Article 15 prohibits communal electorates in local bodies. Nain Sukh Dasv. State of U.P., \.1.R. 1953 S.C. 384. 19. Representation of the People Act, 1951 (Act 43 of 1951). The smaller deposits apply in contesting non-reserved as well as reserved seats. A proposal by Dr. Punjabrao Deshmukh, leading spokesman for the Other Backward Classes, to extend the smaller deposits concession to that category elicited a gruff dismissal from Dr. Ambedkar, then

the Law Minister: “this is the first time that I hear that such a concession should be

extended to the Backward Classes. Hitherto the concessions that have been spoken of as

46

Setting and Policies

Unlike the general authorization of special treatment and the provision for reserved posts in government service, the reserved seats in legislatures are subject to a constitutional time limit. It was originally provided that such reservations should expire ten years after the commencement of the Constitution.”° In 1959 these provisions were extended for another ten years, ?1 in 1969 for another ten,?? and in 1980

for yet another ten.

The demarcation of constituencies and designation of those reserved for SC and ST are entrusted to a Delimitation Commission. The courts are constitutionally barred from intervening in the process of delimitation or allotment of seats.25 Generally, constituencies are formed on the basis of contiguity, communications, absence of natural

barriers, and cultural and ethnic homogeneity. After they are formed, certain constituencies are selected as reserved.6 In the case of the being necessary for the upliftment of the Backward Classes are educational concessions

and concessions in the services of the country. . . .{S]o far as the Members representing

the Backward Classes in this House . . they are capable of paying not only their own election deposits, but also the election deposits of many others.” The government refused to accept the amendment and it was withdrawn. Parliamentary Debates, Vol. XII-XIII (Part II), Cols. 9141-49 (19 May 1951). 20. Article 334. 21. Constitution (Eighth Amendment) Act, 1959.

22. Constitution (Twenty-third Amendment) Act, 1969.

23. Constitution (Forty-fifth Amendment) Act, 1960.

24. The Delimitation Commission consists ofa retired Supreme Court judge, a sitting

High Court judge, and the Chief Election Commissioner (Delimitation Act, 1972 [Act

76 of 1972]). In each state, the Commission is assisted by ten non-voting associate members (5 Members of Parliament and 5 Members of the Legislative Assembly)

nominated by their respective Speakers (§5). The associated members from each state are consulted after tentative delimitation. After changes at this stage, the proposed delimitations are published and objections invited. Where there are objections, hearings are held. The location of reserved seats is the most frequent source of objections, but they are rarely shifted. 25. Article 329. 26. During the 1950s, most of the reserved seats were in double-member constituencies, in which one of the seats could be filled only by a member of the “privileged” group and the other was open to general competition. Thus in 1952, all of the Scheduled Caste seats and half of the Scheduled Tribe seats in the Lok Sabha were double-member constituencies. In 1957, 467 out of 470 SC seats in the Vidhan Sabhas and 115 of the 221

seats reserved for ST were also in double-member constituencies. (On the functioning of

the double-member constituencies and their abolition, see Dushkin 1972 : 189 ff.) Double-member constituencies were abolished in 1961 (the Two Member Consti-

tuency [Abolition] Act, 1961 [Act 1 of 1961]). This aboliton seems to have been inspired

principally by the concern of politicians with the greater difficulty and expense of campaigning in a double-sized district. This was compounded by some resentment of the infrequent but well-publicized instances in which the unreserved “general” seat as well as the reserved seat was won by SC or ST candidates, who were able to compete for

both seats, while their opponents could compete for only the unreserved seat. This did not happen very often. In nearly two hundred double-member constituency contests for Lok Sabha seats in 1952 and

1957, the general seat was won by a reserved-seat

Programs and Their Implementation Scheduled

Tribes

seats

there is a single standard

47

for selection—

concentration of the ST population.?? Since typically the tribal popu-

lation lives in areas of concentration, the constituencies reserved for

them contain a high proportion of ST.”* In 1962, the ST constituted more than 50% of the population in 18 of the 31 Lok Sabha constituencies reserved for them.?9 Conversely, most of the major concentrations of ST population are located in reserved constituencies. Of the 29 constituencies in which ST make up over 40% of the population, 23 are reserved. Thus the reserved seats encompass the bulk of the ST population— perhaps close to 70%. These constituencies are, not surprisingly, more isolated and much less urban than are the general constituencies. In the case of seats reserved for Scheduled Castes, the Delimitation

Commission is instructed to apply two standards— concentration of SC population and dispersal of reservations. The Delimitation Act instructs the Commission to locate seats reserved for SC “in different parts of the State and . . .as faras practicable, in those areas where the proportion of their population to the total is comparatively large.”3° The Delimitation Commission achieves this by first reserving the constituencies with the highest concentration of SC—usually not much more than 20% — and then selecting the others by the dispersal criterion. Since the SC population, unlike the ST, is rather evenly dispersed, the constituencies reserved for them contain far fewer SC members. Among the 76 Lok Sabha constituencies reserved for SC in the 1962 elections, there was

candidate in one in 1952 and in five in 1957. It was apparently even less frequent in the legislative elections, as indicated by the very small number of non-reserved seats occupied by SC and ST. A well-known 1959 Supreme Court decision upheld the right of an ST candidate to take the unreserved seat, sealing the defeat of the Labour Minister (V.V. Giri v. D. Sura Dora, A.1.R. 1959 S.C. 1318, discussed below in chap. 13,§ B). This

decision triggered the move for abolition (Dushkin 1972 : 192). New single-member constituencies were designated for the 1962 elections. In most cases this was done simply by dividing the double-member constituency and allotting the reserved seat to the one with the greatest proportion of SC or ST in its population (Election Commission 1966 : 1. 8). 27. §9(1) (d). 28. In the following pages, comparisons of the reserved-seat holders in the Lok Sabha with the occupants of general seats are based on data made available to me by Professor Henry C. Hart of the University of Wisconsin. These data consisted of (1) coded cards on all 497 members of the Third Lok Sabha from 1962 to 1965 and (2) coded cards based on his interviews with 189 members of the Lok Sabha in 1963-64. Of the sample, 108

were a random sample (30%) of Congress members. The remaining 79 were 60% of the opposition and independent members. The sample of the latter is not random, but is weighted slightly toward those active in the affairs of the Lok Sabha. In the sequel a source given simply as “Hart Data” refers to the larger set of cards on all members; the smaller set is referred to as the “Interview Sample.” The Hart data is analyzed in more detail in Galanter 1979. Other data from the Hart study are presented in Hart 1971. 29. Krishna (1966 : 28) reports 19. 30. § 9 (1) (c).

48

Setting and Policies

none in which they were a majority of the population and only 13 in which they were as much as 30%. The bulk of SC constituencies contain between 10% and 30% of Scheduled Castes. Thus, the vast majority of SC— perhaps 75% —live outside the constituencies reserved for SC. Of the 111 Lok Sabha constituencies where SC constituted more than 20% of the population in 1961, only 46 were reserved, whereas 29 constituencies with less than 20% were reserved. On the whole, consti-

tuencies reserved for SC tend to be political backwaters—slightly less urban, with less newspaper circulation and a slightly greater percentage of agricultural laborers. The same pattern of concentration of Scheduled Tribes and dispersion of Scheduled Castes reappears in the reserved seats in the state

legislatures. In Andhra Pradesh, for example, SC make up only 21% of

the population of the constituencies reserved for them (and 13% of the

population of the general constituencies). But ST make up 43% of the

population of the constituencies reserved for ST and only 2% of the general constituencies.>? Designation of a constituency as reserved usually encounters opposition.

While everyone expressed anxiety that the Scheduled Castes and the Scheduled Tribes should get their full quota of reserved seats, very rarely

indeed would anybody (unless he himself belonged to any of these castes or tribes) agree to any seat being reserved for them in a constituency in which

he was personally interested.34

The selection of reserved constituencies gives rise to the bulk of the objections to the proposed delimitation. As the Election Commissioner

reports,

generally speaking the vocal sections of the public objected to their own areas becoming part of a reserved constituency and put forward ingenious

suggestions for reshaping two or more constituencies so that some particular

constituency would be a general constituency.>4

However, these objections rarely bring about a change. The courts will not provide any relief so long as the Delimitation Commission has taken into account the factors specified by the Delimitation Act.*5 31. My figures are taken from the Hart data. Krishna (1966: 28) gives slightly different figures, but the overall picture is not very different. 32. Election Commission 1976b: II-A, 265. Inspection of the Election Commission tables reveals that Andhra is not atypical in this regard. 33. Election Commission 1959: I, 57. For an instance of alleged manipulation in the opposite direction, see Saberwal 1972 : 74. 34. Election Commission 1968: I, 12.

35. Mastanaiah v. Delimitation Commissioner, A.1.R. 1969 A.P. 1. The Court rejected the

argument that the Delimitation Commission was required to locate a reserved constituency in an adjoining district with a slightly higher percentage of Scheduled Castes.

Programs and Their Implementation

49

Proposals for rotation of reserved constituencies have been rejected, and their location has remained relatively stable.>°

Of course, Scheduled Caste and Tribe candidates may stand for

non-reserved seats. However, they have been notably unsuccessful in

winning elections to them. In the first six Lok Sabhas, only a handful of

candidates from SC or ST filled unreserved seats. Five were elected in 1971 and three in 1977.37 In the lower houses of the state legislatures, the record is no more heartening, even though the smaller constituencies mean that local concentrations of population and resources should offer more opportunities for political success. For example, in 1970-71, there were only three SC and two ST representatives sitting in the 2,853 non-reserved seats in the Vidhan Sabhas.3® There is no reservation of seats in the indirectly elected upper houses at both Centre and states. The Rajya Sabha at the Centre is elected by

proportional representation by the members of the state Legislative

Assemblies.*® The upper houses (Vidhan Parishads) in those state legislatures that have them are chosen by a more complicated formula, combining election by the lower house, special qualified constituencies,

and governmental nomination.” Here, too, the number of SC and ST

has remained fairly small. In 1969-70, there were 13 SC and 5 ST among the 228 members of the Rajya Sabha, and 18 SC and 9 ST

among some 703 members of Vidhan Parishads.** There is no constitutional requirement or statutory provision for reservations of political appointments within the legislatures or the government, such as cabinet ministerships*? or membership on stand-

ing committees. However, it early became a convention to have at least one SC cabinet minister at the Centre and in each of the states. The

number of SC ministers at the Centre and the number of ST ministers in

the states has slowly risen.

36. The Election Commissioner characterized a proposal for rotation (on the ground that no area should be pernanently burdened with a reserved seat) as “mere lip service to the principle of reservation” and concluded that the “real purpose of such reservations would be largely defeated if the proposal be adopted” (Election Commission

1959:1, 75). In the 1969 debate preceding extension for the third ten-year period, an amendment proposing rotation and supported by a disingenuous plea to afford op-

portunities for election to Harijans in other localities was negatived without a division of the House (LSD (4th Series), Vol. 34, No. 17, Cols. 361, 379 [9 Dec. 1969]). 37. RCSCST 1971-73 :244; 1975-77 :1, 20-21. 38. RCSCST 1970-71: 122. 39. Article 80.

40. Article 171. It should be noted that in only seven of the larger states is there a second chamber (Art. 168).

41. RCSCST 1969-70:92.

42. Article 164(1) provides that the states of Bihar, Madhya Pradesh, and Orissa shall have a minister in charge of tribal welfare who may in addition be in charge of

Scheduled Castes and Backward Classes or any other work.

50

Setting and Policies

Reservation of seats on elective bodies at the local level is not specifically provided by the Constitution, but such arrangements are allowed by Article 15(4), and nearly all the states provide such reservations by election, co-optation, or nomination in various local bodies and

panchayats. Similarly, most of the states have made some provision for representation of SC and ST in panchayat bodies. There is little data

available on the extent to which

these reservations have been im-

plemented, much less on how effective they are. We may safely conclude that the presence of the Scheduled Castes and Tribes in legislative bodies is accounted for largely by the provisions for reserved seats. Very few members of these groups have been successful in open competition. Even their presence in bodies where there is no reservation (upper houses, cabinets) is due largely to the political base provided by reservations. Thus reservations do provide for a substantial quantitative presence that would otherwise be lacking. We cannot infer that without reservations, representation would be reduced to the present level of success in non-reserved seats. Their geographical concentration would assure the Scheduled Tribes at least a modicum of representation even in the absence of reservations. The more dispersed Scheduled Castes would undoubtedly obtain far fewer seats than they now enjoy, but we should not conclude that they would have none at all. If there were no reservations, some of the political talent among the SC now competing for reserved seats would compete for non-reserved seats. Parties would then need appealing candidates in areas with heavy concentrations of SC; perhaps slate-makers would be less moved by the sense of “they already have their share” that now seems to affect the giving of tickets. (And of course, some representation would be provided by legislators from other groups, who would be accountable to constituencies with sizable numbers of SC voters.)

Perhaps the best indication of the scale of representation that might

be expected in the absence of reserved seats is provided by some data on elections to local bodies where no reservations were in force. In the early

1960s local bodies in West Bengal with a total of 1,081 members contained only 41 SC members (3.8%) and 16 ST members (1.5%).* Among the 66 presidents and chairmen there were 3 SC members and | ST. (According to the 1960 census, 19.84% of the population of West Bengal were SC and 5.91% were ST.)“ A similar picture is presented by some fragmentary data from Gujarat where only 35 (0.5%) of 6,863 43, RCSCST 1963-64:1, 163.

44. The population figures here are somewhat overstated, since SC and ST are less

urban than the general population. But offsetting this, the small territorial con-

stituencies ought to give these groups an advantage. Also, Bengal is reputed to be less resistant to the mobility of these groups than many other areas.

Programs and Their Implementation

51

Sarpanches were from the SC.‘5 The electoral experience of India’s other large and (mostly) low-status minority— the Muslims—does not offer additional reassurance.*6 The effect of reservation of seats on the quality of representation is more difficult to assess. How well served are the residents of the reserved constituencies? And, apart from them, how well served are the

members of Scheduled Castes and Scheduled Tribes—a great majority of the former—who

reside outside

the reserved constituencies? In

what sense do the reserved seat legislators represent them? Does the presence of their fellows in reserved seats assure that they are better served by their own representatives? Finally, is the nation as a whole

well served by these legislators? What effect do they have on the working of the legislatures? In assessing the quality of representation afforded by these representatives, we must recall the structural constraints imposed by the arrangements for reserved seats. The reservations are designed in a way which filters and muffles the representation of the interests of the Scheduled Castes and Tribes insofar as they diverge from the interests of other groups. The reserved-seat legislator—especially from the SC— is responsible to and dependent upon a constituency made up overwhelmingly of non-members of his group. To the extent that these reserved-seat legislators are assumed to be responsible and prudent politicians, this arrangement serves as a filter to keep the divergent interests of these groups from being expressed too directly or forcefully. This is especially the case with the SC representatives, whose constituencies are populated overwhelmingly by others. From a handful of studies we may draw a composite portrait of the Members of Parliament who occupy reserved seats.*7 They are younger and more rural in residence than their fellow legislators; fewer have professions (other than landholding), and they have less formal

education. The median age of reserved seat M.P.’s is in the early forties; that of their fellows in general seats is in the early fifties. The disparity in age has remained constant over successive Lok Sabhas, but the gap in educational attainments has narrowed appreciably. Those in reserved 45. RCSCST

1969-70: 51

46. The percentage of Muslims (who make up about | 1% of the population of India) in legislative bodies is small and declining. Muslim membership declined from 7.2% of

the Lok Sabha in 1952 to 4.6% in 1971 (Gauba 1973:245). Muslim participation in the State Assemblies is at roughly the same level and has been undergoinga similar decline

(Gauba 1973 :262-65; Krishna 1967: 187). 47. My account is drawn from the comparisons of the occupants of reserved and

general

seats found

in Chauhan

and Chopra

1969; Dutta

1969; Narayana

1978;

Mohopatra 1978. These accounts comport with the profile ofScheduled Caste M.P.’s in Parvathamma and Jangam

1969.

52

Setting and Policies

seats are younger when they are first elected to Parliament, and, unsurprisingly, come there with less legislative experience at the state

level.#* We may surmise that similar disparities might be found in state legislatures.*9

It is widely believed that reserved-seat legislators are less articulate,

less assertive, and less independent than thejr fellows. Data on the Third Lok Sabha collected by Professor Henry Hart supplies some evidence for the view that reserved-seat holders were as a group less active and less influential parliamentarians than those elected to general seats.5° They were less likely to hold party office, to serve on select

or standing committees, to hold ministerships, to contribute to debate,

and to talk to ministers. Few, compared with holders of general seats, felt that they were effective advocates of their views.5t One may, of

course, wonder whether the most important sorts of representation are

captured by these measures of participation.5? Obviously, one must be extremely cautious in extrapolating from this

data about the Third Lok Sabha to more recent performance under changed conditions, much less to conditions that might prevail in the future. Since the reserved-seat legislators have come to resemble their

fellows in terms of education, occupation, and urban residence, there

may also be a convergence in legislative behavior. Or it may be that

changes in political alignments have accorded reserved-seat legislators a greater voice and elicited new patterns of participation.

Low rates of participation do not imply a proportionate absence of

effective representation of the interests of Scheduled Castes and Tribes. Presumably the activities of reserved-seat legislators are concentrated 48. Mohopatra 1978.

49. Mohopatra’s

(1973) useful study of the Orissa Legislative study in 1969 (n=85 of

139) found that SC and ST legislators were younger and less educated, came from more agricultural backgrounds, and that fewer of them had independent professions.

50. The data referred to here is that described in n. 28 above. It would of corse be illegitimate to take this picture as reflective of current conditions. And it should also be emphasized that we would be on shaky ground generalizing from data on Parliament to

conclusions about the working of reserved seats in the state legislatures. Nevertheless, the data does enable us to see the legislative reservation device in action, and it may provide a useful base line for comparison with later or state data. 51. The data on which these conclusions are based are presented in Galanter 1979.

52. They provide no measure, for example, of the vision or sustained energy whose absence is marked by Saberwal (1972 : 79) when he notes that the Harijans in the reserved constituency that he studied had ‘‘no leaders who understood the larger institutional mechanisms which underlay and renewed the current inequalities, nor any who had the stamina for changing the larger system so that its proceses would generate effective . . equality.” He does not argue that this absence is produced by the reserva-

tion of seats, but such an argument is made elsewhere in connection with criticism of

the dependence of reserved-seat legislators.

Programs and Their Implementation

53

on matters that affect these groups,* and their effectiveness in representing the interests of these groups should be most evident in their dealing with matters that contribute directly to their welfare: (1) establishing programs of which they are the beneficiaries; (2) insuring deployment of resources to these programs;

(3) pressing for

better administration of these programs; (4) resisting cut-backs and premature termination of such programs. Although I have no systematic data, there is some evidence to suggest that, in the Lok Sabha at least, reserved-seat legislators have been prominent and sometimes influential in matters concerning the working of the prefer-

ence policy.

The presence of the reserved-seat legislators, if not in itself likely to

raise the level of benefits, does seem

to serve as a warrant of the

continuation of these programs at their existing levels. Furthermore, their presence in such numbers provides a quantitative basis for Scheduled Caste and Tribe participation in leadership at the cabinet level; it is unlikely that there would be many persons from these groups in cabinets were it not for the reserved seats. It appears that the initiation and expansion of programs for these groups is associated with the presence of group members in cabinet posts. The first reservations in government service were established by the efforts of the first Scheduled Caste member of the Viceroy’s cabinet. Job reservations were first extended from initial appointments to the promotion stage under: the auspices of the next SC politician to become a minister. Similarly, at the state level, the first extension of reservations to the

53. Kochanek (1968:48) found Harijan M.P.’s in 1960 among the strong supporters of land ceilings and urban income ceilings and generally giving “higher than average support for any process of leveling.” But cf. Mohapatra’s (1973:22) finding that

reserved-seat holders in Orissa displayed no overwhelming concern for these groups. The pattern in state legislatures may differ sharply from that in the Lok Sabha, as suggested by Dushkin 1972:210. 54. Narayana (1980) describes the heavy participation on such issues by Scheduled

Caste representatives in the Lok Sabha from 1962 to 1971 and their success in securing the appointment of the Elayaperumal Committee and the standing Joint Committee on Scheduled Castes and Scheduled Tribes. The reserved-seat legislators demonstrated

their political prowess by killing the 1965 Lokur Committee report, which called for the de-scheduling of some tribes and castes (see chap. 5, §B, below). The dissatisfaction of the reserved-seat members was instrumental in bringing about the Government's first defeat in the Lok Sabha in 20 years, a stunning snap vote on a 1967 motion that the constitutional safeguards for SC and ST were not being fully implemented. See Dushkin 1972:206-8, who notes that this new assertiveness was followed by government’s Tepudiation of the Lokur recommendations, appointment of new deputy ministers from among the Scheduled Castes, liberalization of the rules for reserving promotion posts in

government services, increasing the reservations in direct recruitment, and (after years of promises) making converts to Buddhism and Christianity eligible for post-

matriculation scholarships.

54

Setting and Policies

promotion stage occurred when a Scheduled Caste politician became

chief minister of Andhra;

the ascension of Scheduled Caste Home

Minister in Madras was followed by a dramatic increase in the vigor of prosecution under the Untouchability (Offences) Act.55 The potential leverage enjoyed by the occupants of reserved seats

varies with the balance of party power in the legislatures. The more evenly balanced the party alignment, the greater the potential impact of the reserved seats.5° Of course, representation is not the only virtue claimed for reserved seats—and misrepresentation is not the only sin attributed to them. At the end of this chapter I shall sketch out the range of claims about costs and benefits that might be made in connection with compensatory discrimination programs. For example, do these schemes produce acceptance of Scheduled Castes or do they lead to rejection of them? Do they promote or retard the self-sufficiency and self-regard of the beneficiaries? Optimally, we should like to be able to answer such questions for each preferential program.5”? The data to do this are not available. In any event, it is difficult to separate the effects of reserved seats from the other components of the compensatory discrimination policy. We shall postpone further assessment untill we have surveyed the whole preference policy. But before leaving legislative reservations, we should note their special relationship to the whole of the compensatory discrimination policy. Reservations of government posts or university admissions often operate as guaranteed minimums rather than as allotments of places over

and above those gained on merit. Therefore the net amount of preference they confer automatically declines as members of the preferred groups gain places on merit. Reservations in the legislatures do not have this automatic self-liquidating feature. However, unlike other preferences, they alone are subject toa constitutional time limit— this limit has been thrice extended. Since the presence of the reserved-seat legislators undoubtedly contributed to this result, these reservations are ina literal sense self-perpetuating in a way that others are not. Furthermore, their presence helps to perpetuate other forms of preference; indeed, that is their purpose. They can only perpetuate preferences to the extent that they are influential—so, self-perpetuation is a mark of some success in fulfilling their function of representation. Thus self-perpetuation is not to be accounted a negative feature unless the whole system of com55. Galanter 1972a. 56. E.g., Krishna (1966:32) notes that in 1962, SC legislators in reserved seats

supplied the entire Congress plurality in 8 of 13 state legislatures, Dushkin (1972:205) calculated that in March 1967, at least 9 state governments were dependent on reserved-seat legislators. 57. Galanter (1979) attempts to do this for the reserved legislative seats.

Programs and Their Implementation

55

pensatory preference is thought to be unjustified or to have outlived its justification. Although there is no time limit on the constitutional authorization of other forms of special treatment for Scheduled Castes and Tribes, one of the curious and significant things about the debate preceding each extension of reserved seats has been the extent to which the public, the

press, and many parliamentarians proceeded as if an end to reserved seats was the end of all benefits and concessions to these groups. There is no constitutional or legal foundation

for such an equation. The

political reservations seemed in many minds to encompass the entire system of compensatory preference, and their extension was tantamount to heading off expiration of all benefits. This symbolic fusion may tell us something important about the way in which political reservations function

to lock into place all the other programs

Scheduled Castes and Tribes.*®

for

When we focus (in Parts Two and Three) on the judicial encounter

with compensatory discrimination, these reserved seats retreat into the background. Election disputes about reserved seats have been the source of almost all the litigation about group membership (taken up in chap. 9) because the greatest tangible prize for asserting membership is combined there with the greatest incentive for rivals to disprove it. But with that major exception,5® legislative reservations have remained outside the ambit of the courts. Courts do not get involved in the allocation of reserved seats or the delimitation of constituencies. They have no occasion to assess the working of reserved seats. Our discussion

of the working of legislative reservations should remind us that the

courts are not the only and perhaps not the most crucial shapers of compensatory discrimination policy. If the courts, as we shall see, act to

brake and confine the preference policy, the reserved legislative seats may serve to lend it motive power.

B. EDUCATION AND OTHER SCHEMES The total number of persons directly benefitted by the reservation of

legislative bodies and of government jobs constitutes only a small

58. Of course that insurance function might be performed by other means—e.g., by a

constitutional guarantee of continued preferential treatment until SC and ST enjoyed specified levels of well-being relative to the rest of the population. It might thus be possible to de-link the assurance of continued special treatment from the reserved-seat device, permitting reserved seats to be evaluated in terms of their specific benefits and costs.

59. Legislative reservations problem of the double-member ory change made this doctrine has general implications in the

were also the arena for important litigation about the constituency. After it was settled by the courts, statutmoot in the legislative representation area, although it administration of preferences. See chap. 13, §B, below.

56

Setting and Policies

fraction of the backward classes. They are far outnumbered by those who benefit from a vast array of special schemes for education and

economic

uplift. Among

educational

schemes

are scholarship

pro-

grams, provision of meals, supplies, and hostels for Scheduled Caste

(SC) and Scheduled Tribe (ST) students; and the construction and

maintenance of schools intended primarily for these groups. Joined to these are a lengthy catalog of measures to improve the economic status

of the SC and ST: allotments of (and subsidies to purchase) agricultural land; installation of irrigation wells; provision of livestock, seeds, and tools; encouragement of cottage industries by the provision of training,

supplies, and equipment; support for agricultural and craft cooperatives. Finally, there are schemes for providing housing, roads (in tribal areas),

drinking-water

wells,

health clinics, legal aid, and nursery

schools. The “Backward Classes Sector” of the successive Five Year Plans provides a rough measure of expenditures on special measures for the disadvantaged. Planned expenditures for all categories for backward classes rose from Rs.300 million in the First Plan (1951-56) to Rs. 794 million in the Second Plan (1956-61) to Rs. 1,139 million in the Third Plan (1961-66) to Rs. 1,403 million in the Fourth Plan (1969-74) to Rs. 3,269 million in the Fifth Plan (1974-79) A breakdown of plan

expenditure®! reveals a continuing emphasis on education as the major undertaking for SC. For ST, educational expenditures, now the predominant category, were earlier overshadowed by communications

and more recently by economic uplift for isolated tribal areas. The plan provisions for Other Backward Classes (OBC) have followed a more

irregular course owing to the controversy about the membership of that category.©2 Expenditure on OBC was 9.84 crores in the First Plan, 2.63 in the Second, and 17.2 in the Third Plan. The projected expenditure on

OBC in the Fourth Plan was 8 crores on education, 2 crores on economic development, and |crore on health, housing, etc. There was no

provision under this heading in the Fifth Plan. These special undertakings for the backward form but a small fraction of plan expenditures. India’s planners have been convinced that the improvement of the conditions of these “weaker sections” would flow from overall development. The special schemes for them represent a redistributive sub-theme to narrow the gap that separates them from the rest of the population. It is a distinctly minor theme. In spite of the 60. Planning Commission 1961 : 701; 1970 : 415. The scheme for post-matriculation scholarships is included, except for the Fourth Plan. 61. “Planned expenditure” refers to new developmental spending and does not include spending to maintain established “regular” government programs. 62. See chap. 6, below. 63. Department of Social Welfare 1969 : 162.

Programs and Their Implementation

57

rise in absolute expenditures, the backward classes sector (including the post-matriculation scholarship expenditures of the Education

Ministry) accounted for 1.63% of the planned expenditure in the First

Plan period; 1.87% in the Second Plan; 1.52% in the Third Plan; 1.12%

in the Fourth Plan; and 0.83% in the Fifth Plan.

The amounts in question are a tiny fraction of the planned expenditure, and when we consider the vast population of the groups to be benefitted, we see how thin this special treatment is spread. Table |

shows the per capita planned expenditure for SC and ST in the succes-

sive Plan periods. TABLE

|

ANNUAL Per Capra EXPENDITURE ON SCHEDULED CASTES AND

ScHEDULED TRIBES FOR THE Five Five-YEAR PLaws [IN RUPEES] Scheduled Castes

Education

Economic Uplift

Health, Housing, etc. Torats

1951-56 Ist 0.20

1956-61 2nd 0.66

1961-66 3rd 0.98

1968-74 4th 0.85

1974-79 th 1.44

0.10

0.39

0.25

0.33

0.52

0.31

1.20

1.40

1,38

2.59

0.01

0.15

0.17

0.20

0.51

Scheduled Tribes Ast

2nd

3rd

4th

1.78 1.34

1,82 0.78

2.20 0.29

1.45 0.75

3.90

3.58

3.72

4.05

Education

0.48

0.78

Torats

1.79

Economic Uplift Health, Housing, etc.

Sources:

Note:

RCSCST 18-20.

0.41 0.90

1965-66: 118; RCSCST

0.98

1.21

1970-71 : 116-17; RCSCST

Sth

1.83

1975-77 : I. 16; II.

— The exchange value of the rupee decreased from $ 0.21 in 1951 to $ 0.12 in 1979.

The multitude of schemes, the paucity of useful information about their working, and limitations of space and energy make it impossible to undertake even a rough sketch of these programs. To avoid being swallowed up in a sea of detail, we shall concentrate on the largest single

item among these distributive programs—education. It entails the highest expenditures and directly affects the largest number of indi64. Department of Social Welfare 1969: 162; Planning Commission 1976: 115, 159.

58

Setting and Policies

viduals. Education is widely regarded as the prime instrumentality for improving the condition of the backward; and the elevation of the backward is seen as one of the most important social products of

education.®,

Independent India has undergone a remarkable educational explosion. Expenditures and enrolments have increased steadily and rapidly.

Governinent expenditures increased from Rs. 1,143 million in 1950-51

to Rs. 6,000 million in 1965-66 to an estimated Rs. 14,500 million in

1974-75. Enrolments increased correspondingly from a total of 18.9 million in 1950-51 to over 58.2 million in 1965-66 to over 85 million in

1973-74.°7 In spite of rapid population growth, the growth of educa-

tional facilities has resulted in a great increase in the proportion of the population receiving instruction at every level. For example, enrolment in the elementary grades rose from 37.8% of the corresponding age group in 1950-51 to 83.5% in 1973-74. Enrolment in higher education had risen from 0.7% of the 18-23 age group in 1950-51 to 2.1% of

vastly larger group by 1965-66.

Included in this general increase is an even more pronounced increase in the portion of Scheduled Castes and Scheduled Tribes children who go to school. These groups are still far below the general population in educational accomplishments, but this must be seen against a starting point of almost total exclusion from educational opportunities until fairly recently. By the 1920s, there were thousands of Depressed Class (as they were

then called) students in the primary schools— many or most of them in segregated inferior schools.’° Few went beyond the primary level. In 1927, the Hartog Committee found that enrolment of Depressed Classes children in secondary schools was 2.6% of enrolment in primary schools.71 Only a handful of students from these groups managed to get higher education. In 1931, only 1.9% of the “Exterior Castes” and 0.8% of the “Primitive Tribes” were literate, compared with 8% of the total population.”? In 1961 7.4% of Scheduled Castes and 6.28% of 65. “One of the most important social objects of education is to equalize opportunity, enabling the backward or underprivileged classes and individuals to use their education asa

lever for the improvement of their condition” (Ministry of Education 1966 : 108).

66. Planning Commission 1976 : 75.

67. RCSCST 1975-77 : II, 32ff. The figure of 85 million does not include higher education, which enrolled over one million by 1965-66 (Ministry of Education 1966 : 300). 68. Ministry of Education 1966 : 161; RCSCST 1975-77 : II, 32.

69. 70. 71. 72.

Education Commission 1966 : 300. Indian Statutory Commission (Hartog Committee) 1929 : 226. Indian Statutory Commission (Hartog Committee) 1929 : 220. Natarajan 1972 : 81, 93, 95. The 1931 and 1961 Census Reports are the only ones

that give information useful in making these comparisons. “Exterior Castes,” as they

Programs and Their Implementation

59

Scheduled Tribes were literate, compared with 24.02% of the total population.”> Thus the recent figures reflect a vast increase in educational attainments for the lowest groups, though they still lag far behind

the rest of the population.”*

The expansion of educational opportunity for the general population during the decades following Independence was accompanied by a

special commitment to raise the educational level of the Scheduled

Castes and Tribes. While free primary education was gradually extended to larger portions of the whole population, education through

the secondary level was made free to SC and ST pupils. By 1955 most

states allowed full fee concessions to SC and ST students at all levels.75 Mere exemption from fees was found insufficient. Economic distress, apathy, and reliance on school-age children to earn or look after younger children disinclined many parents from enrolling their children in schools and encouraging them to stay in school.” Some states have encouraged enrolment in the first years of primary school by providing free mid-day meals, free books and supplies, and a clothing

allowance.” And for a fortunate minority, the few who reach the upper primary

grades,

there were pre-matriculation

scholarships.7° Even

among these, the vast majority do not recieve such scholarship help. In

some states pre-matriculation scholarships are awarded on merit rather than as of right. And state governments have sometimes failed to spend

were called, do not correspond exactly to the later-designated Scheduled Castes, but they are close enough for this crude comparison. Similarly, “Primitive Tribes” are not precisely equivalent to Scheduled Tribes. The 1931 tribe figures are for those “20 years and older’’ and may thus understate recent improvement. 73. Natarajan 1972 : 118. The gap between SC and ST and the rest of the population is understated by comparing them with a “total” of which they make up almost

one-quarter.

74. It should be noted that there was a change in the operational definition of literacy. In 193] it meant those who had completed grade 4 in primary education; in 1961 it was based on a simple test of literacy for those who claim to be literate without having

completed primary school.

75. RCSCST 1955 : IT, 106. 76. RCSCST 1958-59 : I, 48 (on the need for incentives for attendance); RCSCST 1975-77: I, 89-90. 77. RCSCST 1960-61 : 161-68; RCSCST 1971-73 : 55-56.

78. In about 1956, a scheme of pre-matriculation scholarships was introduced for the

SC, ST, and OBC. Payments under these scholarships start at grade 5 or 6 (age 11) and Tange from 23 to 72 rupees a year, increasing with each standard so that by standard 1! they range from 32 to 150 a year (Department of Social Welfare 1969 : 191-92). During the Third Plan period a total of 1.7 million pre-matriculation scholarships or stipends were awarded to Scheduled Castes and over one million to Scheduled Tribes (Planning

Commission

1970 : 416). This is less than halfa million a year—a small fraction of

population of these groups. Most states have extended the schemes for prematriculation scholarships to OBC. Incomplete figures for ten states show over 88,000 pre-matriculation scholarships for OBC in 1960-61.

60

Setting and Policies

the amounts allocated to pre-matriculation Elayaperumal Committee found that

scholarships.

The

the governments of A.P., Bihar, Punjab, Rajasthan, and West Bengal have

lapsed the amounts allocated for pre-matriculation scholarships, while they

deny the scholarships [to applicants] on the ground that the fund is not

available. . . .This is due no doubt to lack of interest on the part of these Governments.”?

Lowered barriers, a changing atmosphere, availability of schools, and government aid have contributed to a great spurt in the enrolment of SC and ST students. In the lowest grades the percentage of children from these groups trailed behind the rest of the population, and in the higher grades there were still massive disparities (table 2). TABLE 2

ScHoor ENROLMENT AS A PERCENTA43. OF CORRESPONDING AGE. Group, sy Communities 1973-1974 Classes

Others

sc

ST

I-V VI-VIII IX+

83.5 35.6 21.2

68.9 22.1 11.0

59.3 15.7 77

Source: RCSCST 1975-77 : 11, 32-34. For many purposes such figures should be regarded with

utmost suspicion (suspicion not allayed by the fact that a number of states report

enrolments of over 100% of the relevant age groups). But for the purpose at hand there is

no reason to think they do not givea rough idea of the disparities among SC. ST. and the rest of the population. A slightly different set of figures is found elsewhere in the same report: RCSCST

1975-77 : 1, 88.

With expanding enrolments, wastage (dropping out) and stagnation (repeating a grade) were accentuated as major problems in Indian education generally. Of 100 pupils enrolled in grade 1 only half reached grade 3 in the prescribed 2 years, and only halfof that half finish grade 6 on time. The rate of attrition is even more pronounced among

Scheduled Castes and Tribes.® There is a strong correlation between

social status and the proclivity to stay in school.®1 In one revealing study, comparing the influence of a multitude of factors, caste was the

aspect of social structure with the strongest effect on wastage.®? 79. Department of Social Welfare 1969: 193.

80. Sharma and Sapra 1969: I. 81. Sharma and Sapra 1969 :

81.

82. Sharma and Sapra 1969: 96 ff. The more influential variables were behavioral

(e.g. steady attendance) and attitudinal (e.g. motivation for learning from home). It

Programs and Their Implementation

61

For those who manage to remain in school there are serious problems of the suitability and quality of the training they receive. The former policy of creating special schools for these groups has been severely criticized for perpetuating their segregation. There is some indication, too, that the standards in schools intended primarily for Scheduled Castes and Scheduled Tribes are much lower than in general schools. It is now generally agreed that mixed schools are preferable

wherever possible.®

Even where they are located in areas without special problems and enrolled in mixed schools, children from low-status groups tend to attend schools with the least-adequate facilities, the least-qualified teachers, and the least-advantaged fellow students. Combined with lower

resources

and

motivation

in the family

setting,

these disad-

vantages cumulate as they move up the educational pyramid. Of those

who survive the course, most emerge with academic records that are not

prepossessing.®5

Notwithstanding these difficulties, the number of Scheduled Caste students who reach higher education increases steadily. The contrast

with the situation before- Independence is stunning. In 1927, out ofa total of almost 55,000 college students in all of India (excluding Bengal), there were 82 students from the Depressed Classes—less than 15/100 of one percent of total enrolment.® By the 1961 Census, SC and ST were still only a tiny portion of the nation’s graduates—a very generous estimate would be about 1.6%, a more cautious one closer to 1%.87 Access to higher education has remained a matter of prime concern

might be thought that these things are associated with the educational attainments of

the parents. But of course the parents with the lowest educational status are most likely to belong to the lower castes. D'Souza 1969 : 91. 83. The number of pupils in special schools declined from 829,000 in 1951-52 to 465,000 in 1960-61 (Ministry of Education 1955: 111; 1968: 135). One of the causes of low enrolment is lack of schools in depressed areas, and as schools are established to

meet this need. more children will in fact be receiving separate education. This is especially so for Tribals where the policy of bringing schooling to remote areas and making the first years available in tribal languages effectively separates them. 84. In a study of 21 schools in Chandigarh the schools were stratified on the basis of fees: in the higher-grade schools the pupils were exclusively of the “Advanced Castes,” in third-grade schools there were 25% “other castes” (D‘Souza 1969 : 100-5).

85. Shah and Patel 1977 : 59. 86. Indian Statutory

Commission (Hartog Committee)

1929: 128, 220.

87. There were, according to the 1961 Census, 1.147 million graduates. There were 6,307 SC. and 1,087 ST graduates counted in urban areas. The rural SC and ST were

listed only as “matriculation and above.” If we make the unlikely assumption that the same proportion of rural as urban matriculates were graduates, we arrive at an

estimated total of 16,223 SC and 2.938 ST graduates—about 1.6% of the total.

62

Setting and Policies

to leaders of all backward class groups. A centrally funded scheme of

post-matriculation scholarships for Scheduled Castes was founded in 1944. It was extended to Scheduled Tribes in 1948 and Other Backward

Classes in 1949. There was an early commitment to provide a scholarship for every qualified applicant from the SC and ST. Total scholarships for these groups grew from 731 in 1948 to 37,372 in 1958, to 156,834 in 1968, and to 350,000 in 1975. Expenditures rose from Rs.

500,000 to Rs. 136 million.®* Virtually every applicant from these groups received a scholarship. Scholarships for Other Backward Classes were more highly competitive.89 The number of scholarships for OBC increased rapidly in the early years, but more slowly after 1955 as the whole question of special

treatment for this group became controverted. Until 1961 the OBC for purposes of this scheme were lists of castes (and communities) supplied by each state. In 1961 the Center abandoned

the earlier notion of

drawing up an authoritative list of OBC and recommended to the states a changeover to an economic basis for determining backwardness.%° From 1963-64, post-matriculation scholarships were awarded on an income basis. The applicants were divided into two slabs: those with annual family incomes of less than Rs. 1,500 and those with income between Rs. 1,500 and Rs. 2,000 (except in the case of technical courses,

where the upper limit is Rs. 2,400). Students from the first slab had first priority for scholarships; any remaining funds were used for those in the higher slab. Within the slabs, students from families with lower income

were to be given priority, but where income was the same and there were insufficient funds available, selction was to be on merit.9?

For SC and ST applicants, there was a much higher income ceiling: scholarships were available to those whose monthly income is below Rs. 500.9? Several states pay far less than the stipends recommended by the central government— Rs. 40 per month to hostellers and Rs. 27

to day scholars for graduate and undergraduate courses, and Rs. 75 for

hostellers and Rs. 60 for day scholars in Medicine and Engineering.*? 88. RCSCST

1975-77:1. 93; RCSCST

1971-73 : 274-75.

89. E.g., in 1958-59, there were 37,654 applications but only 12.390 scholarships

awarded. 90. See chap. 6, below. 91. Government of Maharashtra, Education and Social Welfare Dept., Resol. No. CSP 1065-M(SG) Statement C, pp. 15-16 (1966). The recipients under this new dispensation were called “Economically Backward Classes” from 1963 to 1965, when their name was changed to “Lower Income Group Students.” 92. There had originally been an income ceiling, which was withdrawn in 1957 and

re-imposed in 1961-62. 93. E.g., Punjab pays Rs. 15 to intermediate students and Rs. 20 to B.A. and B.Sc. students.

Programs and Their Implementation

63

The central government’s standards were set in 1954-55 and have not been raised, in spite of great increases in the cost of living. As one might expect, SC and ST students are concentrated in the less prestigious, less demanding, and less potentially renumerative subjets. Students from the OBC have a much different profile of enrolment—a

much larger portion are students in the highly competitive engineering and

medical

fields.

In

1958-59,

37.8%

OBC

students

at the post-

matriculate level were studying medicine or engineering; in contrast,

only 6.4% of post-matriculate SC and ST students were enrolled in these subjects. These students tend on the whole to be less well prepared than their fellows. As we have noted, many of them come from schools with poor

facilities, as well as from families with little educational background. They tend, as might be expected, to be poor academic performers. Although little data is available, it seems that the wastage (drop-out) rate among SC and ST college students is staggeringly high. For example, a survey conducted in Maharashtra found that 25% to 40% failed in their first-year exams, about 70%

failed in intermediate, 10% in their

junior year, and between 40% and 60% in their final exams. Only 8%

received their degrees within the prescribed four years, and about 85% left college without a degree.95 Those who do get through tend to get low grades.*6

The policy of reservations has been widely adopted in higher education, but the general expansion of facilities has meant that there is room enough for everyone except in a few highly coveted fields. Reservations are of consequence only where opportunities are scarce—notably in institutions for professional and technical training. The Ministry of Educa-

tion requested that all states and universities reserve 15% of places in these

institutions for SC and 5% for three years and the minimum Most states have provided for tions. But although accepted

ST; that the maximum age be relaxed by marks for admission be reduced by 5%. such reservations in at least some instituin theory, they have not been fully im-

plemented. A rough indication of the actual extent of such reservation

and of its effectiveness is provided by the results of a survey of medical and engineering colleges conducted by the Elayaperumal Committee. In the 42 colleges that replied, less than 11% of seats were reserved for SC and far less than half of these were actually filled by SC

candidates.9” The general level of inclusion reflected in these figures comports with several later readings. From data available on 67 of 97 medical colleges in 1972-73, Dushkin computed that 6.6% of medical 94. RCSCST 1958-59 : 176. 96. RCSCST 1968-69 : 65.

95. Department of Social Welfare 1969 : 203. 97. Department of Social Welfare 1969 : 237-44.

64

Setting and Policies

students in 1972-73 were Scheduled Castes.® Available data show that 157 of 2,301 (6.8%) admitted to 22 medical colleges in 1975-76 were

Scheduled Castes.9® None of these estimates. tells us about the percentage of SC students who graduate. Reservations in professional-school admissions have been of great concern to OBC. Such reservations are in force in at least nine states.1° Indeed, reservation of these highly prized medical admissions can be taken as a crude measure of state commitment to divert resources to the groups denominated as OBC. The composition of these groups differs greatly from state to state. As reservations for OBC are sometimes larger than for SC and ST and they never go unfilled, they represent a major claim on a much-coveted and scarce opportunity. In a number of states reservations for OBC in professional colleges have become a focus of conflict between putative OBC and other groups.1°? We shall encounter these reservations frequently in the sequel, for virtually all of the litigation about the compensatory discrimination policies in education has been about reservations in medical and engineering colleges for the Backward Classes. Although it remains unclear just how much of the increase in education for the backward classes represents “special treatment” over and

above what is provided for the whole population, it is clear that com-

pensatory discrimination provisions can be credited with securing the inclusion of these groups as beneficiaries of the educational explosion.

And it has enabled them to “catch up” to some extent. If “catching up”

is taken to mean greater percentages gains over a base-line period for the backward than for others, there clearly has been catching up in this sense. But “‘catching up” may also mean narrowing the gap (expressed, e.g., in percentages) between the backward and the rest of the population. Here, it seems, there has been substantial catching up at the lower educational levels, but not at the higher levels. C. ADMINISTRATION Limitations of space and energy do not allow even a cursory review of protective measures, of the development programs in the tribal areas, and of the many distributive schemes outside the field of education.

These

latter include

provision

98. Dushkin 1979 : 663. 100. See table 20, in chap. 6, below.

of free house-sites,

99. RCSCST

subsidies

for the

1975-77 : II, 36.

101. The politicizauton of medical admissions is apotheosized by the arrangement in

Jammu.and Kashmir in which the admissions list was issued directly by the Chief Minister of the State. See Subhash Mohan Jalili v. Principal, A.1.R. 1967 J. & K. 106.

Programs and Their Implementation

65,

construction of houses, allotments of acreage to the landless, improvement of methods of agriculture and animal husbandry; health measures such as provision of wells, mobile dispensaries, and maternity centers; encouragement of cooperatives and cottage industries. Such schemes

have been mainly for the Scheduled Castes (SC) and Scheduled Tribes

(ST). There has been only a small expenditure on welfare measures for Other Backward Classes (OBC).'°? A reading of the official reports suggests that such distributive schemes have on the whole been less intensive and less successful than efforts in education. Since any assessment of compensatory discrimination policies requires that we consider not only the announced schemes but the level of performance, we conclude with a very brief characterization of the administration of compensatory discrimination schemes. Our gross characterization will undoubtedly do some injustice to detail. Hidden in the dismal aggregates are some gems of devoted and effective implementation. But all accounts point in the same direction as far as the overall character of implementation is concerned. There are persistent “shortfalls” in expenditures. That is, the allocation of government moneys for these programs has been greater than the ability of the various agencies to spend it. Over and over again there are reports of failure to spend more than a fraction of the money allocated to programs for the backward classess.*° Shortfalls in expenditure reflect lack of vigorous and imaginative leadership, inadequacy of planning, absence of coordination between agencies, and deficient arrangements for collecting data on the progress of schemes. There are delays at the Centre in making grants; states are disinclined to make sufficient matching grants where that is required. Skill, enthusiasm, and political push are in short supply.*%* How much of this spending actually represents preferential treatment for the backward classes? There is some evidence that adminis-

trators sometimes feel that it is not necessary to include Scheduled

102. Tribes Tribes. various

The annual reports of the Commissioner for Scheduled Castes and Scheduled provide data on central and state schemes for Scheduled Castes and Scheduled On schemes for Backward Classes it is necessary to consult reports issued by the states.

103, Department of Social Welfare 1969 : 130, 144-45. RCSCST

1963-64 : 27, 33,

110; RCSCST 1959-60 : I, 17. (The Commissiner notes that part of the spending is only on paper. He estimates that fully one-third of the spending reported represented transfers to other accounts to prevent lapse ofappropriations.) SASTC :I, 100, observes

that shortfalls are greater in economic and health programs than in education. This suggests that the money is more readily spent where the administration’s job is to pass it on to interested and alert recipients to spend in existing institutions than where the

administration is called upon to erect and operate programs of its own. 104. RCSCST 1959-60 : I, 13-14; RCSCST 1960-61 : 66; SASTC : 1,78, 102.

66

Setting and Policies

Castes within general schemes, so that in part the budget for backward classes replaces rather than enlarges the benefits to which they are entitled under general welfare measures. 165 This is most clearly documented in the case of the Scheduled Tribes. In 1960 the Dhebar Commission pointed out that although the Centre’s special provision for Scheduled Tribes were meant to supplement the general welfare programs of the states, many states had not pursued normal development programs for the ST and instead used the central funds as a substitute rather than a supplement.?% A decade later a Study Team noted “with regret”: Though the attention of State Governments has been repeatedly drawn to the fact that the special provision in the plan for S.T. and O.B.C. is not in lieu of the provision in the general development programme but is intended tosupplement it . . .no conscious attempt has been made by any of the State

Governments with the solitary exception of A.P. to ensure that the tribals receive a reasonable share of benefits of the general development programme to which they . . . are legitimately entitled.”

But persistent complaints suggest that it is also true of the various distributive programs (other than education) for the Scheduled Castes. This point emerges starkly from Aggarwal and Ashraf’s study of Scheduled Castes in Haryana, where we see these policies through the

eyes of their beneficiaries.

A majority of the 500 household heads we interviewed for our study

complained that they were excluded from the general aid programme of the

government. Many asserted that their applications were either not accepted or they were rejected mainly on the plea that there were special provisions for them.

Some

said that they were clearly told to go to the “Harijan

department.” We learned that the general impression among the people, including the government officials, was that “the Harijans were lucky to have all those special provisions.” Most of our high caste informants, too, shared the same view. They said that Harijans should feel satisfied with the special privileges. In fact there was complete ignorance about the smallness of the budget and the myriad administrative problems. People who normally evince sympathy for the weaker sections and support the idea of special privileges know very little about the programme and feel complacent. Even the government officials were rather poorly informed on the details of the social welfare programme and therefore assumed that enough was being done by the govenment for the Scheduled Castes. . . . 105. Dushkin 1961 : 1702. The Commissioner pointed to the need to make benefits

additional rather than substitutes (RCSCST

Welfare 1969: 166).

1960-61

106. SASTC : I, 73.

: 75; Department of Social

107. STTDP: 13.

Programs and Their Implementation

67

In order to get some idea of what this exclusion means for the Harijans, we may compare just one set of figures ...only Rs. 200,000 was provided in the 1973-74 budget of the Welfare Department for the Scheduled Castes and Backward Classes, for business loans. Compared to this the Haryana Government allocated over 220 million rupees in the year 1972-73 as takavi and crop loans to the farmers. As we can see, this figure is a thousand times the amount provided for the Harijans. Had there been no special provisions for the Harijans and they were given a proportionate share of the bigger amount they would have got 44 million rupees. That would have been 240 times the amount that they now get.1%

Sometimes the schemes turn out to be special in the sense of separate

rather than additional. Programs may be administered in such a way as

to perpetuate the segregation they are intended to overcome. Particularly in housing, which tends to be located away from higher-caste localities—and consequently in schools, wells and meeting halls—the provision of special facilities has tended to perpetuate isolation. (Similarly, officials faced with “controversial” claims by Scheduled Castes for the use of facilities, are tempted to achieve a settlement by diverting development funds into provision of what are in effect separate wells, etc.) Although the total expenditure on these schemes is small, there is a great plethora of schemes. The proliferation of small schemes consumes a large portion of expenditures in their administration. Benefits can be made available to only a small fraction of those who are nominally eligible under any given program. Locating programs and establishing eligibility is a demanding undertaking for would-be beneficiaries, made more demanding by passive and torpid administration. Those most in need of help are least likely to have the literacy, information, contacts, and skills to navigate the tortuous channels leading to inclusion.11° These programs are administered in a routine, mechanical fashion. There is no attempt to generate feedback from those affected. Recordkeeping about actual results is fragmentary and incoherent. There is no serious attempt to monitor the effectiveness of programs. One officer in a state office that had been administering a number of schemes for fifteen years reported that the office had never discontinued any scheme on grounds of ineffectiveness. Many programs not only are adminis108. Aggarwal and Ashraf 1976 : 171-72.

109. Thus out ofa sum of Rs. 100,000 sanctioned for a Tribal Development Block,

nearly three-quarters was spent on staff salaries (SSTDP : 29). 110. On the problem of illiterate tribals whose applications for land allotments are Tejected as improperly filled in, see STTDP : 21. Cf. the problem of securing approvals for schemes from standing committees of Panchayat Samitis and Zilla Parishads, which meet infrequently and irregularly (RCSCST 1962-63 : 5).

68

Setting and Policies

tered in a lackadaisical fashion but are of doubtful suitability to local

conditions.111 Lower officers enjoy little discretion to modify schemes

to suit local conditions. Nor is there much enthusiasm, initiative, or

inventiveness ward classes departments, officers move

in the higher echelons. Departments dealing with backattract little talent. Experienced staff tend to flee these which offer-few channels for promotion. Well-qualified on to better prospects elsewhere.!!? The ranks are filled

with rejects from other departments.143

The listlessness of administering the means is matched by thoughtlessness about goals. The mindless, going-through-the-motions quality of administration is nicely captured by Aggarwal and Ashraf: Tn our discussions with the Haryana government officials, from the State to

the tehsil level, we found that no one talked of a goal to be achieved through

the special privileges programme. Outside the government, too, neither the Scheduled Caste leaders nor the high caste persons were aware of any goal. In other words nobody seemed to know what was to be achieved and by what date. The Haryana Welfare Department publishes a report each year... These reports are built around the previous year’s budget which contains all the details about the schemes, the sources of funds, the amount, allocation

for each district on the basis of its population, and even the number of cases to be served. The reports then simply mention the amount spent and the number of grants and loans given. There is no projection of what is to be achieved and by what date. If equality is the objective, which is implied, no mention is made with whom and by when it is to be achieved. All of this clearly indicates that this programme isactivity centered and not goal

centered. Such a goalless programme suffers from many deficiencies. One is that it is impossible to evaluate. Only activities can be listed and expenditure accounted for. Beyond that no one can say whether the programme is achieving anything because achievement can be measured only against a

goal. Furthermore,

it induces a false sense of accomplishment in the ad-

ministrator. He feels satisfied as long as he is able to spend the money and

cater to a specified number of clients. A feeling of complacency is generated

which in turn kills initiative for greater efficiency and achievement. With an easily achievable target of activities already set; only an extraordinary

official will strive on his own.*1*

Generally, administration of these programs is a matter of low priority, which commands little attention from ministries or from legisla111. On the failure of housing colonization schemes for tribals, see STTDP : 8, 16-17

112. RCSCST 1962-63 : 2. 113. RCSCST 1960-61:13; STTDP : I, 49. 114, Aggarwal and Ashraf 1976: 176-77.

Programs and Their Implementation

69

tors.145 Thus legislators have paid little attention to the reports of the Commissioner for Scheduled Castes and Scheduled Tribes. State offi-

gials have been less than assiduous in responding to requests for information. There seems to be little genuine curiosity about how the schemes do work in fact. In spite of the vast quantity of paper that is generated by the administrators, there is remarkably little data about

how the schemes work.

For the most part the departments of social welfare or Harijan welfare found in each state Government are not independent administrative departments, but have only advising and coordinating functions in relation to the regular governmental departments which administer

these programs.16 Administration of special benefits is by the departments concerned with the general welfare programs of housing, educa-

tion, agricultural loans and so forth.

Government departments whose major task is dealing with the disadvantaged tend to have little prestige or influence within the government. As one M.P. observed of the Central Department of Social Welfare: Nobody takes the decisions of this department into consideration and no other department cares for this department. No department complies with

the directions and promises of this department. The conditions prevalant in

the State level [are) worse.117

This is reflected in the surprised observation of the Elayaperumal Committee that although State Ministers in charge of Welfare of B.C.s had unanimously

decided that the principle of earmarking funds in the general sector for the Welfare of the SCs should be implemented this decision was not carried

out.118

One well-placed officer offered the suggestive generalization that the

minister in charge of backward classes or social welfare is usually the weakest in the state cabinet. 115. See Schermerhorn (1969) for an interesting attempt to measure the priority levels of Scheduled caste programs in various states.

116. At the Centre these functions were housed in the Home Ministry until 1964, when all aspects other than government employment and the Scheduled Areas were transferred to a new Department of Social Security, later renamed the Department of

Social Welfare. In a subsequent reorganization they were returned to the Ministry of Home Affairs in 1973 (RCSCST 1964-65 : I, 1; 1972-73 : I, 251). In 1978 for the first time backward classes matters became the subject ofa separate wing within the Home

Ministry. 117, Shri Baidhar Behara at LSD (4th Series), Vol. 16, No. 50 : Col. 3152 (25 April 1968). This was said of the Social Welfare Department. See n.116 above. 118. Department of Social Welfare 1969 : 166.

70

Setting and Policies

Problems in carrying out constitutional policies for these groups were not unanticipated. The Constitution established a special ‘“‘watchdog” agency to monitor the working of the “safeguards” for Scheduled Castes and Scheduled Tribes. Article 338 provides: (1) There shall be a Special Officer for the Scheduled Castes and Scheduled

Tribes to be appointed by the President. (2) It shall be the duty of the Special Officer to investigate all matters relating to the safeguards provided for the Scheduled Castes and Scheduled Tribes under this Constitution and report to the President upon the working of those safeguards at such intervals as the President may direct, and the President shall cause all such reports to be laid before each House of Parliament. 119

The Office of the Commissioner for Scheduled Castes and Scheduled Tribes was established pursuant to this provision. The original conception that the Commissioner would serve as an independent critic of government was gradually effaced as the office took on tasks of oversight and evaluation of programs. '?° It proved no match for the problems of resistance, low priority, poor planning, and lack of coordination that beset these programs. Its investigative potential is limited by lack of funds, inadequate staffing, and the absence of power to require testimony and records. It has no capacity to generate data of its own, and no power to induce other agencies to provide it with information. The Commissioner’s annual reports are filled with accounts of the unwillingness of other agencies to supply information or even to answer

specific inquiries. 121

The original aspiration to establish the Commissioner as an independent body (like the Election Commission or the Union Public Service

Commission) was forgotten over the years. In 1967 the Commissioner’s

field organization was dismantled and replaced by a regional organization subordinate to the Central Department of Social Welfare. To the Elayaperumal Committee this was the culmination of a process in which the Commissioner’s organisation was gradually . . . reduced to the position ofa subordinate office of the Government of India with the result that he became powerless not only in respect of the various State Governments but even with the various Departments/Offices of the Central Government. He 119. This provision refers only to the SC and ST. Article 338(3) says that these

references “shall be construed as including references to such other backward classes as the President may, on receipt of the Commission appointed under (Art. 340(1)] .. .by order specify.” The abandonment of the expectation that there would be a Presidential Order specifying the Other Backward Classes is recounted in chap, 6, §3, below. 120. CWSCST (4th) 1 (1969). 121. On the working of this officer, see Fisher 1970.

Programs and Their Implementation

71

has been powerless even to collect the basic and requisite data for his assessment in fulfillment of his constitutional duties. His annual reports are a record of his bewailings and beseachings [sic] year after year. 27 Without any investigative capacity of his own the Commissioner was reduced, in the words of another Study Team, to [discharging]

his constitutional functions through those very agencies on

whose performance in certain spheres he may be called upon, by virtue ofhis office, to sit in judgment.123

In 1977 the Commissioner reported that his office had “only two Research Officers and two Investigators.” 1%

In 1968, Parliament established its own “watchdog” institution, the

Joint Parliamentary Committee on the Welfare of Scheduled Castes and Scheduled Tribes. This is a thirty-member body elected by the two

houses of Parliament (20 from Lok Sabha, 10 from the Rajya Sabha) for two-year terms. The Committee is mandated to recommend measures to

implement the Commissioner’s Reports, and to monitor central government programs and practices. Unlike the Commissioner’s Office, the Joint Committee has investigative powers. The Committee has issued a series of detailed reports (there were 65 by early 1976), each concerned

with a single agency or problem. A number of these are follow-up

reports, tracing the response mittee has focussed heavily tions in central government higher-echelon educational committees were established

to its earlier recommendations. The Comon extending and implementing reservajobs and in public corporations, and in opportunities.1*5 Comparable oversight in some states in the 1970s.

Rising concern over atrocities against Scheduled Castes and eager-

ness to allay charges of Commissioner’s Office, “‘watchdog” institution Scheduled Castes and

neglect, combined with dissatisfaction with the led to the establishment of yet a third central in 1978. A five-member Commission for the Scheduled Tribes was appointed and em-

powered to investigate the implementation of safeguards (particularly

job reservations) and anti-disabilities laws. The functions of the new Commission were described as corresponding broadly to those entrusted to the Commissioner, and the government suggested that the Constitution would be amended to reflect this replacement.126 122. Department of Social Welfare 1969 : 399. 123. STTDP : 89. 124. RCSCST

1975-77 : 1, 65.

125. On the establishment of the Joint Committee and its performance during its first years, see Narayana 1980. 126. LSD (6th Series), Vol.16, No.5 : Col. 262-64 (21 Jul. 1978). For an interpreta-

tion of this puzzling event, see Economic and Political Weekly 13 : 1159 (July 28, 1978).

72

Setting and Policies

On a day-to-day basis the function of prodding the administration is performed not by these formal “watchdog” institutions but by the

M.P.’s and M.L.A.’s—usually the occupants of reserved seats. They intervene with the administration to champion beneficiaries’ grie-

vances, urge priority for programs, overcome inertia, and move things through channels. As one backward classes leader observed, “If reservations in the legislature are takén away, educational schemes will not

work, for there will be no one to press the ministers.””!27

D. CRITICS, COSTS, AND BENEFITS Few in independent India have voiced disagreement with the propo-

sition that the disadvantaged sections of the population deserve and

need “‘special help.” But there has been considerable disagreement about exactly who is deserving of such help, about the form this help ought to take, and about the efficacy and propriety of what the government has done under this head. The major lines of criticism may conveniently be sorted into objections to what the government has not done and objections to what it has done. Critics of the first (omission) type may in turn be divided into those who criticize the Government for

doing not enough and those who criticize it for doing not the right thing.

Not enough criticisms include insufficiency of appropriations and expenditures, inadequacy in the extent and scope of reservation, failure to implement programs with sufficient zeal. Not the right thing criticisms point to the government's failure to do things it ought to have done— €.g., to aim its educational efforts at preparing backward class students for offices rather than trades (or vice versa). Those who make these criticisms of non-performance or omission have generally been favourable to the main lines of the government’s policy of special treatment

and seek to see it augmented or redirected. 128

A much more widely articulated set of criticisms has come from those who question the main lines of the Government’s policy of special treatment.

These

(commission)

critics, too, may

127. Interview with G. Latchanna, Hyderabad,

be conveniently di-

1965.

128. A great deal of criticism along these lines is found in government sources, especially the topical reports of the Committee on the Welfare of the Scheduled Castes and Scheduled ‘Irribes, and the annual reports of the Commissioner for Scheduled Castes and Scheduled Tribes. A massive compilation of such objections to government policy may be found in the report of the Elayaperumal Committee (Department of Social Welfare 1969). All of the above deal with Scheduled Castes and Tribes. For assessment

of Schemes for Backward Classes, see Government of Tamil Nadu

1971; Governmentof

Kamataka 1975. Views along this line are also forthcoming trom organizations of beneficiary groups.

Programs and Their Implementation

73

vided into those who think the government has done too much and those who think it has done the wrong thing—although many of these critics link the two. The to much view asserts that preferential treatment, especially reservations of various kinds, has been overdone and costly in terms of efficiency, utilization of talent, morale, and basic fairness. But

the great focus of dissatisfaction has been the use of caste and community groups in selecting the beneficiaries. In what has become the conventional wisdom on the matter, the use of caste units has been blamed

for perpetuating the caste system, accentuating caste consciousness, alienating and demoralizing the groups who are benefitted, augmenting political abuse of caste by promoting “vested interests in backwardness,” impeding the development of secularism, and other sins as well. Thus it has been asserted by two sociologists that reservation on the basis of caste . . .has effected the morals, the administra-

tion and the society adversely. The attempt to atone for the injustice

perpetuated on the lower caste [sic] in times past has in effect become a tool

of aggrandisement in the hands of the numerically large and hence politically dominant caste groups in majority of the states. 129

A political scientist has observed that governmental definition of backwardness in terms of caste has perpetuated and accentuated the caste consciousness. The judiciary by

upholding classifications based on caste, has also helped the growth of the virus of casteism. This has undermined the foundations of the secular state and has been detrimental to the development of secularism.

Another sociologist has observed that reservations for backward groups are bound

to prove

perpetuate, more classes. . . .134

harmful

or less,

from a broader

the

existence

point of view, since it will

of the

so-called

depressed

129. Barnabas and Mehta 1965:76-77. 130. B. A. V. Sharma 1968:70. 131. R. N. Saksena 1965:25. A scholar who could fairly be called the dean of Indian sociology expressed dismay and anger at “the Process of proliferation and entrenchment of caste under the specious plea of social justice and upgrading-levelling of the so-called Backward

Classes.

...”(Ghurye

1969:430-31.)

He concludes a lengthy

ical analysis with the prediction that if the reservation of legislative seats is extended, ‘“[w]e will thus have a few caste and sub-caste groups turned into political parties'in miniature and vested interests for preserving their special spoil of the social cake of the Government jobs. From there accretion of other economic and political interests would not fail to begin and thus to-solidify them as exclusivist and separatist communities,” (1969:441).

74

Setting and Policies

This kind of sweeping and drastic condemnation has not been confined to academic observers.43? Such views have been expressed by ad-

ministrators and official committees,9 and by a wide array of intellectuals and publicists. Whether complacent about the progress that has been made or despairing at the lack of progress, participants in public discussion of preferences overwhelmingly depict them as pernici-

ous. Two recent examples convey some of the flavor of this discourse.

The first, an article entitled “Reservation: Blessing or Curse” that appeared in the Sunday Statesman Magazine in February 1978, begins with this assertion: Constitutional reservations have reduced the mass of untouchables into beggars endlessly waiting for the caste-Hindu crumbs of bread, thrown at their whims and fancies . . . . [T]he majority of caste Hindus grow increasingly jealous and curse the recipients as “Government Brahmins.” .. . Except producing some elites among Dalits, the system has done the greatest disservice to the Dalits as a whole. . . . The alleged rising number of atrocities against Harijans is an off-shoot of this upper-caste jealousy and prejudice against these reservations ....[T]he major portion [of the benefits] has gone to create . . . professional elite . . . [which] has turned into a set of new-Brahmins living the life of leaches [sic], sucking the blood of their own brothers and sisters. . . . The system of reservation.... kills the beneficiaries’ initiative, drive and capacity. Everybody admits that the concessions extended to Dalits have not reached them. . . . [T]wenty-six years of reservations have not made any dent on untouchablity—which is the root cause of all discriminations

against Dalits. . . .

The author concludes by recommending that reservations be scrapped

in the hope that “abrupt withdrawal of reservation, which have

pampered them so long will rouse them to action.” Wounding the sleeping tiger (the author’s image) will hasten social change. About two weeks later an article appeared in the Indian Express entitled “Emancipation or Exploitation.” The author finds that “‘there 132. A number of foreign students of India have also voiced negative assessments. Thus the leading study of Indian secularism (Smith 1963:316) found “a profound

contradiction between the objective of a casteless society and the method of evaluating

the backward on a caste basis.” Similarly, a renowned study of Indian politics (Rudolph and Rudolph 1967: 150) describes the caste basis as the “critical contradiction” in the government's efforts to eradicate the effects of untouchability—‘‘a kind of blackmail in reverse,” which “is not only profoundly disturbing but also an important source of alienation and rebellion.” 133. See, e.g., the covering letter of the Chairman of the Backward Classes Commission, the Memorandum of the Home Minister on the Commission’s Report (1956), and

the various reports cited in chap. 6, §C, below. 134, V. T. Rajshekar Shetty in Sunday Stateman Magazine, Feb. 19, 1978.

Programs and Their Implementation

75

is no untouchability today in the manner which existed before Gandhiji

except in rural areas. . . .” If the cheris (untouchables’ hamlets) in the

villages are broken up “untouchability will cease to exist as a fact. The Brahmins, kshatriyas and vysyas are getting thoroughly mixed up through business, marriage and their non-recognition for any special representation, job or privilege.” It is only the Harijans that remain a distinctive element, and their perpetuation for the purpose of political exploitation is the result of “defining the Scheduled Castes by birth.”35 The solution, again, is to phase out preferences on an ascriptive basis and substitute economic and educational measures of backwardness. Without taking up the task of assembling the evidence relevant to the sweeping assertions in these arguments, I present them as specimens of a style of discourse that remains relatively constant. They display a notable unanimity: whether you take the Statesman view that untouchability is undented or the Indian Express view that it is vanishing, the conclusion is that preferential treatment for low castes should be abolished. There is no comprehensive study from which we might gauge the extent of public approval or disapproval of preferential treatment, but some scattered survey data enable us to sketch a very rough profile of public views. Anant’s useful 1968 survey of three localities in North India reveals a consistent pattern of sharp division about preferential treatment. For example, just over half (52%) of his respondents favor reservations of legislative seats for Scheduled Castes, but 62% thought they should be terminated (in 1970, when they were to expire). Seventythree percent favored special educational facilities for Scheduled Castes.136 Anant conducted a follow-up study four years later after

the flowering of Mrs. Gandhi’s egalitarian rhetoric and after the 1969

renewal of reserved seats for another ten years. In 1972 he found that an even higher percentage (86%) of residents of these localities favored special educational facilities, and there had been a striking increase in 135. K. Santhanam in Indian Express, March 6, 1978. The view that only compensat-

ory treatment prevents the disappearance of caste inequality had greater currency in the 1960s, before the upsurge of rural violence against untouchables. For example:

“There is at present no problem [of untouchability] and all that need to be done has

been done and more. .. .It is the perpetuation of propaganda on behalf of ‘untouchables’ in this country that serves as a constant reminder to those born in the ‘untouchable’ castes that they are ‘untouchables.’ The sooner the propaganda is stopped, the

better it will be for all concerned, and for the country” (letter from B. G. Balakrishnan in

the Economic Weekly 17:624 [1965]). 136. Anant

1972: 92, 97. Cf. Mohopatra’s (1974) finding that ofa 1973 sample of

political science majors (N= 328) in Orissa, 45% favored abolition of reservation of seats in the state legislature, and 38% opposed.

76

Setting and Policies

support for reservation of seats in legislative bodies, which was now approved by two-thirds of the respondents. '37 Rural, lower caste, and less-educated people were consistently more supportive of preferential treatment than urban, higher caste, and moreeducated respondents.1* This pattern is strikingly illustrated (in table 3) in the 1968 responses to the question of whether reservations for Scheduled Castes in government employment were “necessary” or “ruinous.” One recent study reports that “‘the so-called middle range backward

castes offer a stiffer resistance to the rising status of the Harijan than the TABLE 3 Views oF RESERVATION IN GOVERNMENT EMPLOYMENT FOR SCHEDULED Castes : THREE Nortu INpIAN Loca.ittes tn 1968 Responses

N

Groups By Location

Urban Rural

Necessary %

Ruinous %

_No Response %

371

43.1

56.6

03

140

33.6

66.4

0

53.1

05

23.2

06

12.4

0.8

374

65.5

33.7

08

By Caste

Brahmin

Kshatriya

115

46.1

Vaishya

209

46.4

Harijan

168

76.2

Sudra

113

70.8

53.9 27.4

0

18

By Educational Level Illiterate

Less than H.S. High School College, but less than B.A. B.A.

Postgraduate

Tora croup Source:

121

86.8

235 151 91 95

68.1 49.0 33.0 27.4

30.2 51.0 67.0 72.6

78.4

20

749

54.3

45.0

0.7

51

19.6

17 0 0 0

Anant 1972:100. Calculations of significance will be found there.

137. Anant 1978.

138. The pattern of higher caste hostility to government help for Scheduled Castes reappears in the responses to Anant’s open-ended inquiry about “what else should the Government do to help the Scheduled Castes.” See Anant 1972, chap. 10

Programs and Their Implementation

77

very highest castes. ... [T]he former . . feel that their position is

threatened by the growing power of the scheduled castes. They are also jealous of the benefits and advantages available to the scheduled castes under the constitution.” °° The survey data concur that educated and urban people are personally more accepting of mobile untouchables and more opposed in theory to the continuation of the caste system.1 But even though less-advanced segments of the population may have more to fear from the direct competition of Scheduled Castes, it is the

higher castes, the educated, and the urban who seem most opposed to preferential treatment. The articulate criticism of the policies by intellectuals taps a broad vein of dissatisfaction among elite groups. A sizable minority of highranking government officials disapprove these policies.141 There is intense resentment of them among government servants, epitomized in the mocking reference to the Scheduled Caste officer as a ‘‘son-in-law of the government” or a “Government Brahmin.”4? Although the educated elite have found little to say in its favor, preferential treatment has proved durable if not popular among politicians.43 And members of the beneficiary groups remain generally favorable. While not unaware of some of the costs, Scheduled Castes and Tribes are convinced that preferential treatment is benefi139. Sachchidananda 1977: 172.

140. See Anant 1972, chaps. 4, 5, 6. Ina survey of seven selected Rajasthan villages on

readiness to accept social change, Oommen found that admission of lower castes to educational institutions engendered much less resistance than special development Programs and economic concessions. He notes that his more change-prone, educated respondents shared a “general disapproval...in regard to accepting caste as the crite-

rion for extending economic benefits... [T]he educated and the literate are likely to feel that special concessions based on caste deprive them of their legitimate opportunities” (1968 :935). 141. Taub

(1969: 183) asked 55 high-ranking officials in Orissa whether they ap-

or disapproved

of reserved placés in legislatures and administration:

18

(32. 7%) disapproved. These results are remarkably close to those of Subramaniam

(1971: 118), who surveyed 98 young I.A.S. recruits and 46 older 1.A.S. officers in 1963. Thirty-seven percent of the veterans and 34% of the recruits favored abolition of special concessions for SC and ST in recruitment for public services; only 8% of the veterans and 13% of the recruits thought preferential treatment should be expanded; the ethers

chose, ‘somewhat unenthusiastically,” to retain it. 142. The term “son-in-law” is used, Chitnis (1977:210) observes, to connote “a

person whose every whim is indulged . . .indiscriminately.” Taub (1969: 187) reports the “explosive hostility” that inquiry about reservations elicited from “sophisticated and highly educated” I.A.S. officers. On resistance to reservation orders in government offices, see chap. 4, below. 143. Some scattered survey data suggest that clites are arrayed along a scale from politicians, who are most favorable, to intellectuals, who are most opposed. In 1964 and

1965 surveys of attitudes toward community development programs, various groups of

78 = Setting and Policies cial. In Anant’s 1968 survey of three areas of North India, Harijan respondents supported reserved legislative seats by four to one. While 72% of other respondents would have terminated reservations in 1970, only 26% of Harijans would, far less than any other segment of the population.'*5 Unlike all other segments, a majority of Harijans disagreed with the assertion that reservation of seats tended to separate

them." There was among the Harijans a significant minority: 26%

would have abolished reservations in 1970, and 23% thought reserved posts “ruinous.” But on every measure there was more support for preferential treatment from Harijans than from higher caste groups. Four years later the support of Harijans for preferential policies was

even higher. 47

Similar overwhelming support for preferential treatment is shown in Aggarwal and Ashraf’s survey of 500 Scheduled Castes household heads in Haryana. Only 2.8% thought that guaranteed political representation was not a good thing.’* Although there was less agreement about the usefulness of job reservations, sentiment to continue them for a long time was almost unanimous. '*9 respondents were asked whether they agreed development programmes would automatically ing poorer sections; no special programmes for who chose this statement over an alternative treatment was essential is as follows:

that “proper implementation of the contribute to the welfare of all includthem are necessary.” The percentage which stated or implied that special

Opposition to Special Treatment Among Various Elite Groups (1964-65)

Respondents Opposition M.P.’s (1965) Congress M.P.’s (1965)

Senior Press and Public Figures (1965) Officials (1964)

Pradhans, Sarpanches, Panches (1964) Executives of cooperative societies (1964)

Educationists (1965)

(n)

Percent agreeing that no

Special programmes are

(23) (77)

(63) (360)

(1,440) (900)

(37)

necessary 0 20

30 34 38 40

46

Sources: Indian Institute of Public Opinion, Monthly Public Opinion Surveys No.

109-10:59:No. 124: 19.

Note: Although the statement refers to the “poorer sections,” it was prefaced in the interview by a reference to the “weaker sections.” The alternative: ‘Benefits of the development p1 mmes go normally to the relatively stronger [and economically powerful] sections in the village.” The alternative given to the three large groups of respondents omitted the words “and economically powerful” and included an additional clause: “‘a special programme for the poor sections is, therefore, essential.”

144. Computed from Anant 1972:97. 146. Anant 1972:96.

148. Aggarwal and Ashraf 1976: 145.

145 Anant 1972:92.

147. Anant 1978.

149. Aggarwal and Ashraf 1976: 132-33.

Programs and Their Implementation

79

There is similar endorsement of preferential treatment by the next generation of educated untouchables. Thus Chitnis finds that Scheduled Caste college students and high school students believe that the programs are beneficial.5° Almost two-thirds of the college students disagreed with the statement that “‘the Scheduled Castes have advanced so much that they can stand ontheir own in open competition and do not need reservation any more.’’154 That beneath their overwhelming approval of special treatment there is considerable unease and awareness of the costs is suggested by responses to ‘three of Chitnis’s items: the large sample of school and college students were asked whether preferences damage self-respect, foster dependence, and destroy initiative. In each case, almost three-

quarters of all the respondents end up in the “can’t say/don’t know” column—a striking departure from the 5% to 30% ‘“‘can’t say” re-

sponses on other items. 15?

There is no open public defense of the ancien régime. Everyone is against untouchability and against caste. Public debate takes the form of argument among competing views of what is really good for the lowest castes and for the country. These views involve a host of assertions about the effects— beneficial and deleterious—of compensatory

discrimination policies.

We shall address some of these contentions in the course of this study. For now, a summary of the main arguments for and against preferential

treatment may help to sensitize us to the issues and alert us to be on the look-out for relevant evidence as we examine the programs, their implementation, and their treatment by the courts. As we proceed to discuss specific preferential arrangements, we shall from time to time attempt to sort out the claimed benefits and costs. Here, it may be appropriate to sketch out in the most general terms the full range of claims that are made as to benefits and. costs—the various ways in which the policy of compensatory discrimination allegedly helps or hurts the protected groups, others, and India as a whole. Rough and redundant as it is, this anthology of claims will provide us with.a checklist that will help in devising appropriate standards for evaluating specific schemes. For convenience each claimed benefit is paired with the opposite claim of cost. It would simplify matters if each of these pairs represented points on a single dimension that could be unambiguously measured. Unfortunately, the most that can be claimed

for them is that each is a composite of sometimes reinforcing but

150. Chitnis 1977:231. A much more primitive study almost 20 years earlier found

similar results :of 286 Backward Classes students receiving government scholarships,

88% favored reservations in admissions and 79% favored reservations in government services (Ayyar 1965:60-61). 151. Chitnis 1977 :233.

152. Chitnis 1977 :239.

80

Setting and Policies

occasionally conflicting qualities that may conveniently be grouped

together. Since the lines between the claimed effects are not always distinct, some overlap and redundancy is unavoidable. It was necessary, for convenience in labelling, to devise some rubrics that are not

found in ordinary talk of these matters, imparting a somewhat stilted quality to the list. These claims are listed here (in table 4) in an order that proceeds roughly from the most focussed and immediate to the

most capacious and remote, from those which speak of impacts primarily on the beneficiary groups to those which concern the shape and

career of the whole society. This catalog is clearly not a set of explanations of why these policies were adopted, although such goals undoubtedly played a part. It is a set of standards for judging these policies. But whose standards? Obviously

they are the author’s but I would claim that they are more than the

author’s—a claim supported by their provenance, for they are refinements and generalizations of arguments ‘found in current Indian discourse about these policies. I would claim further that the list en-

compasses most of the standards that would occur to a disinterested policy-maker.

(By this I mean a policy-maker concerned with these

policies per se rather than with their implications for his political fortunes.) In this accounting of costs and benefits I have put to one side those benefits (and costs) that accrue to individual actors from supporting or opposing a particular program (apart from some that enter

incidentally under the heading of diversion). This is not because I suppose that policy-makers omit consideration of the personal and political gains and losses that such positions entail. The course that

policy takes is very much shaped by this second level of costs and

benefits. But although any given actor may have his own schedule of priorities and his own admixture of second-level objectives, almost all

would share at least some part of the goals implied by these standards. And virtually all would appeal for support in terms of these standards. That actors differ in their priorities and goals as well as in their estimates of fact does not reduce the usefulness of exposing and articulating standards for judging these policies. The evaluation of these compensatory programs involves a two-stage inquiry. First, there is what we might call the problem of performance. Does the program actually deliver the goods (more jobs or housing or better performance in schools or whatever)? In making such judgments we must be wary of all the pitfalls of measuring program effects. Having satisfied ourselves that the program has the projected effect, we then face what we might call the problem of achievement. Has the program

produced the results that it is supposed to achieve—do more jobs for Scheduled Castes produce considerate treatment by officials, or stimu-

Programs and Their Implementation

TABLE 4

81

ALLEGED BENEFITS AND Costs oF PoLicy oF ComPENSATORY DISCRIMINATION REDISTRIBUTION

Preferences provide a direct flow of

These resources are enjoyed by a small

valuable resources to the beneficiaries in larger measure than

segment of the intended beneficiaries

and do not benefit the group as a

they would otherwise enjoy.

REPRESENTATION

whole. vs.

Preferences provide for participation

MISREPRESENTATION

By creating new interests which diverge from those of the beneficiaries, preferences obstruct accurate repre-

decision-making by those who effectively represent the interests of the

beneficiaries. interests that would otherwise be under-represented or neglected. INTEGRATION

DIVERSION

vs.

sentation of their interests.

vs.

By affording opportunities for participation and well being, preferences promote feelings of belonging and foyalty

ALIENATION

By emphasizing the separateness of these groups. preferences reduce their

opportunities (and feelings of) com-

mon participation.

among the beneficiaries, thereby

promoting the social and political integration of these groups into Indian society. ACCEPTANCE

vs.

Preferences induce in others an awareness that the beneficiary groups are participants in Indian life whose in-

Preferences frustrate other:

they consider unfair favoritism and educate them to regard the beneficiaries as separate elements who, enjoy their own facilities and have no claim on general public facilities.

terests and views have to be taken into

account and adjusted to.

INTEGRITY

vs.

achievement and personal efficacy that enable the beneficiaries to contribute to national development as willing partners.

By broadening opportunities, preferences stimulate the acquisition of skills and resources needed to compete successfully in open competition.

MANIPULATION

Preferences subject these groups to manipulation by others, aggravate their dependency, and undermine their sense of dignity, pride, selfsufficiency and personal efficacy.

Preferences permit forms ofaction that

promote pride, self-respect, a sense of

INCUBATION

REJECTION

Vs.

OVER PROTECTION

Preferences provide artificial protec-

tion, which blunts the development of the skills and resources needed to succeed without them.

MOBILIZATION

vs.

ENERVATION

By cultivating talents, providing op-

By making them dependent, blunting

consciousness, preferences enhance

ing self-respect, preferences lessen the capacity for organized effort in their

portunities and incentives and promoting their awareness and self-

the development of talent, undermin-

own behalf.

the capacity of the beneficiary groups to undertake organized collective action. STIMULATION

vs.

By projecting an image of comprehen-

By increasing the visibility of the beneficiary groups, promoting their placement in strategic locations, and emphasizing the national commitment to remedy their condition, preferences serve as a stimulus and catalyst of en-

sive governmental protection and pref-

erment, preferences stir the resent-

ment of others, allaying their concern and undermining initiatives for measures on behalf of the beneficiary

larged efforts for their uplift and inclusion. SELF-LIQUIDATION

groups.

vs.

The benefits of preferential treatment

terest in their continuation, while discouraging the development of skills,

resources, and attitudes that would en-

able the beneficiaries to prosper without special treatment. vs.

Preferences compensate for and help to offset the accumulated disablements

portunities they deserve on merit.

vs.

By reducing tangible disparities among groups and directing attention

to mundane rather than ritual stand-

Preferences contribute to national de-

velopment by providing incentive, opportunity, and resources to utilize neglected talent.

COMMUNALISM By recognizing and stimulating group identity, preferences perpetuate invidious distinctions, thereby undermining secularism.

ing, preferences promote the development ofa secular society. DEVELOPMENT

UNFAIRNESS Preferences place an unfair handicap on individuals who are deprived of op-

resulting from past deprivation of advantages and opportunities. SECULARISM

SELF-PERPETUATION These arrangements create vested in-

are mutually reinforcing and will eventually render unnecessary any special treatment.

FAIRNESS

SEDATION

vs.

STAGNATION

Preferences impede development by misallocation of resources, lowering of

morale and incentive, and waste of talent.

Programs and Their Implementation

83

late educational accomplishment, or produce social integration? To

what extent does delivering the jobs entail the costs alleged by critics of

preferential treatment—such as stigmatizing the beneficiaries, fomenting group resentments, lowering self-esteem? This is not a list of the direct effects of compensatory programs—e.g., more jobs, higher literacy, or better nutrition. It is a list of the good and_ bad effects that-are alleged to flow from the performance promised by the program, including the effects attributable to the specifically preferential aspects of the program. Even if all of these dimensions enter into estimation of the overall working of the compensatory discrimination policy, it is not implied that all of them are involved in every specific scheme. Nor is it implied that they are to be accorded equal weight in making such evaluations. Presumably, specific schemes in different fields (education, housing, etc.) or for different groups (SC, ST, OBC)

have a different mix of intended effects. And, the relative weights to be

assigned to those effects (and to unintended by-products) will differ among various participants and observers.

4

Reservations in Government Employment: A Closer Look at the Paradigm Program A

MORE

DETAILED

account

of one

policy

area

may

serve to give us a firmer sense of how such programs operate and an

appreciation of the complexities of evaluating them. If only one policy area is to be developed at length, government employment is an obvious choice, for in a number of ways it is the paradigm case of the Indian policy of compensatory discrimination. It relies on the reservation device as the core of the program; it is promoted by both state and central government; policy in this area is made with due deliberation by informed agencies; the various groups of interested participants are attentive and responsive; the actors have recurrently invoked judicial intervention. It is an area that in the eyes of proponents and opponents occupies a central symbolic position in envisioning the compensatory discrimination policy. To all of this must be added one more factor: the data that make assessment possible, while insufficient in many respects,

are more abundant, accessible, and readily comparable than in any

other field of policy.

Government in India is regarded, not as just another employer, but as.one that affords a degree of security, prestige, and authority not obtainable elsewhere. For a broad sector of the India population, including both established middle classes and many aspiring groups,

government employment is the prime focus of ambition.’ (Perhaps it

would not be unfair to say that the social role of government employ-

1. The overwhelming preference for government over private employment is confirmed by some survey evidence. Eldersveld, Jagannadham, and Barnabas (1968 : 31-32) report that in Delhi State 76% of urban and 89% of rural respondents preferred government over private-firm employment “if the pay were the same.” Similar percentages accorded greater prestige to doctors, clerks, and watchmen in government employ than to those privately employed.

Reservations in Government Employment

85,

ment in India is comparable to that of corporate employment in the

United States.)? As one Indian observer put it: With

us the Government,

be it liked or not, be it respected or not, is

surrounded with an aura of glamour. It is the fount of power, of prestige, of patronage, and last but not least, of employment. If we are on good terms with it, and its representatives, life is all the more easy for us.?

Because government has been their benefactor and because their access to employment in the private sector has been severely limited, dependence upon government for employment opportunities is espe-

cially pronounced among Scheduled Castes (and to some extent among other backward groups).* Government employment is a matter not only of private ambition but of group advancement. Employment in the

upper reaches of government service is often used as a measure of

advancement and security.5 The presence of its members in the administration had been regarded by many groups in India as an assurance of accessibility and sympathetic response to their interests and a warrant of fair dealing on the part of the government. Having men belonging to their own fold at different levels, gives the peoplea sense of confidence that they will get a fair deal, they will be treated with consideration and no officer will scorn them as people of no consequence.®

Thus, from the outset “untouchable” leaders have seen in govern-

ment employment

not only career opportunites, but relief from un-

sympathetic and oppressive administation, facilitation of the utilization of opportunities, and readier implementation of programs for their benefit.” Furthermore, government work is regarded as a source of prestige for both the individual and the group. A leading Scheduled

Caste politician observes: 2. Blaug,

Layard, and

Woodhall

(1969:94 ff.) estimate that over two-thirds of

employed graduates and nearly two-thirds of matriculates work in the public sector, which makes up over 58% of the “organized sector” (defined as public sector plus private firms employing at least 25 persons). 3. Bonarji 1956. 4. Cf. Dahl’s (1961 : 294) observations that American Blacks have had little choice but to seek employment in the public sector because of their exclusion from the private sector of the economy and that upwardly mobile Blacks have regarded government as warm and receptive. 5. Cf. Dushkin 1972 : 179. This measuring rod is used not only by untouchables but by others as well. 6. Tamil Nadu Backward Classes Commission 1971 : I, 84. 7, M. C. Rajah 1925; R. Srinivasan at Indian Central Committee

Ambedkar cited at Austin 1966:20.

1929 : 384;

86

Setting and Policies If today scheduled castes and scheduled tribes are anxious to get proper representation in the Government service it is not that it presents a solution for their economic difficulties but it adds to their prestige and gives them status in society.®

Over the years spokesmen for the backward classes have expended

more concern on government employment than on any other scheme of

preferences.

A. THE SCOPE OF RESERVATIONS

The central government has provided reservations in government employment for Scheduled Castes (SC) since 1943,9 and for Scheduled Tribes (ST) since 1950.'° Since 1947, of posts recruited directly on an all-India basis by open competitive examination, 122% are reserved for SC and 5% for ST. (These percentages were raised to 15% and 7¥2% in 1970.)*1 Of posts filled on an all-India basis other than by

open competition,

167/3% are reserved for SC and 5% (raised to 742%

in 1970) for ST.!? In the case of lower posts likely to attract candidates

8. Jagjivan Ram at Planning Commission 1965:21. Similarly, the Tamil Nadu Backward Classes Commission observed that government service is primarily a status symbol and that “[t]he synibolic and prestige value of Government Service is in almost direct proportion to the extent of the castes’ social backwardness” (Tamil Nadu Backward Classes Commission

1971:1, 39, 88). They find the craze for government

service part ofan “aversion to follow the traditional pursuits . . .due to social stigma and low ritual status” even where wages are satisfactory (id., at 129). 9. Instructions were issued in July 1934 that duly qualified candidates belonging to the Depressed Classes should not be deprived of fair opportunities for appointment because they could not succeed in open competition, but no definite percentage was specified. In 1942 the government concluded that the slow progress was due mainly to the unavailability of suitable candidates, but that reservation of a definite percentage might provide a needed stimulus. In August 1943 orders were issued reserving 8Y3% of posts filled by direct recruitment, raising the age limit and lowering the examination fee for members of these castes. In June 1946 the percentage of reserved vacancies was raised to 12¥2%, corresponding to their percentage in the population (12%% in undivided India). In August 1947 reservation for other minorities in posts filled by open recruitment was withdrawn, though it was continued in respect to posts filled otherwise

than by open competition. Reserved for Scheduled Castes were 164% of posts filled on other than open competition, a percentage apportioned to their presence in the population of divided India (152%) (RCSCST 1951:27 ff).

10. In 1947 the Government decided that no useful purpose would be served by

reservations for Scheduled Tribes, as too few were available for appointments. In December 1947 instructions were issued that appointing authorities should keep in view the desirability of recruiting suitable tribal candidates to vacancies reserved for other minorities (RCSCST 1951 : 29). 11. Times of India, 4 April 1970. 12. The basic policy document is the Ministry of Home Affairs Resolution of 13 September 1950, reprinted at RCSCST 1951:120-22. The stated percentages apply

Reservations in Government Employment

87

only from the locality, percentages are fixed on the basis of the population of SC and ST in the area. The central government has not at any time provided any reservation in government employment for Other Backward Classes.

At the state level there are similar reservations for SC and ST. Reser-

vations range from 5% to 25% of posts for SC and 3% to 80% for ST. Ina few places, both groups share a common reservation. The combined reservation in higher posts’ for both groups ranges from a low of 10% in Kerala toa high of 40% in Orissa. Several states reserve a slightly larger percentage of lower-grade posts.’* All the states except Orissa, Madhya Pradesh, and Jammu and Kashmir had reservations in effect by 1951. 15 In several states these reservations were part of wider schemes for promoting entry of backward classes into government service. Over the years, the reservations for SC and ST have remained largely unchanged, although they have increased by a few percentage points in erstwhile

Madras,

West

Bengal,

and

Punjab,

and almost

doubled in Uttar Pradesh, from 10 to 18%. In several states where these

groups had been treated together with Other Backward Classes, reservations for them were separated out at the insistence of the central government. Although there are no reservations for Other Backward Classes at the

Centre, such reservations are not uncommon at the state level.16 In

1978, at least thirteen states reserved posts for backward classes other than the Scheduled Castes and Tribes. Reservations for Other Backward Classes were found throughout southern India, in Maharashtra and Gujarat, and across the north from Bihar to Kashmir. The proportion of reservations for these groups was heaviest in the south. Reservations are accompanied by an array of other special provisions, designed to enhance the ability of the preferred groups to compete successfully for government posts. These include age concessions—i.e., relaxation of the maximum age for entry into the service; fee concessions—i. ¢., the waiver or reduction of examination fees; reduction of the minimum qualifying marks on examinations and

primarily to posts in Classes I and II. The Resolution provides that for lower posts likely to attract candidates from the locality only, the Government shall fix percentages of reservations upon taking into account the population of SC and ST in that area. 13. Central government jobs are divided into four grades: Class I (senior administrative), Class II (other administrative), Class III (clerical), and Class IV (attendants, peons). On the classification, see Pai Panandiker 1964. The states and public sector employers employ similar classifications. 14. Planning Commission 1965:285 ff. 15. For a survey of reservations in employment dating from 1925, see BCC 1: 127 ff. 16. See table 6 for current figures, and chap. 6 for details. Fora profile of the situation in the early 1960s, see Planning Commission 1965:285 ff.

88

Setting and Policies

the waiver of a passing mark on viva voce examinations; specific safe-

guards against retrenchment; authorization for Public Service Com-

missions and appointing authorities to relax minimum standards; free-

dom for Scheduled Caste and Scheduled Tribe jobholders to register for better jobs while working for the government; travel allowance (to reach the interviewing center), pre-examination training for Scheduled Castes and Tribe aspirants for high-echelon posts, etc. Thus preferences not only exclude others from competing for some portion of the available posts (reservations per se), but let into the competition those who would not otherwise be eligible (e.g., age and minimum-qualification waivers), make it easier for them to enter the

competition (e.g., fee waivers, travelling allowance), and in a few cases

enhance their capacity to compete (i.e., through coaching). It should be added that not all of these devices are in effect in all places or for all

posts. There has been, during the period that reservations have been in

effect, a striking

increase

in the numbers

of Scheduled

Caste

and

Scheduled Tribe government servants. In 1953 there were 20 SC (0.35%) in Class I jobs. In 1975 there were 1,201 —3.4% ofa much expanded work force. Similarly there was a jump from 1.3% to 5.0% of

Class II posts and 4.5%

to 10.7% of Class III. Scheduled Tribes

increased from 0.14% to 0.6% of Class I; from 0.24% to 0.6% of Class

IT; and from 0.47% to 2.3% of Class III. The small percentage changes

conceal

a great

increase

in absolute

numbers

of these groups.

For

example, during this period the total number of employees in Classes I

and II more than sextupled (from 14,500 to 89,200). But the number of SC and ST in these classes increased almost 25 times, from 160 to 4,436.

In spite of this increase these groups, making up population, held less than 5% of these important and The increase at lower levels was less spectacular in but more substantial in absolute numbers.!7 Table

position in the central government in 1953 and 1975.

over 21% of the coveted positions. percentage terms 5 summarizes the

During this period, the number of Scheduled Castes and Scheduled Tribes in upper grades increased manyfold over the very low starting point. The upward movement in their proportion of total employment in these grades was modest but steady. At the same time there was a very 17, Available data does not indicate how much of the intake of Scheduled Castes and

Tribes can be attributed directly to reservations—i.e., how many of the successful candidates would have failed to obtain the post on merit. In the upper echelons at least the portion is high. Of the 59 SC and ST recruited to the I.A.S. in the years 1959, 1961, and 1962, only three would have succeeded without the reservation (Planning Commission 1965:279). Presumably the ratio of candidates who succeed on merit is higher

in less-exalted posts.

Reservations in Government Employment

89

TABLE 5

Scueputep Caste AND SCHEDULED Tripe EMPLOYMENT IN CENTRAL GOVERNMENT SERVICE, 1953 AND 1975

Scheduled Castes

Scheduled Tribes

1953

1953

No.

No.

(%)

1975 No.

(%)

(%)

CLASS I (Higher Administrative)

20 (0.35%)

1,201 (3.4%)

6 (0.10%)

218 (0.6%)

CLASS II (Lower Administrative)

113 (1.29%)

2,695 (5.0%)

21 (0.24%)

322 (0.6%)

CLASS III (Clerical)

24,819 (4.52%)

174,119 (10.7%)

2,548 (0.47%)

36,893 (2.3%)

CLASS IV (Menial)

161,958" (20.52%)

230,864" (18.6%)

14,512 (1.80%)

49,464 (4.0%)

Sources: I BCG 134; RCSCST 1974-75 : 82. Notes: # Includes Sweepers. Excludes Sweepers. substantial rise in the numbers of SC and ST in the inferior grades anda sizable increase in their proportion of the overall figures. The situation in the states, though somewhat uneven, is roughly similar. There have been small gains in higher services from practically no members of these groups to at least token representation. In lower services there are many of these groups in menial grades; there are beginning to be substantial numbers in clerical grades. Table 6 summarizes the situation in the states, comparing the reserved percentages with percentages of places actually held. Data on places actually filled by Other Backward Classes (OBC) is difficult to come by. The census gathers no statistics on these groups; the central government no longer recognizes them as a category; the states that maintain such reservations tend not to be forthcoming with data about them. The general impression is that there is no lack of candidates from the OBC to fill all of the posts reserved for them.1 This

18. Thus OBC made up well over halfof the “qualified applicants” for every grade of

posts in Andhra Pradesh for the years 1962-63 and 1963-64 (Andhra Pradesh Back-

Setting and Policies go

° v Vv

Vv v

IN IN IN

IN IN IN

IN IN

4 + +1

ol zl £

9% 0% G

ol S

IN ol os

IN IN IN

IN IN IN

IN IN

4 + +1

ol rd} £

9% ol tl

ol S

%

PPE

20 ee

IN IN IN

IN IN IN

IN IN

%

«prssssy

S% + +1

ol él L

9% ol tl

%

play

IN IN IN

IN IN IN

OL suleld

9% ol +1

6G6I 4AL4V SNOLLVAUSaY ON ol IN ol S IN S

v Vv v IN os orl

%

peassy

ISsv1D

%

= Padasay’

&% + tl

L

a Vv v

%

PPH

90 ve

OF 98 %

PaAasay

90 os

xy y PP

£% Lol

aoun0og

IL SswId

ISsv1D

%

AISSVTO

SLSOd JO AGVUD

IN L6¢1 eo

ZL6l 161 SL6l

dO As OS

Dao LS Os

UVHI

vaya

LS OS

luowe

aWLNAD

-AOA0F)

dnoig

VUHUNY

INVSSV

8261 9261 9201

1vaQ

cL6l tL61 +161

IN LB Ith

8S og

£69 6e' Fl

-uonendog

adequaieg

9L6I Lnogy ‘aTa} Sisog GNV GaAUASAY SLSOg :SOIALAS ANAWNYZAOD) NI SASSVIQ GUVMYOVG YsHLO GNV ‘saa, GIINGIHS ‘saisv GINGaHOS 9ATAVL

91 Reservations in Government Employment

90 0%

IN 90 6%

OF z 8

8 sl

+1 t él

IN 961 oes

+108

Let +261 +261

9L61 9L61

dO As Os

LS DS

a

IN 80 ve

IN 0 0%

VANNUVN

vy

Hsdavud VAHVIN

dO LS VaLHsva VHVIN DS

+1 t gl 80 61 OF z 8

LS OS

ds DS Hsdavud

PIQUNINVE

MINHSVN

2961 L961

DS

VNV.QIVH

“INHOVINTH

Lol

9961 Lol Lol

IN 8% el 0% 9 IN 0 Le

261 £261

dO DS

se'e 009

+1 L él 8% 09 OF z 8

el +l

£261 9261

+1 t él

0% 9 IN ot OL

€ cl

9e'8

IN os zS%

V Vv OF z 8

*0 Ut

at 8

H v v

9 v v

€ sl

IN IN

¢ GS

688!

:(6)

OF

90 i+

a 8

“il 1

0%

{ce}

IN

€ ¢I

IN IN

S $e

o€

OF

80 Vt

ae 8

68 9L

0%

IN

£ cl

IN IN

S G8

9%

OF

6 Vt a. 8

LZ $8

0

IN

V V IN IN

S G%*

rz)

OF

a Vv rs ES

0%

IN

v Vv

912

ad

Vv

HOARY Bl “uDmsADIS ALL: J 'ZP-GE6 EG

S sl

ol ve

S $I

‘O66 “D'S ELL

ayeIG.“sNSUID | 6] WY LS pue Dg 10) sandy uonendog ——: 104,

“aiqepieae uoneuojut OU ={ Ny ‘UONOWOY = q “UIUINIIIY = Y

“WLW “Uysoy pup muwolfo 2015-0 posvig ryuvf’: Q “BL

$ SI

v0 zz

¢ st

IN Ie OOS

ols 06°61

8261 9461 9L61

6961 6961

dO Ls DS

LS DS

‘IWoNaa

LSOSIY :V_:S222N0g

dy g ‘sours

s0 ve

(91 z 8

(agi

IN” _ Ve

(a)¢

(W9I z 8

MLL

IN 0 9€

[ePOS Jo luaUNIedag Yseperg BIYPUY :g SLI: L-GL6L

WoWUI2A00)

fo Img : 7 ‘8: SL-FLEL LS: BLLI“LLLI O'S 1261 WTV 797 Pad *2 geting

“Appanoadsas ‘A] PuR TLL TL "1 8288819 spun umoYs ae ‘q “2 "A “V S28EIC 2910195

BLEL

i 2 M8261 aun’ g ‘g26l Yue GI ‘uousag ayy : [11 2 EL-ZLEL SP8SEI PAeMyDeG Jo a1BJ01I04 Jo Ss¥iQ) Plemypeg ENYSEIEYEWN | HERES 1BEZ “D'S 9L6L “WIV “Pay fo MIG “2 sass000f": D 2L1E-91E ‘1: SLE] EAELEUILYY td

+% Le

unpsmpunpy soasi99g + 2 ‘E-1 : 9L-S461 98pNg 2oUEULO}12g ‘220y]2.M

Vv Vv (wor z 8

(a)s1

IN z0 08

(a)s1

(aot z 8

1€ ot

AVN .

is DS NVHISVIVE

LS DS

HIAOL)

quow

9261 9L61

9261 9261

dnoasy

a

Jeok

+961 9461

dO Ls : OF rd gt

+ 91

INI (dey £9

%

2261 9261 9261 l¢ gt I

er

1 (duel

+0

IN +0 gal 1g 9 a yl

IN Le

+2 9

f Vv Vv ool v8 al

I (A)0z

al 91

ol

IN UBL

HB 91

I Vv vol +81

€% 99

%

Issvil)

IN (24

aa ce OLLI

v v IN wle ¥ 9

PPH — pa.uasoy,

a 91

1 Vv LL +l

MISTY

L re

v v

PPP,

%

PUASoy

%

PH

%

PI4aaY_

%

d

PPH

%

ssviID

ls

D4N0g—-

%

ALSSWID

SOd JO ACN

Reservations in Government Employment

93

is particularly the case where the OBC category is broad and includes some prosperous and politically potent groups. In some cases the actual posts obtained by OBC are augmented beyond the stated reservation by the carryover of unfilled reservations from the SC and ST. (Thus the Punjab Evaluation Committee found that where reservations were made interchangeable ‘backward classes fed at the cost of scheduled

caste vacancies out of all proportion.”)*?

B. THE EFFECTIVENESS OF RESERVATIONS In estimating the effectiveness of reservations, we must remember that reservations are not of the full complement of posts but only of current appointments.2° Therefore, if the size of a service remained constant and the reservations were filled every year, it would take the

length of an average service career (approximately 33 years) for the percentage of posts held to reach the percentage of appointments

reserved.?! At the time our most recent data was collected, reserva-

tions had been in effect at the Centre (and in most states) for a period roughly three-quarters the length of a service career. We might then expect that if reservations were fully utilized, the percentage of posts filled by these groups should now be approaching three-quarters of the ward Classes Commission 1970: 120). See also Department of Social Welfare 1969: 290. R. D. Bhandare (Government of Maharashtra 1964: 152 ff) gives the following figures for posts held by OBC in Maharashtra in 1961: Class of Posts Reservation % Held I

Ir

Il IV

12%2%

0.4

9% 11%

9.3 77

(shared with SC and ST)

21

19. Punjab, Welfare Department 1966:91. 20. There have been instances of states reserving much higher portions of current intake, the enhanced figure to apply until the reserved percentage was reached. E.g., in the earty

1950s, 50%

of all vacancies were reserved in Madhya

Bharat (RCSCST

1952:69) and 100% in Saurashtra (Backward Classes Commission 1955:132). In 1960 Bihar announced a policy of hiring only SC in Class IV posts until the quota was reached (Department of Social Welfare 1969:295). This policy was withdrawn after Devadasan v. Union of India, A.1.R. 1964 S.C. 179 (discussed in chap.12,§B, below) established an upper limit of 50% on such temporarily enhanced reservations. In the late 1960s the Punjab was reported to have adopted a policy of reserving 50% of the intake of constables (Department of Social Welfare 1969:372); U.P. had a policy of hiring 45% SC in inferior services until its 18% quota was reached (id., at 389). 21. With a policy of massive reservation for all communities other than Brahmins (i.e., for 95% of its population), “it took [Mysore] a period of 30 years to reduce the percentage of Brahmins from 69% to about 39%” in 1952 (I BCC 138).

94

Setting and Policies

reserved percentage of appointments. But this would be a considerable understatement, for during this period the size of the government’s work force has more than doubled. Far more than three-quarters of present government employees have entered the service during the period since reservations came into effect. So we might expect, very conservatively, that if reservations had been fully utilized each year,

Scheduled Castes and Scheduled Tribes would be present in govern-

ment service at about three-quarters of the percentage reserved for them. Actual representation falls short of the expected. Thus we would expect, conservatively,?? to find over 9% of Scheduled Caste emp-

loyees in Class I central government jobs, but by 1975 there were only 3.4%; we would expect 3.75% Scheduled Tribes, but there were only 0.6%.

However, at the very peak of the system, in the elite Central Government Services, and down in the more populous clerk grades that make up the bulk of the pyramid, actual employment comes much closer to the expected. We would expect to find over 9% SC and 3.75% ST in the Indian Administrative Service (IAS) and in fact we find 8.56% and

4.08%, respectively.23 In Class III posts, we find 10.7% SC and 2.3% ST.

This general profile is confirmed by a somewhat more cumbrous and slightly more refined way of measuring the effectiveness of re-

servations.25 By comparing employment for the years 1953 and 1975 (the earliest and latest for which complete figures are available) we can estimate the number of new jobs. To these are added the estimated

22. Our expectation is that the levels of reservation used to compute the expected presence of SC and ST are 124%2% and 5% respectively, the levels that obtained until 1970, when they were raised to 15% and 7%. 23. RCSCST 1975-77:1, 51 (this is as of Jan.1, 1976). The representation of SC and

ST in the various other elite service corps (Indian Police Service, Indian Forestry Service, etc.) lags behind that in the I.A.S., so the figures here should not be taken as

applicable to the whole of this stratum of jobs (id.). 24. The rate of effectiveness for Class ITI is overstated by our use of the 12¥2% figure. Many of these posts are no doubt covered by the 167 reservation for posts not recruited directly on an all-India basis. 25. All of these measures of “effectiveness” are subject to several distorting factors. First, they overestimate the “expected” total (by assuming that the coverage of reservations is universal) and thus underestimate the effectiveness of reservations. Second, they overestimate effectiveness by underestimating the reservation in the lower grades and in the later years by using the 124%2% and 5% figures throughout. Third, they overestimate the effect of reservations by attributing all of the increases of these groups to the reservation even though some portion of that increase would have been attained on “merit” without any reservation. But this seems fair in so far as the reservation is intended not only to secure places to those who would otherwise not get them, but also to motivate aspirants from these groups and to insure that deserving candidates are not eliminated by prejudice or misjudgement.

Reservations in Government Employment

95

number of jobholders who were replaced (owing to retirement or death) during that period, and a total of the number of jobs to be filled is arrived at. Then we can calculate the number of Scheduled Caste employees who would have been hired were thé reservation universally applicable and completely filled. This is the number by which SC employees should have increased by 1975 (assuming that none of them were among the deceased or retired, an assumption that introduces little distortion since there were so few SC among older employees). Then the actual increase in SC employees is compared with the expected increase. Table 7 confirms that actual increases most closely approximate the

projected increase at the very peak and in the clerical grades. In the wide ranges of administrative grades, the increase falls far short of that projected. In Class I posts, the actual increase is a bit more than a quarter of the expected for the Scheduled Castes and a bit more than a eighth for Scheduled Tribes. In Class II it is four-tenths for SC but less than an eighth for ST. The gap is considerably greater for ST at every level except at the very top. Both calculations depict this U-shaped distribution. Twenty-five years of reservations have produced solid representation in the lower ranks and a strong and growing presence at the very pinnacle, but there remains great under-representation throughout the middle to higher ranges. By dividing our period into three periods of approximately equal length and making the same computation of effectiveness for each period, we can gauge changes in the effectiveness of reservations over time. There has been a great increase in the effectiveness of reservations at the peak. In the latest period Scheduled Castes and Tribes have been coming into the all-India services in numbers close to the reserved portion. In Class I and Class II there have been substantial increases in effectiveness for SC, but reservations for ST in Class I and Class IT have

shown much smaller improvements. This reflects the siphoning effect of the availability of the highest positions and the limited number who can meet

the basic educational

qualifications.26 In Class

III, there has

been improvement in the recruitment of ST, while hiring of SC remained stable. In explaining the profile of effectiveness and ineffectiveness, we must take into account the somewhat deceptive character of the announced reservations. A very large number of positions are effectively out26. There was not a complete count of SC and ST graduates in the 1961 census. There

is a count of urban graduates (6,307 SC, 1,087 ST). If we assume (undoubtedly an

overestimation) that a comparable percentage of rural matriculates from these groups had earned a university degree, we derive an estimated total of about 16,000 SC graduates and 3,000 ST graduates in 1961 (based on census [1966]).

9

4

¢

IS/S AO. ASVAYONI WAIN

9L61 Asenuel’] 01 ¢CH] AJENUe | Wod sUTU

gq

aoLuog aANRASIUMpY ULIPUY 4Oj poling ,

WG

INUUNISO UO PIseg »

“(% 199+ | ‘[o9) stuswode]daa pur sisod Mau JO

L60"E80'1

S061 AONIS 3

:NOY

i

I

SV

Wl II

SV

SLSOd 40 davay

}

1

FEL DO [18 sandy woy paayiaq_ :204n0g

G

. INIS SIVAGLAICNI AON “ADV Id NINLIST

782 SEFLGL LSS

1) 490489 aotasasJo YABUd] aesV.Vw JO} SIBIS EF Jo

EINIS ONTIH aL ILLS &

(GL6I-€S61) Gortag YVaA %17Z V YIAC SADIAWAG LNAWNAIAOSD TVALNAL) NI SaaIY | GA1NGAHDE ANV SALSV7) GITNGAHDG UOd SNOLLVAWISTY AO SSANFALLOAIY ILVWIXOUdY

2 FTaVL

Reservations in Government Employment

97

side the scope of the announced reservations. Some departments are not covered; posts filled by promotion (over one-quarter of all posts) are not

covered (except in the few instances indicated above); posts filled at the creation of a new office are not covered; nor, in some cases, are tempor-

ary posts. A very considerable proportion of government posts are thus exempt from the coverage of reservation—just how many is hard to say.27 Nonetheless, a very large number of reservations have gone unfilled. The poor showing is attributed by the appointing authorities to the lack of qualified candidates among the Scheduled Castes and Scheduled Tribes. This is undoubtedly a factor in some cases, but the existence of large numbers of persons in these groups with the requisite educational credentials lends some credence to the frequent complaints that lack of qualification is not the whole reason.?8 Whatever the shortcomings of the candidates, they are compounded by the shortcomings that characterize the administrative implementation of reservations. Indifference or hostility on the part of the appointing authorities, insufficient publicization of vacancies, the sheer expense of application, all contribute to the under-utilization of reservations. Although there is no data from which it is possible to estimate their telative contribution, it is clear that inflexibility and apathy are endemic in the administation of reservation policy. A committee investigating reservations in government service in Maharashtra observed: Generally speaking, there exists no thoughtful awareness of the intentions

and purpose of Government in reserving posts for the Backward Classes and consequently, the obvious and immediate reasons for the unavailability of candidates are accepted as a matter of course. . . . The picture that arises is ...one of stolid apathy, lack of sympathy, understanding and flexi-

bility. . . . Years of routine approaches and looking to the formal interpreta-

tion of orders and rules have contributed to the elimination of initiative,

enthusiasm and thinking which is particularly essential when dealing with problems arising from a social malaise.?9

27. Some figures from the working of the Employment Exchanges give a rough indication of the shrinkage between announced and actual reservation coverage. In 1966, only 3.8% of the vacancies notified to the Employment Exchanges by the central

government were reserved for SC and only 2.37% for ST. Among vacancies notified by

the state governments, only 0.7% were reserved for SC and only 0.14% for Scheduled

Tribes, (RCSCST 1966-67: 101-3). Of course, the disparity might also result from

failure to notify the Employment Exchanges of reservations. 28. At the end of 1968 there were about 94,000 “‘matriculates and above” SC on the live register of employment exchanges, including over 7,000 graduates. There were

about 14,000 matriculates and above ST, about 1,000 of them graduates, on the Live Register (RCSCST 1968-69:211). 29. Government of Maharashtra 1964: 15.

g8

Setting and Policies

Implementation of reservation policy is almost invariably a matter of very low priority for administrators.°° The intricate maze of provisions

for implementing reservations often simply fails to secure adequate attention

in

offices

overburdened

with

other

matters.

Thus

the

Maharashtra committee reported that “the directions concerning the

carrying over of vacancies and to make up the existing shortfalls are either not understood or not followed.”31 A more recent investigation revealed the same state of affairs. On the basis of spot checks in twenty government offices the Commissioner for Scheduled Castes and Scheduled Tribes concluded that “reservation orders and various other relaxations concessions were being ignored by almost all the or-

ganizations.”5? Officers and Staff charged with implementing conces-

sions were “not fully conversant” with the relevant rules and instructions; there was an inclination to apply “whatever interpretation suits a

particular officer.”33

This administrative obduracy is compounded by a genuine concern for efficiency. Many administrators believe that Scheduled Caste and Tribe candidates are of inferior abilities. As one recent account puts it, the “attitude .. .[of] a major section management, is one of ‘protect your organization against SC/ST.”’* Even if one regards the formal requirements with deserved skepticism, there is very likely some factual basis for such reactions. Reservations, as we have seen, have-a siphon-

ing effect: they draw in candidates with lower performance on various

indicators. In many cases, lower performance undoubedtly reflects inexperience, lack of exposure, or other shortcomings that are temporary and readily remedied on the job. But to some extent the scores reflect limitations of ability or training that are not so readily overcome—the less readily since others, expecting poor performance, expend less effort in on-the-job training. So any original deficiencies may in many cases be amplified by the reaction to them—and the expectation of poor performance becomes a self-fulfilling prophecy! The Centre and some state governments have adopted various de-

vices to offset the lack of zeal of appointing authorities: requirements

that de-reservation be authorized by (or notified to) high ranking officers; requirements of periodic returns by appointing authorities and departments; creation of special grievance cells.>5 30. Nor is it a matter that engages those involved in devising policy for reform of the

administration. For example, the Administrative Reforms Commission does not discuss

reservations anywhere in its 1969 Report on Personnel Administration. 31. Government of Maharashtra 1964: 15.

33. RCSCST 1975-77:1, 45. 35.

Department

1966 :96.

of Social

Welfare

1969:295~-96;

32. RCSCST

1975-77:1, 65.

34. Kirpal and Kelkar 1976:7. Punjab

Welfare

Department

Reservations in Government Employment

99

The most visible and dramatic attempt to improve chances of Scheduled Caste and Scheduled Tribe applicants has been the establishment of pre-examination training centers to prepare them for highlevel examinations. The first of these, funded by the Home Ministry,

was established in Allahabad in 1958. Of 411 candidates who prepared

at this Centre between 1959 and 1966, 153 (37.2%) were selected for the Indian Administrative Service, Indian Police Service or the other cent-

ral services.3® By the late 1970s there were six such establishments operated by the central government, fifteen at the state level, and four new “‘coaching cum guidance centres.’3” The performance of these centers has declined as their numbers have increased. From 1975 the three longest established centers

(Allahabad,

1970 to

Madras,

and

Patiala) trained 652 SC candidates and 202 ST candidates. Of these,

only 49 of the SC (7.5%) and 43 of the ST (21.2%) were finally selected—just 10.8% of the total.>* That such a program could go into its twentieth year without any evident urgency about upgrading this

deteriorated performance is a telling indicator of energy and critical

intelligence expended on the design and administration of preferential treatment. (It may also indicate that a large and inefficient program is preferred to a more select and effective one—in spite of the administrative and human costs.) It is often contended that SC and ST candidates are unfairly eliminated at the later stages of the selection process, particularly where there is scope for personal evaluation in the form ofan inerview or personality

test.39 Scattered data on the selection process at the state level indicate

very severe attrition at the interview stage. For example, over a two-year period in Madhya Pradesh only 21% of SC candidates (who had fulfilled all of the educational and tested qualifications) survived the interview, compared with 41% of other candidates.*° On the other hand, figures on the work of the Union Public Service Commission suggest that the higher the qualifications for the post in question, the less the elimination of Scheduled Castes and Tribes in the later stages of

the appointing process.*?

Preferences may extend not only to initial recruitment into govern-

ment service but to advancement within it. Promotion depends in part on favourable evaluation by superior officers who rate each subordinate 36. Planning Commission 1970:416. 38. RCSCST 1975-77:1I, 11.

37. RCSCST 1975-77:1, 53.

39. E.g., RCSCST 1959-60:1, 143; Government of Maharashtra 1964: 154 ff. (Separate Note of Prof. R. D. Bhandare); Department of Social Welfare 1969: 280 (disqualifi-

cation by “personality tests”’).

40. Department of Social Welfare 1969:314. 41. Planning Commission 1965: 266 ff.

100

Setting and Policies

in a “Confidential Report” as Outstanding, Very Good, Good, etc.4? It is widely believed by Scheduled Caste employees that superior

officers deliberately obstruct their promotion by giving them unfavorable ratings.43 The Commissioner for Scheduled Castes and

Scheduled Tribes concurs that such reports are “generally biased.’44

The difficulties that Scheduled Castes and Tribes have experienced in securing promotions have combined with the demand to be represented

at higher levels to generate an intense concern to insure their access to promotion posts. Government has responded to this concern in various ways. Standards for promotion may be relaxed, as they are in some states,

or there may be a definite reservation of posts to be filled by promotion. The

central government’s

policy on reservation in promotions

has

vacillated. Somewhere in the range of one-quarter to one-third of all

central government posts are filled by promotion, rather than by direct recruitment,‘5 including virtually all of the upper ranges of Class II and Class III. Promotions are made on three different bases and by different procedures: (1) by competitive examination open only to candidates within the department (‘departmental examination’’); (2) by selection (“merit”); (3) on groundsof “seniority-cum-fitness.”

Before 1957, there was no reservation of any promotion posts. In 1957 the Home Ministry recommended that existing reservations be ex-

tended to those promotion posts filled by departmental examination—a suggestion implemented by some departments and later held to be constitutional by the Supreme Court.*6 In 1963 the Ministry of Home Affairs severely curtailed reservations in promotion. It extended them to selection posts as well as those filled by departmental examination, but it confined them to Classes III and IV and to only those posts to which there was no direct recruitment. This virtually ended them. Two and a half years later an organization of Scheduled Castes and Scheduled Tribes Central Government Employees declared: Since the withdrawal of the concessions in November, 1963 . . . not a single

Scheduled Castes/Tribes candidate has been promoted either on the basis of departmental promotion examination or on the basis of ad hoc selections.47 42. Dave 1966 :536.

43. Department of Social Welfare 1969:278.

44. RCSCST 1975-77:1, 43-47. Bad confidential reports can lead not only to loss of

promotions, but to enforced early retirement. The incidence of such enforced retire-

ments of SC employees had led to special procedures for review at high levels.

45. Dave 1966:534.

46. In General Manager v. Rangachari, A.1.R. 1962 S.C. 36, discussed in chap.11. §B,

._

below.

47. Central Government Scheduled Castes and Scheduled Tribes Employees Welfare Association 1966: 11.

Reservations in Government Employment

101

In 1968 the orders were further refined, and somewhat more scope was allowed for reservation in promotions in posts within grades II, II, and IV filled by departmental examination and for which direct recruitment did not exceed 50%. In promotion by selection from grade ITE to II, within grade II, and from grade I to the lowest rung of grade I, there was provision for weightage of individual SC and ST candidates in competing for up to 25% of the available vacancies.** Some state governments have followed the central government in extending reservations to the promotion stage.*® It is promotions rather

than initial recruitment that have been the focus of the most contro-

versy and the most adamant resistance from administrators and from other employees. The Commissioner complains, for example, of “the tendency of administrative authorities to continue ad hoc promotions for long periods to disallow the SC and ST their due share.”5° Another measure of the sensitivity of this issue is provided by our litigation data: an overwhelming majority of the cases about reserved posts concern promotion rather than initial recruitment.51 We have discussed state and central government services in the narrow sense: our figures do not include “government employment” in local bodies or in public sector undertakings. Reservations are applic48. Department of Social Welfare 1969: 252 ff. The weightage takes the form of giving SC and ST employees who are not considered unfit for promotion “one grading higher than

the grading

otherwise

assignable

to them

on

the basis of their record of

service—i.c., if any Scheduled Caste .. .employee has been categorized ..., on the basis of his record of service as ‘good’, he should be recategorized as ‘Very Good,’

(MHA [O.M. 68: 120).

No.

1/12/67}, dated

11th July 1968, reprinted at RCSCST

1967-

49. There is little information available on the working of the short-lived (1961-64)

promotion reservations in Andhra. The Andhra Pradesh Backward Classes Commission (1970:116—17) recommended against re-instituting promotion reservations because of the “demoralizing effect” on other employees. Punjab adopted a policy of reservation in promotions in 1963, reserving 10% ofall promotion posts for members of SC, ST, and OBC “provided [they] . . .possess the minimum necessary qualifications and have at least a satisfactory record of service,” (Punjab, Welfare Department 1966:94). In 1966, the state government revised its policy, aligning it with the then central government policy—expanded promotion reservations to 20% for SC and 2% for BC—but confined these reservations to those Class III and IV posts filled by selection or departmental examination. An Evaluation Committee found the change “demoralizing, retrograde and injurious to the interests of the Scheduled Castes. It has practically denied all chances of promotion for Harijans to higher posts. In actual practice, ever since reservation in promotions was laid down, there have hardly been any benefits to the Harijans in Class II and higher posts, but with this change in policy, whatever semblance of hope there was, has been completely ruled out” (id., at 95). 50. RCSCST 1975-77:1, 45. And see the Commissioner's complaint about the unwil-

lingness of the personnel authorities to modify a zoning scheme, designed to even out promotion prospects among departments, which was felt to undercut the reservation policy (id., at 50). 51. See chap. 14, §C below.

102

Setting and Policies

able to employment by local bodies in some states but not in others, but

application of them has been rather lackadaisical.5? The percentage of SC in these jobs is less than in state services. Although extension of reservations to public sector undertakings was agreed to “in principle” as far back as 1954, such a policy had not been implemented by the mid-1960s. A comparison of government and public sector employment at that point and subsequently is instructive. The Elayaperumal Committee collected data on Scheduled Caste employment in 159 public undertakings whose total employment seems to be about equal to that of the entire central government. It was found that in Classes I, II, and

III combined, these undertakings employed only one-eighth as many Scheduled Castes as did the Centre—and this in spite of the fact that most of them were new enough to have done virtually all of their hiring after reservation had become accepted public policy. The comparison is given in table 8. A similar comparison for Scheduled Tribes isin table 9. The situation in public sector undertakings by state and local governments seemed not very different.53 The sequel offers dramatic testimony to the efficacy of reservations in

securing the inclusion of the beneficiary groups. Reservations in public TABLE8 Comparison OF SCHEDULED CASTE AND SCHEDULED TRIBE EMPLOY MENT 1N GOVERNMENT SERVICE AND Pustic Secror UnperTaAkincs, MiD- 1960s Central Government Posts (1966) Class

Total

I 18 ul

20,379 30,001 1,117,554 Total

20,379

I

30,001 1,117,554

II a0 Source: Now

sc 361 974 99,017 ST

103

80 12,356

Public Sector Undertakings (1965)

% SC

‘Total

sc

% SC

1,77 3.25 8.86

35,512 15,820 1,261,166

73 274 11,608

0.20 1.07 0.92

%ST

0.52

0.27 1.10

17,521"

23,730° 246,654"

% ST

ST

Total

#

3° 1,516"

0,023

0.034 0.65

Department of Social Welfare 1969 : 257. 275; STTDP 257.261. his information is based on the returns from 56 Public Sector undertakings,

whereas the information on Scheduled Castes is based on 159 undertakings.

52. See Department of Social Welfare 1969: 292, 298, 321, 370.

53. E.g., Department of Social Welfare 1969: 298, 353-54.

Reservations in Government Employment

103

sector employment were introduced in the late 1960s.5* By 1975 the numbers of Scheduled Castes and Scheduled Tribes in the administrative grades were still below those found in government, but they were

beginning to be more than token. And in Class III the totals were comparable to— perhaps beyond— those in government service. Incomplete returns show an addition of some 2,000 Scheduled Tribes and Castes in administrative posts—and almost 150,000 in clerical jobs. No attempt is made to extend the principle of reservation to the private sector. There is no provision for reservation by those enterprises

which are aided by the government or do business with it. (Some states have made requests along this line to their contractors.) Nor is there any requirement that employers in the private sector hire members of these groups. Until recently it was not required that private employers refrain from discriminating against them. The Untouchability (Offences) Act, the central government’s anti-discrimination law, did not

reach private employment until its amendment in 1976.55 Nor has any state attempted to enforce its own anti-discrimination policies in private sector employment. It is not clear whether or how much there has TABLE

9

ComPaARISON OF SCHEDULED CASTE AND SCHEDULED TRIBE EMPLOYMENT IN GOVERNMENT SERVICE AND PuBLic SECTOR UNDERTAKINGS, 1975 Scheduled Castes

Class

Government

I I HI

3.4% 5.0% 10.7%

Iv°

18.6%

Scheduled Tribes

Public Sector? Government 1.4% 3.0% 13.7%

26.2%

0.6% 0.6% 2.3%

4.0%

Public Sector* 0.3% 0.4% 6.0%

11.9%

Source: RCSCST 1974-75 : 82, 93.

* Based on returns from 120 undertakings.

> Sweepers omitted from Class IV.

54. The CWSCST (4th) 2 (1969) traces the process through more than a decade of

urgings, admonitions, and occasional compliance

to establishment of a system of

implementation of reservations in 1969. 55. The original Untouchability (Offences) Act 1955 (Act 22 of 1955) was amended by the Untouchability (Offences) Amendment and Miscellaneous Provision Act, 1976, (Act 106 of 1976) and has become the Protection of Civil Rights Act. An addition to §4 (iii) makes it an offense to discriminate on grounds of untouchability in connection with “employment in any job.” The 1976 amendments strengthen the old act in many ways. For a discussion of the weaknesses of the old act and the tendency of the courts to construe its provisions narrowly, see Galanter 1972a.

104

Setting and Policies

been any improvement in the employment position of Scheduled Castes and Tribes in the private sector. Given the enhanced opportunities in

government service and the difficulties in the private sector, it is doubt-

ful if many ambitious Scheduled Castes and Scheduled Tribes have attempted to make careers there. The greater presence of Scheduled Castes and Tribes in government employment than in the private sector reflects greater desire for government work, greater assurance of acceptance upon applying, the existence of personal networks of information about openings, and a selec-

tion process that makes some allowances for their deficiencies and lack

of resources. It would seem to be the absence of these, not the presence of greater ‘“‘prejudice”’ that is responsible for the poorer showing in posts where there are no reservations. For example, let us compare government jobs with Khadi and Village Industries Boards, establishments which we may assume would harbor a goodly share of devoted Gand-

hians who would score low in prejudice against untouchables. Yet we find that their performance in hiring Scheduled Castes is far worse than that of the government in general. For example, in Maharashtra the KV Board had only 27. SC employees out of 653 (4.1%)—and all of them in lower grades.°¢ In Orissa, the KVI Board had 29 SC (4.9%) among its 588 employees—again all of them in lower grades.5” If the data suggest a picture of slow but steady overall progress, it is progress against resistance that is seldom articulated but nonetheless dogged. It is resistance compounded of ignorance, apathy, prejudice, and resentment mixed with genuine concern for efficiency and fairness. ‘The data on litigation allows us to locate the strain and tension associated with implementation of these preferences.5® Reservations in government employment gave rise to some 27 reported instances of

litigation during the years 1950—77—not an impressively large total in a country where government servants are not loathe to litigate about seniority, discipline, and other matters. In particular, there has been

surprisingly little litigation in regard to preferences in initial recruit-

ment into government service. Of the 27 cases, 19 were concerned with reservations in promotions. (Of the remaining 8, 5 were brought by

lawyers seeking to obtain posts of munsif [a minor judge]).

‘These cases involved, not rival contenders fighting over entrance into government service, but government servants (and lawyers) fighting over opportunities for advancement. The battle focusses not at the top, where reservations have relatively little impact on career chances, nor at the bottom, where they are hardly needed (and those involved might 56. Department of Social Welfare 1969: 329.

57. Id., 370.

. The universe of cases that is the basis of these computations is described at

chap.14,n. 45.

Reservations in Government Employment

105

not be able to afford litigation), but in the middle: particularly in the

upper ranges of Class III and the jump from Class III to Class II. Naturally, such litigation clusters around those departments and posts where provision for reservations exist and are implemented with some vigor. For example, of the 13 cases involving the central government, 11 involved three departments (Railways, Posts and Telegraphs, Central Secretariat). As we shall see, the courts have provided resisting government servants with a resource which may be used to curtail and restrict the operation of compensatory discrimination. This incomplete and spotty data enables us to make an initial quantitative sketch of the performance of reservations of government posts: 1. reservations succeed, where exhortation and good will do not, in

getting members of the beneficiary groups into government service; 2. generally, the reservations fall short of announced goals;

3. this is more so the higher the posts, except at the very peak; 4. reservations are far fewer in number than the announced level conveys; 5. reservations tend to be clustered in certain services, departments and grades; 6. the process of achieving substantial representation is slower than indicated by the announced level and scope of reservation; 7. reservations are resented and resisted; resentment and resistance

is more articulate and focussed at the middle levels and at the promotion stage than at the initial recruitment stage. C. ASSESSING THE COSTS AND BENEFITS

Our crude assessment of the sheer quantitative impact of the reservation policy leads us to the question of its overall effects. What are the consequences of propelling members of backward groups into government posts by:a policy of preferential treatment? There has been no systematic study of these consequences and none can be attempted here. Here we can only try to formulate the issues and sort out some of the scattered and inconclusive but suggestive bits of evidence that are available (and indicate areas where there is no evidence at all). The

assessment that results is necessarily incomplete and tentative. It is meant to launch inquiry, not to conclude it. We may, conveniently, reformulate our general list of alleged costs and benefits with specific reference to compensatory discrimination in government employment (table 10).

TABLE 10 ALLEGED BENEFITS ANDCOSTS OF PREFERENCES INGOVERNMENT EMPLOYMENT REDISTRIBUTION

vs.

Reservations benefit directly only a small segment within these groups, and the benefits do not spread widely.

Reservations provide a direct flow of

valuable resources (e.g., salaries,

status, information, contacts) to the backward classes. REPRESENTATION

vs.

sympathetic consideration of their

needs in policy formulation and

classes. If there is effective representation, it is of some smaller

administration.

segment within these groups. vs.

Reservation in government jobs stimulates the educational and other accomplishments necessary to utilize them. INTEGRATION

competition, blunt the incentive to

acquire skills needed to compete ona

“merit” basis.

vs.

Holding of government jobs promotes pride, self-respect, and a sense of . achievement and personal efficacy that enable these groups to contribute to national development as willing partners.

ALIENATION

By emphasizing the separateness of these groups, preferences reduce their sense of common participation.

vs.

The presence of backward classes afforded by preferences induces others to accept them as fully entitled participants in Indian life.

INTEGRITY

OVERPROTECTION

Reservations provide artificial protection, remove the necessity for

By affording opportunities for constructive and prestigious ‘icipation, preferences promote feelings of belonging and loyalty among the employees (and others) from these groups. ACCEPTANCE

MISREPRESENTATION

Preferential treatment in government jobs creates new interests which diverge from those of the backward

The presence of backward classes in government jobs facilitates the

INCUBATION

DIVERSION

REJECTION

Preferences frustrate others by what

they view as unfair favoritism, educate

them to regard the beneficiaries as separate elements who enjoy their own facilities and have no claims on general public opportunities. vs.

MANIPULATION

The separate treatment involved in preference in government jobs

undermines their sense of dignity, selfsufficiency, and personal efficacy, and

aggravates the dependency of these groups by depriving them of potential ders.

MOBILIZATION

vs.

By imparting skills and providing incentive for cultivating talents, reserved posts increase the capacity of the beneficiary groups to undertake organized action on behalf of their collective interests. STIMULATION

By siphoning the most talented into

government service, reservations

deplete the pool of organizational skills available to backward communities and direct attention away from

collective action to individual advancement through reservations. vs.

positions, reserved posts stimulate their inclusion even in settings where it is not mandated by law.

measure to include them where it is not

required by government. VS.

Preferences utilize neglected and under-utilized talents and provide administrators sensitive to. problems of these groups. FAIRNESS

less-qualified persons and lower the

morale and efficiency of other

employees. VS.

disablements resulting from past deprivations of advantages and opportunities.

acquisition of skills, and by

encouraging acceptance, reservations

will render themselves unnecessary by

increasing the number of backward classes who can enter without preference.

UNFAIRNESS

Preferences place an unfair handicap on individuals who are deprived of opportunities they deserve on merit.

Preferences compensate for, and help

By increasing confidence through stimulating development of talent and

iNEFFICIENCY

Preferences lead to employment of

to offset, the accumulated

SELF-LIQUIDATION

SEDATION

By portraying the inclusion of the backward classes as a responsibility of government, reserved posts make others feel it is unnecessary to take any

By providing a model for the presence of backward classes in responsible

EFFICIENCY

ENERVATION

vs.

SELF-PERPETUATION

Reservations, by lowering incentives, encouraging alienation and rejection, creating vested interests in separate

and preferential treatment, lead to

perpetuation of reservations.

Note: This table is an adaptation of table “Development” there is specified here as “Efficiency.” The “Secularism vs. Communalism” item, omitted here because it does not raise issues that are distinctive to the employment area, is discussed in chap. 16.

108

Setting and Policies

Redistribution vs. Diversion; Representation vs. Misrepresentation;

Incubation vs. Overprotection

Reservations in government services clearly entail some substantial

redistribution of resources to the Scheduled Castes and Scheduled

Tribes—not only of salaries and perquisites, but of information, contacts, prestige, and patronage. Although the percentage increases in the representation of these groups are small, and although the fortunate beneficiaries of reserved seats remain a small portion of the total population of these groups, we should not overlook the significance of the absolute increase. In the early 1970s Aggarwal and Ashraf found that

every eighth Scheduled Caste household in their survey of Haryana

State had one member or more in a reserved job.5? Although the vast majority are not directly benefited, reserved jobs bring to the beneficiary groups a manyfold increase in a number of families liberated from circumscribing subservient roles, able to utilize expanding opportunities and support high educational attainments. Although such families are only a tiny fraction—and optimistic guess might be 6% ©—

of all Scheduled Caste families they provide the critical leaven from which effective leadership might emerge.® It is not clear how widely

these benefits have been spread among

the various parts of the

Scheduled Castes and Tribes population or how much they have entailed any spillover from the direct recipients to others in these

groups.® Scattered evidence suggests that the mobile minority utilizing the reservations in government employment is by no means evenly spread through the SC and ST. There is a severe clustering of op59. Aggarwal and Ashraf 1976: 125.

60. Isaacs (1965: 111) estimates that perhaps as many as 10% of the SC population is

“coming up,” largely through reserved posts. Our figures suggest that his estimate is probably high. By 1975, about 180,000 SC were in Class ITI or higher service with the central government. Let us assume (generously) about an equal number of Class III or higher in public sector enterprises and the same number in state services. That gives us about 500,000. If we make the very optimistic assumption that for every such person there is another person similarly situated in local government or private sector employment, we come toa total of | million. If we assume that each of these persons is the head

of a family of five, we come to a total of 5 over 6% of the entire SC population. Ifwe as having similar potential for mobility, come out with something close to Isaacs’s

milliop persons in such mobile families—just were willing to include Class IV employment we would more than double our figures and ten percent.

61. Cf. Zelliot’s (1969:4, 53) discussion of the way in which army service in the 19th

century provided the most important source of leadership for modernization and mobility among the Mahars. 62. In his study of the Agra Jatavs, Lynch (1969:108) suggests the absence of such spillover: “individual education and advancement do not necessarily rebound to

the benefit of the rest of the caste.”

Reservations in Government Employment

109

portunities in some of the larger and more advanced groups among them. In some cases the presence of Scheduled Castes and Scheduled Tribes does seem to produce more sympathetic administration and increased responsiveness to the needs of these groups. At least one would assume that their presence might inhibit the overt mistreatment of members of these groups by other officers. But few SC and ST officers have risen to policy-making positions. Whatever the benefits, they are at the cost of the channelling of the most talented and qualified members of these communities into government service, rendering them unable to contribute directly to community leadership (and perhaps at the cost of luring many into dead-end white-collar employment in which

their capacity to advance either themselves or their community remains limited). Knowledge that jobs are accessible to them through reservations acts as a stimulus to educational accomplishment. It is often asserted that feeling assured of jobs, members of these groups feel no need to apply themselves strenuously to education. Aggarwal, on the basis of interviews with successful Harijans, portrays preferential treatment in a much more positive light. He concludes that for those starting in illiterate families “the availability of privileges was the major facilitator.” Integration vs. Alienation; Acceptance vs. Rejection; Integrity vs. Manipulation There is clearly some integration and acceptance in the office setting.

The presence of Scheduled Castes and Scheduled Tribes on the job creates relations of interdependence with both fellow-officers and

clientele. To some extent this is offset by the alienation experienced by the SC and ST employees which they attribute to rejection on the part

of their fellow-officers.®

63. There is evidence of severe clustering in the distribution of educational benefits

(see above, chap. 3, §B, and below, chap. 13, §C) and presumably this is reflected in

differential capacity to utilize reservations in employment. 64. Aggarwal 1980:3. 65. A measure of SC and ST employees’ perceptions of resistance to their presence is provided by the number of complaints forwarded to the Commissioner for Scheduled Castes and Scheduled Tribes about service matters. In 1976 the Commissioner’s office received 17,000 complaints, more than ten times the total of five years earlier (RCSCST 1975-77: 1, 66; RCSCST 1973-74: 167). If we estimate that the total number of SC and

ST employees in all grades of government or public sector jobs was less than one million, we conclude that in a single year almost 2% of these made written representations of mistreatment to a single remote (and powerless) officer. This reveals not only much

(perhaps increasing) unhappiness on the job but also an absence of effective channels of redress.

110

Setting and Policies

The Scheduled Castes/Tribes employees are treated as trespassers, coming into the services at the expense of really deserving hands and recruited on the basis of “birth” and not “worth.” They thus start with an initial disadvantage in the sense that others are instinctively prejudiced against

them. They are harshly judged by their superiors. . . . [They] find it very difficult to establish a personal rapport with their bosses and colleagues. . . .&

As this plaint suggests, even the beneficiaries assign much of this

resentment to the preference policy per se. The antagonism that exists

toward the beneficiary groups seems to be accentuated by antagonism toward the policy that produces their presence. Educated untouchables, Aggarwal observes, are “taunted and slighted and constantly reminded of the unfairness of their privileges.”®7? So while preferences do get them into the office setting, they may augment other barriers to integration. Unfortunately, we have no studies which trace the interplay between antagonism to reserved-seat holders and the momentum of office activities. To assess the effect of preferences on integration and acceptance, one would want to compare situations in which Scheduled Castes and Tribes were present on a “merit” basis. This may be complicated by one of the by-products of reservation: the widespread overestimation of their scope and the consequent perception of all SC and ST as recipients of preferential treatment. Reservations can probably be credited with promoting feelings of integrity and self-respect associated with prestigious and responsible government employment. However, this is undermined by feelings of receiving separate and second-class treatment, by resentment of features of preference schemes seen as demeaning, and by aggravation ofa

sense of dependency.®

Mobilization vs. Enervation; Stimulation vs. Sedation

Reserved posts provide a powerful incentive for undergoing the sacrifices required to cultivate talents by obtaining education. The growing number of government servants includes many who devote their talents to group betterment through educational and social work. But government servants are not permitted to participate in political activity, so backward groups are deprived of the skills of those who might be the most effective organizers. By tendering inactive many of the most 66. Central Government Scheduled Welfare Association 1966: 11-12.

67. Aggarwal 1980: 10.

Castes and

Scheduled Tribes Employees

68. Isaacs 1965: 1 16ff. Cf. the “no response” data discussed in chap. 3, §D, above.

Reservations in Government Employment

111

talented and qualified, reserved seats deprive the community of leader-

ship and

reliant.@

render it more manipulable,

less assertive, and less self-

The record of private sector employment (and public sector employment before the introduction of reservations) undermines the stimulation argument. It seems hard to deny that reservations have been accompanied by lack of concern to include these groups in settings where such inclusion is not mandated. But against a long history of the absence

of such

concern,

it js difficult to conclude

absence is attributable to reservations.

that its current

Efficiency os. Inefficiency It is widely believed that compensatory discrimination in government employment is responsible for a decline in the efficiency of government operations. It is clear that reservations bring about the presence of government servants with less qualification according to existing methods of measuring it. We should hesitate to assume that there is a direct connection between the tested accomplishments of entrants and the effectiveness of government administration. In any event, the effect n “quality” of Scheduled Caste and Scheduled Tribe recruitment is overshadowed by rapid expansion of government service and a general lowering of standards. A Study Team of the Administrative Reforms Commission found that in the All-India non-technical services (Indian Administrative Service, Indian Forest Service, Indian Police Service,

Central Government Class I), the percentage of first-division entrants dropped from 42% in 1950-55 to 25% in 1960~65, while third-class degrees rose from 4% to 19%.7° Since this shift in educational profile

has been cited as evidence of the deleterious effects of reservation,”? it

deserves close examination. The falling off of first-division entrants can hardly be attributed to reservations.7? It reflects a general decline in the 69. In his study of the Jatavs of Agra, Lynch (1468: 107) concludes that reservations deprive the community of leadership from its best-qualified members. 70. Administrative Reforms Commission, Study Team 1969:5.

71. Ghurye (1969:440) cites a newspaper account of these figures as conus uutiy, “common knowledge” and infers that “‘{t]o a fair-minded person and to any patriotic Indian it must become clear that the proviso of the maintenance . . .of efficiency of administration of Article 335 requiring the Government to take the claims of the Scheduled Castes to be represented in the services has been set at naught.” 72. Except as the declining appeal of government service may reflect some apprehension that opportunities to rise in government service might be constricted by the ce of reservations. I know of no study which has attempted to isolate the contribution of this factor. I would surmise that such an effect has been produced by the expansive reservations for Backward Classes in several states, where others’ op-

112

Setting and Policies

appeal of government service to the academically most qualified.”* The rise in entrants with third-class degrees is attributable in part to the

intake of Scheduled Castes and Tribes under reservations. But the Study

Team’s own figures show that only a minority (160 out of 377) of the third-class degree holders selected in 1960—65 were Scheduled Caste and Tribes.74 By rearranging the Study Team’s data we can compare the educational performance of Scheduled Castes and Tribes with that of the other successful candidates (table 11). In estimating the impact of the reservations on the academic profile of the intake, it js important to isolate their net effect. Let us assume that reservations were responsible for the selection of all of the third-class and some large portion of the second-class degree holders among the Scheduled Castes and Scheduled Tribes. If there were no reservations,

they would be replaced by ‘‘others.” Presumably, it was the less qualified others who were eliminated by the reservation. So the academic profile of these eliminated others is bound to be lower than the others TABLE

11

Prior EDUCATIONAL PERFORMANCE OF SUCCESSFUL CANDIDATES FOR

ALL-INp1aA HicHer Non-TECHNICAL SERVICES, 1960—1965 Scheduled Castes and Tribes Type of Degree I

n 18

%

Others n

%

II 100)

178 160

50% 45%

5%

484

29.5%

Total

356

100%

1,637

100%

Source: Derived from Administrative Reforms Commission

936 217

57.2% 13.3%

1969:5, 6.

portunities for entrance and advancement are severely restricted. But it seems less likely that comparable discouragement would be produced by the modest reservations for SC and ST in central services. There is also a question of whether such discouragement is a bad thing: do talented individuals make more of a contribution to development outside government service? 73. In 1950-55, an average of | in every 3.5 holders of first-class degrees took the Combined Competitive Examination for higher central services: there was an average of3 firsts competing for each vacancy. In 1960-65, only | in every 13 first-class degree holders took the Examination, and there were only 1.8 First Classes competing for each vacancy (Administrative Reforms Commission, Study Team 1969:4). 74, By subtracting the figures given by the Study Team on p.6 from the figures it gives on p.5, we can determine that 160 of the 356 (44.9%) SC and ST recommended in 1960-65 held third-class degrees, as did 217 of the 1,637 (13.3%) of the other candidates.

Reservations in Government Employment.

113

who were successful in spite of the reservation. This successful group included over 13% third-class degrees. We must assume that the next two or three hundred down the list would include an even higher percentage of third-class degrees. Thus the net difference between the two groups is overstated by our table (for it compares the SC and ST only with those competitors who weathered the “bumping” caused by the reservations). Thus it overstates the effect of the reservation on the

academic profile of the intake.

The translation of lower academic accomplishment into inefficiency in administration is difficult to trace. It is not clear how well academic performance correlates with administrative talent. Nor is it clear that differences in the level of such talents are directly reflected in efficiency

or inefficiency of administration. In the light of the many forces which affect administrative inefficiency, it seems appropriate to confront assertions about the effect of reservations with some skepticism, if only to guard against the widespread tendency to attribute any inefficiency in

governmental operations to the presence of Scheduled Castes.’5 Another “efficiency” argument is that reservations injure the morale of other workers, lowering the general level of administration. Low morale might stem from (1) inferior performance by SC and ST (which in turn might be caused by rejection—i.e., abrasive social relations, absence

of informal

on-the-job

training,

etc.);

(2) fear for career

prospects—especially where reservations in promotions give rise to apprehension

that one’s co-worker (and competitor) of today is des-

tined to become undeservedly one’s superior tomorrow;7¢ (3) displeasure at the enforced contact with these groups. Since there are plenty of government offices with no Scheduled Castes or Scheduled Tribes, and many where no reservations in promotions are in effect, a comparative study of effects of these arrangements

on morale is conceivable, but has not been done. For the moment we may accept that reservations entail some cost in morale—but it is not clear how much is due to the reservation device per se and how much is

just rejection of Scheduled Castes and Scheduled Tribes and would be there even if they attained their posts on “merit” alone. The arguments about efficiency lead back to the “‘siphoning effect”

that was mentioned earlier77—a phenomenon with implications for the values that we have summed up under the headings of acceptance

and integrity. Reservations, particularly those for high-echelon posts, 75. For some example, see Isaacs 1965: 133.

76. Thus Kirpal and Kelkar (1976:8) convey widespread reports that “(c]aste Hindus . . .tend to withhold information at the time of training their SC/ST subordinates, who, they fear, may supercede them by virtue of seats reserved for promotion.” 77. See text at n.26 above.

114

Setting and Policies

pull in candidates with lower performance on various indicators. The pool of candidates from the beneficiary groups who have passed the threshold educational qualifications to aspire beyond the clerk’s grade is rather small. Once the threshold is passed, members of these groups tend to shoot for the very top. Thus in 1961 Scheduled Castes candidates made up 8.9% of applicants for I.A.S. and Class I posts, but only 4.5% of applicants for Class II posts. Among Scheduled Tribes, the siphoning is even more pronounced: they made up 1.5% of the appli-

cants for I.A.S., etc., but only 0.06% of applicants for Class II posts.78 Reservations place these candidates in positions alongside peers with higher performance on various indicators (examination scores, etc.). In part, the higher scores of others may reflect cultural advantages which

are irrelevant to the business at hand; in part, the lower scores of

beneficiaries may reflect a remediable lack of polish and experience rather than lack of native ability. But with all due allowance for the irrelevance of the indicators, performance on them probably reveals something about ability to deploy one’s talents. And with all due allowance for “diamond in the rough” qualities, it is clear that not everyone can surmount greater challenges while still suffering the effects of earlier disadvantages. Reservations propel SC and ST candidates into settings in which their ability to perform competently and comfortably is placed under considerable strain—strain compounded by the hostility and resentment of those around them. If this leads to some instance of heroic triumph over challenge, it must in many others be damaging both to performance on the job and personal well-being.

The forced-draft aspect of reservations at the top may carry heavy personal costs for beneficiaries as well as for others. Fairness vs. Unfairness

The fairness argument is of course based on the principle of historical compensation and more immediately on the notion that members of these groups on the whole have less opportunity to cultivate their talents so as to show up well in recruitment testing. The charge of unfairness is based on the notion that it is unfair to individuals that their life chances be impaired to provide such compensation. There is no reason to suppose that the individuals who are “bumped” from valued opportunities were disproportionately benefitted by invidious discrimination in the past. If current ability to perform meritoriously is thought a product of accumulated past advantage, those who are bumped seem to enjoy less rather than more of this advantage. And since reserved

places are distributed among those eligible in order of merit, the ben78. Planning Commission 1965: 266 ff.

Reservations in Government Employment

115,

eficiaries will be those who are least disaavantaged among the eligible

up.

e eause reservation policies tend to vary from department to department in their incidence and in the effectiveness with which they are implemented, there is some clustering—so that the “unfair competition” impinges heavily on some government servants while others are virtually untouched. We may distinguish two separate claims of unfairness: (1) that which rejects the principle of collective recompense for historical disabilities; and (2) that which objects to the adminstration of the principle so that the burden is borne by a few who are no more responsible for the historical deprivations that underlie it than the many who escape. The administration of compensatory discrimination measures does seem to involve considerable “unfairness” of the latter kind. If some is inseparable from compensatory discrimination, it is clear that more could be done to distribute the burden on nonbeneficiaries more widely and more evenly. Self-liquidation vs. Self-perpetuation Reservations in government posts are not self-perpetuating in a literal sense (as are reservations of legislative seats). They set up expectations of and pressures for continuation and expansion (“‘vested interests”):—but in the tield of Government service the pressures seem more directed to

enhanced benefits for existing recipients (e.g., extension of reservation to promotions) than to increasing the beneficiaries. And, in an important sense, reservations in government service are self-liquidating: many reservations operate as guaranteed minima, not as fixed quotas.”? Thus are potentially self-liquidating to the extent that there is adequate incubation of talent. Thus, if programs for delivering education to these groups were fully effective, the reserved posts would become largely redundant.

Considering the readiness with which many Indian intellectuals, including some eminent social scientists, have passed severe judgments on compensatory discrimination, it is surprising how little systematic study of its presumed effects we have. Our cursory sketch suggests several considerations that seem crucial if the heightened interest in these questions is to equip us to arrive at a credible assessment. First, there is the necessity of isolating the net effect of compensatory discrimination from effects of the presence (apart from compensatory discrimination) of Scheduled Castes and Tribes. For example, to the extent that 79. See chap. 13, §B, below.

116

Setting and Policies

rejection,

alienation,

lowered

morale,

etc., are associated with the

presence of Scheduled Castes and Tribes, apart from preferential treatment, they cannot be accounted as costs of compensatory discrimination. Reservations may provide the occasion rather than the source of the negative effects. Second, there is the necessity of considering not only the design of the scheme in assessing its costs and benefits, but also the level of performance at which it is operating. As the government employment area suggests, programs of compensatory discrimination may operate at a

level of performance considerably below their announced promise. One consequence is to project to interested (and resentful) publics an image of ubiquitous and unqualified preference with immense power to affect them adversely, while to the beneficiaries it appears a program of narrow and contingent preference with limited power to benefit them. The output of those values we have labelled costs and benefits may be augmented or diminished by changing levels of performance as well as by changes in authoritative policy. Third, perceptions of costs and effects may have a powerful influence quite apart from the presence of objective indicators of such costs, etc. Compensatory discrimination policies are symbolic as well as administrative realities. Their consequences are not reducible to their perceived consequences, but the perceived consequences of these policies are important components of any assessment. Notwithstanding all the problems of ascertaining their performance and evaluating their achievements, the compensatory discrimination policies mark the inclusion of the beneficiary groups into the visible public life of Indian society to a degree that would have been unimaginable a few decades ago.® Although the beneficiary groups still are significantly lower on virtually all measures, they are no longer completely excluded from the mainstream politically, occupationally, or educationally. Yet it is evident that actual performance of these policies falls far short of announced aspirations. We have seen that reservations go unfilled in good measure; most of those receiving educational benefits drop out early; protective legislation is unenforced and circumvented; allotted funds go unspent; and programs may give substitute services instead of additional ones. The general pattern is that normative com80. Our preliminary assessment of performance has been concerned almost entirely with the preferences for Scheduled Castes and Scheduled Tribes. For the OBC, there is less data available, there is more variation state by state, and the goals are less clear, soit

is difficult to make any generalizations about OBC preferences. But one important by-product should be noted—the operation of OBC preferences has done much to obscure and confound public understanding of the preferences for SC and ST.

Reservations in Government Employment

117

mitments outrun institutional capacities (and political will) to imple-

ment them. Weak and unaggressive implementation combined with a concentration of attention on “higher echelon” benefits (i.e., those which require some degree of previous success) mean that utilization is by those among the potential beneficiaries better supplied with other advantages. Thus the total redistributive effect is smaller than is projected and tends to be concentrated disproportionately in the upper layers of the beneficiary groups. These programs, with their pervasive gap between promise and performance, provide the setting within which we shall examine the work

of the courts as critics, checks, energizers, definers, and legitimators of

this complex of policies. But first we must take up the process by which the beneficiaries of these policies have been selected and identified.

Part Two IDENTIFYING THE BENEFICIARIES

Analytically, the first question that arises about India’s policy of compensatory discrimination is “who are the beneficiaries?” But historically the process is more complex: the identification of some groups as requiring and deserving special treatment gives rise to a concept of compensatory treatment, and this in turn produces controversies about

those who should be included within its scope. Even after compensatory preference is firmly established, who is to benefit continues to be a central question in designing and administering programs—and crucially important for the continuing justification of these programs to

others. Public and professional discourse about the identity of the beneficiaries is thus an integral part of the operation of these programs. These disputes also reveal how (some) Indians see their society and how they relate their aspirations for it to their perceptions of its present

state. We can observe how those perceptions and aspirations are filtered through the institutional style of the Indian legal culture. Chapters 5 to 8 address the general question of the selection of the beneficiary groups. Chapter 5 takes up the groups designated earliest and at the national level. Section A recounts the emergence of the notion that there were a distinct layer of “untouchables” at the bottom of Indian society, the controversies about drawing the line, resolved by the formation of the Scheduled Castes category in the 1930s. Section B

describes the institutionalization and stabilization of that category in

post-Independence India. Section C (of chapter 5) sketches the comparable development of the Scheduled Tribe category.

120

Identifying the Beneficiaries

Chapters 6 to 8 take up the more recently controverted questions of the other beneficiaries of preferential treatment, the groups that have

come to be called the “Other Backward Classes.” Chapter 6 parallels

the account in Chapter 5: it traces governmental action and public controversy in reference to the designation of the Other Backward Classes, a story that is complicated by great diversity among the states and by ‘tensions between central and state governments. There is another dimension to the selection of Other Backward Classes: it is an area that has been the scene of steady and significant judicial intervention; it is an area permeated by resort to the courts and by ideas worked out in the courts. Chapters 7 and 8 analyze the judicial response to the Other Backward Classes. Chapter 7 takes up the controverted question of what are the social units that government may designate as backward classes (here the controversy over the “caste basis” of selection). Chapter 8 takes up the question of the standards and the methods by which the backwardness of these groups is to be measured.

Chapter 9 moves from the question of which groups to the question of

who are members of those groups. This includes disputes concerning both individuals whose personal history makes their membership problematic and groups (e.g., neo-Buddhists) whose inclusion in a larger category is controverted. These controversies reproduce in a more precise form and on a more personal scale some of the conflicts discussed in the earlier chapters. This chapter analyzes the jurisprudence of personal and group identity which the courts have developed in their encounter with these controversies. Chapter 10 summarizes the perplexities of governmental superimposition of official categories of beneficiaries on the complex processes of group and personal identity. It concludes with a discussion of the implications of these perplexities for the role of courts in implementing the compensatory discrimination policy.

a

The Designation of the Scheduled Castes and Scheduled Tribes THe

CONSTITUTION

DOES

not

contemplate

_reserva-

tions and special treatment as a general principle of operation. Such treatment is authorized only for the limited purpose of advancing the

backward. Who, then, are the legitimate recipients of sich preferential

treatment? The Constitution permits preferences for three categories of groups: (a) Scheduled Castes!—i.e., “untouchables”, (b) Sheduled Tribes;? and (c) other (socially and educationally) backward classes.3 The term “backward classes” is commonly used in two senses: (1) as a generic term including the Scheduled Castes and Scheduled Tribes as well as the so-called Other Backward Classes;* or (2) as a designation of those backward groups not included in either of the first two categories (i.e., as equivalent to “Other Backward Classes.”’)5 The Constitution itself does not define these groups, nor does it provide detailed standards by which they may be determined.® In the case of the Scheduled Castes and Tribes, it does prescribe an agency and a method for designating them. Not only are Backward Classes left undefined in the Constitution, but no such method or agency for their determination is provided. Before taking up (chaps. 6, 7, and 8) the 1. Arts.1 5, 16, 46, 341, 335. Cf. Arts. 330 and 332. 2. Arts.1. 5, 16, 335, 342. Art 244, Vth and VIth Schedules. Cf. Arts. 330 and 332. 3. Arts.1 5, 16, 46.

4. See Art.16(4). In General Manager o. Rangachari, A.1.R. 1962S.C. 36, it was held that

“Backward Classes” in Art. 16(4) includes the Scheduled Castes and Scheduled Tribes. 5. See Art. 15(4), as well as Art. 340 and Art. 164 where it is apparently used in this restricted sense. Cf. Sudhir Kumar Ghose v. Ank Kameya, XXVII Cuttack Law Times

(Short Notes) 5 (1961). (Election set aside on ground that candidate was a member ofa

Backward Class, not ofa Scheduled Caste.)

6. Section (24) of Art. 366 defines Scheduled Castes as those so designated under Art. 341. §(25) of Art. fe defines Scheduled Tribes as those designated under Art. 342. No definition of “! Classes” appears at any point in the Constitution.

122

Identifying the Beneficiaries

more complex issues that have arisen in the designation of the Backward Classes, I shall trace the problems of identifying the Scheduled Castes and Scheduled Tribes. The Scheduled Castes and Scheduled Tribes constitute the bulk of those receiving preferential treatment, and they receive a larger quantum of preferences than Other Backward Classes. The Scheduled Caste and Tribe categories are intended to comprise those who were at the bottom or the margins of the Indian social order— those groups who because of their low ritual status in the traditional Hindu hierarchy

or their spatial and cultural isolation were subjected to imposition of disabilities and lack of opportunity. The Constitution prescribes that the Scheduled Castes and Tribes shall be designated by Presidential Order in the first instance, with subsequent modification only by Act of Parliament.” It further provides that the President shall appoint a special officer to investigate all matters relating to the safeguards for the Scheduled Castes and Tribes and report to Parliament on their working.® Accordingly, the office of the Commissioner of Scheduled Castes and Scheduled Tribes was established in 1950 as a continuing body with coordinating and reporting but not administrative functions. A. THE INVENTION OF THE SCHEDULED CASTES “Scheduled

Castes”

is the most

recent of a long line of official

euphemisms for “untouchables.” The Scheduled Caste category is intended to comprise those groups isolated and disadvantaged by their “untouchability” —i.e., their low status in the traditional Hindu caste

hierarchy which exposed them to invidious treatment, severe disabilities, and deprivation of economic, social, cultural, and political

opportunities. In the early years of the century the “depressed classes” (as they were then called) became an important focus of concern among

reformers. After 1901, fears of diminished Hindu majorities and pro-

posals for special legislative representation for these classes propelled “untouchability” from the realm of philanthropy into the political arena. As concern about untouchables grew and political demands were made in their behalf, there was sharp disagreement about the number of persons who belonged in this category. From 1908 on, it was claimed by numerous writers that untouchables

numbered 50 millions or more—about 24% of the Hindu population or just under 16% of the total population of India according to the 1911 Census.? Although Madras compiled a list of Depressed Classes during 7. Arts. 341, 342. 8. Art. 338. 9. Risley (1908) counted 50.6 million Hindu untouchables (in addition to 17.2 million

animists). This estimate was omitted from the 1915 edition of his book, as noted by

Designation of Scheduled Castes and Tribes

123

the 1920s,?° the Government of India refrained from any official

classification (until 1936) on the ground that it would be unfair to stigmatize these groups by official acknowledgement of their low

status.1? But various governmental bodies attempted to assess their numbers. The first official estimates in 1917 and 1918 arrived at the

figure of about 30 millions for British India.1? The Franchise (Southborough) Committee, which was established to advise on the franchise

under the 1919 reforms, found 42.2 millions in British India. Since there were presumably one-quarter to one-third as many more in the

princely states, this did not depart far from the 50 million figure. The

1921 census and the Reforms Enquiry (Muddiman) Committee of 1924

arrived at comparable figures.4* However, in 1928, official estimates reverted to the lower range: two officials identified the number of untouchables as 28.5 millions and 29 millions,45 and the Hartog ComWeitbrecht-Stanton 1920:175. Cf. H. Madras 1909: 1022 (‘‘the outcastes of Hindu society form all over India a distinct section of the population, numbering about 50

millions”); Holderness 1911: 102 (50 to 60 millions); S.N. Singh 1913:376 (“over 50,000,000”); Fisher 1919:97 (53 million). Somewhat higher estimates are given by Ketkar 191]:78 (one-fourth of Indian population) and Brown 1914:548 (one third of Hindu population). A lower estimate was given by Baines 1912 (34.8 million).

10. They were scheduled by notification under the Madras Local Boards Act of 1920 (Franchise Committce 1932:1, 112). 11. “Though officers of Government have from time to time attempted to estimate the total number of the depressed classes in the country, the Government of India have consistently avoided themselves making any precise classification of the groups forming what are called the depressed classes if only on the ground that owing to the social

disabilities to which members of the depressed classes are exposed it would be in the

highest degree undersirable that any official authorization might appear to extend to such classification. The fluidity of social distinctions and the efforts of the classes lowest

in the scale, aided by social reformers, to improve their status make it the more desirable

that Government should abstain from doing any thing which would tend to give rigidity

to these distinctions ” (Indian Statutory Commission 1930: V, 1341). 12. The estimate of Sir Henry Sharp, Educational Commissioner in 1917, was 31.5 millions (Indian Statutory Commission 1930: V, 1350 ff.) In response to the Govern-

ment of India’s inquiry, the provincial governments (except Assam and Punjab) provided lists in 1917-18 of depressed classes whose total population was 27.4 millions, (Indian Statutory Commission

1930: V, 1341 ff). Weitbrecht-Stanton

1920 comes to a

similar figure. 13. Indian Franchise Committee 1919: 1. The figures appear in the appendices under each province. The figures are compiled and discussed, id. 367, in an appended despatch from the Government of India, Home Department, signed by the Chelmsford

Committee.

14. Census 1921:1, Pt. 1, p.193; Reforms Enquiry Committee 1925:56, The Commit-

tee noted that its figures were “a rough estimate of the minimum numbers given in the last census report ”(id.55). 15. In the Indian Legislative Assembly on February 23, 1928, Mr. G. S. Bajpai,

Secretary in the Department of Education, Health and Lands, stated that according to

his calculation based upon available material the depressed or untouchable classes in

124

Identifying the Beneficiaries

mittee set the number at 29.76 millions.1¢ A new round of counting had begun with the impending of further constitutional changes. In 1929, the Indian Central (Nair) Committee, set up to advise the Simon Commission, arrived at a figure of 44.5

millons,!” and the Simon Commission itself estimated the number at

43.6 millions. *® (See table 12.)

The Indian Franchise (Lothian) Committee, established to advise on

the franchise under the new reforms, received conflicting estimates from a number of quarters. The 1931 Census Commissioner estimated untouchables at 31.2 millions excluding Bengal; the Provincial Governments supplied somewhat lower figures, totalling 37.45 millions.2° The Provincial Committees,

TABLE

set up to advise the Franchise Committee,

12

PERCENTAGE OF UNTOUCHABLES IN BritIsH INDIA (According to Indian Statutory [Simon] Commission, 1930) Number of

Province

Untouchables

MADRAS: BOMBAY

6.5 15

BENGAL

UTTAR PRADESH PUNJAB BIHAR& ORISSA

CENTRAL PROVINCES

ASSAM

Torats

(Millions)

% of

% of

Hindu Pop.

Total Pop.

18 if]

15.5 8.0

15"

57

24.5

5.0°

20

145

12.0° 28

3.3 1.0

43.6

31 42

33

24

2842

26.5 13.5 24.5

13.0

19

Source: Indian Statutory Commission 1930a : 40.

Note: 2 Noted as “‘subject to ... warning”.

the Governor's Provinces (excluding Burma and Assam) numbered 28¥2 millions. Later that year, in Parliament on the 23rd of April the Secretary of State for India estimated the total number of untouchables in British India at 29 millions “on information provided in Provincial Educational Reports” (Indian Statutory Commission

1930:218).

16. The Hartog figures were based on the figures given to the Legislative Assembly (Indian Statutory Commission [Hartog Committee] 1929:217-18). 17. Indian Central Committee 1929: 44, 86.

18. Indian Statutory Commission 1930:1, 40. 19. Hutton’s tentative figures to Indian Franchise Committee 1932:1, 119.

20. Indian Franchise Committee 1932:1, 119. The Memoranda from the provincial governments are in vol. ITI.

Designation of Scheduled Castes and Tribes

125

produced figures that totalled less than half of the then-current estimates (18.62 millions).27 The dissenting members of the Franchise Committee minimized the number even further to 14.42 millions.?2 The Franchise Committee itself was unable to arrive at a firm figure.?3 The various estimates are presented in table 13. As table 13 indicates, there was little dispute about who were the untouchables in southern and western India— Madras and Bombay

Presidencies and the Central Provinces. The dispute was about north-

ern and easter India. Which groups there ought to be treated on a par with the untouchables of the south and west? This, in turn, depended on

how the depressed classes were to be defined — whether strictly in terms of specific indicia of ritual and social exclusion or more broadly in terms of low status. The earlier investigations had tended to proceed on the assumption that untouchables throughout India were set off by some uniform and distinctive pattern of practices and that a test could be devised that would identify them on an all-India basis. Thus, the Government of India’s 1919 Despatch, criticizing the inadequacy of the Southborough Committee recommendations for representation for the Depressed Classes, observes that, though they are defined in varying terms, [they] are broadly speaking all the same kind of people. Except for the differences in the rigidity of their exclusion they are all more or less in the position of the Madras Panchamas, definitely outside that part of the Hindu community which is allowed access to their temples.

But this assumption proved difficult to apply in northern and eastern

India. Touch and distance pollution extended throughout the caste system in the south, while in the north touch and distance pollution were not expressed uniformly enough. The Simon Commission, a decade later, conceded that in Bengal, the United Provinces, and Bihar

and Orissa, provinces

there was a wide margin of possible error.

In these

the connection between theoretical untouchability and practical disabilities is less close, and a special investigation might show that the number of those

who are denied equal rights in the matter of schools, water and the like is less

than the total given for the depressed classes in these areas.?5

21. Indian Franchise Committee 1932:1, 119. The Provincial Franchise Committees’

reports are contained in Vol. III.

22. Minutes of dissent of Messrs. S. B. Tainbe, C.Y. Chintamani, and R. R. Bakhale.

Indian Franchise Committee 1932:1, 219.

23. Indian Franchise Committee 1932:1, 118.

24. Fifth Despatch in Indian Constitutional Reforms 1919: 366, 367. 25. Indian Statutory Commission 1930:I, 41.

dt ueipuy souey) uo “16161 SUL)

Teel

3f 32

ZL

=F?

Stl

ee

9

0%

LE

0%

ott

be

08

cet

wt

ort

NIeg,,

e

4

‘S21ON,

Sle

KVSSY

SIVLOL

HSadVad

avaA

aounos

SVUGVN

AVanog

vONae

YV1LA

avinad

VSSI4O 29 UVHIG

SAOINIAOUd 1VHLNIO

La

al

ut

OE

1%

we

8

69

1

Vv

F

2 5

ze FEE

3

g2

BILLET cL6t

a

rt

LER

Hy Bo

vo

06

LG

u

ve

ay

2

6161

Fg

nz

a

(261

@3g

P2

&

3

68 3

a3 é

2 3

z 3 za

FF i?

ro

te

oe

"6d ‘S292, 1261 Poutunuo;) astysuesy ‘OF : PUEGL UcIssIMUC:) LUoIMeIg URIPUT 25 'g1Z : 6261 (2aNTWIWO:) S296 : B61 2nIWUIE,) JeNIU2;) UeIPUT : G ‘EIT C-UGEI - O61 UOISSINUO:) LOIMeIg UEIPUT :\" ;s20INOg

“suoypius ut aue saund

“(wont 96°) stueY Py BuIpnjoxg jod 10U se poqussaq I. “saundy ou 2ae¥ Auofew !quour wo saunty S21 9pun 40 *voRednsaqut spun ‘aneruar “areulrxasdde se a0un0s ut PION,

197

ws

OBS

uel

vent i

BRT

a

[e6L—ObOLGZEL

GF FF

#233

37 Ze

aGOR?& 53

g 3

4

F

Teel

35

2 2

SF FE

Ef 2p

g

£

Se 38

a2

g 3? Fog

gE 3

2 ¢¢

92 72

2Z

33

>

oF

35

2

8%

vs

Ot

we

908

se

Wwe

Wout

ptt

soe

ott

2

Toe

H

sot

1

ot

ot

os

i

Yeol

#32 93

I

23 £2302 4

z

2

Gi

35

x

rr)

{

pe 2

i é

ob

3s

Se

¢GR

Gp

68 §

Gp

sh 5

o$

5. 32

7b53

3

gge °

33

os

23

3

a5

1461-2161 ‘Sata0g TvI014s© SNortvA Ad VIGN] HSLLG NI NOLLVINdOg JTEVHONOLN() JO SALVNLISY €l ATV

Designation of Scheduled Castes and Tribes

127

The very different nature of the matter in the north was brought out nicely in the testimony before the Indian Franchise Commission of Mr. G. S. Pal, representing the United Provinces Hindu Backward Classes

League, which had submitted a list of 115 castes (estimated to make up

over 60% of the population) for inclusion as Depressed Classes in the United Provinces.”© Dr. Ambedkar. My point is this. Is this a list of what might be called backward communities, economically backward and educationally backward? Or is

this a list of pollution? Is tionally and educationally

castes which are called untouchables, whose touch causes this a list of castes which are touchable but which are educaeconomically backward or which are untouchable and also and economically backward?

Mr. Pal. 1 respectfully submit that this is a list of all the castes which in the opinion of the League are depressed classes; untouchability in the sense in which it is understood in Madras or Bombay does not exist here. Bhangi is the only caste that is universally untouchable, pollution by whose touch is regarded as something very undesirable. Here untouchability exists, of course, but in a very mild form or exists in some form or another in all the

castes I have enumerated.

Sir Sunder Singh Majitkia. How would you define depressed class people? Mr. Pal. Not in the sense of the definition in Bombay and Madras but apparently here the Hindu class is divided into high and low castes. The problem is very acute, as acute as untouchability is in Bombay and Madras.

I apply three tests, the first is we are socially despised; then there is

economic backwardness; and educational bankruptcy.?”

In the most elaborate attempt to specify criteria by which to identify

untouchable groups, J.H. Hutton, the 1931 Census Commissioner, pro-

posed a series of tests which revolved around the incidence of disabilities:

1. Whether the caste or class in question can be served by Brahmans or

not.

2. Whether

the caste or class in question can be served by the barbers,

water-carriers, tailors, etc., who serve the caste Hindus.

3. Whether the caste in question pollutes a high-caste Hindu by contact or proximity.

26. The list and estimate are at Indian Franchise Committee 1932: IV, 825-26.

27. Id., 827-28, Cf. another definition of “untouchables” in the U.P. setting, sug-

gested in a memorandum of the United Provinces Adi-Hindu (Depressed Classes) Association: refusal of caste Hindus to take food or water which they touched; refusal of entry into eating-houses; refusal of accommodation in dharamsalas (id., 833). This

memorandum contained a list of 64 castes said to meet these tests.

128 Identifying the Beneficiaries 4. Whether the caste or class in question is one from whose hands a caste

Hindu can take water.

5. Whether the caste or class in question is debarred from using public conveniences, such as roads, ferries, wells or schools.

6. Whether the caste or class in question is debarred from the use of Hindu temples. 7. Whether in ordinary social intercourse a well-educated member of the caste or class in question will be treated as an equal by the high-caste men of

the same educational qualifications. 8. Whether the caste or class in question is merely depressed on account of its own ignorance, illiteracy or poverty and but for that, would be subject tono social disability.

9. Whether it is depressed on account of the occupation followed, and whether but for that occupation it would be subject to no social disability.2*

These tests are not altogether free of difficulty. They involve terms

like “clean Brahmin” and “high-caste Hindu” which have merely a

local and comparative reference. And whether a caste is “‘polluting” or

“debarred” from public facilities or temples may admit of no un-

equivocal answer, since pollution and debarment may exist in grades or

degrees. Some may be forbidden to approach a temple, while others are only excluded from the inner sanctuary. Debarment from wells might

imply use of adequate but separate water supply or onerous exclusion

from any decent water supply. Tests 7 and 8, as Lelah Dushkin points

out, “seem

a bit academic, since the [Census] Superintendents were

often in no position to find out whether they applied or not. . . .”29 Whether a caste “‘can be served” is not necessarily the same as whether

it is habitually served unless one assumes correspondence between local learning on the matter and local practice that does not invariably

obtain.

However, the tests clearly point to the incidence of disabilities as the

crucial test. “[I]t is at least clear that occupation, ‘ignorance, illiteracy

or poverty’ or any other such forms of ‘backwardness’ were not intended

to enter into the definition process.””>° Hutton suggested:

From the point of view of the State the important test is the right to use

public conveniences—roads, wells and schools—and if this be taken as the

primary test, religious disabilities and the social disabilities involved by them may be regarded as contributory only. Some importance must be attached to them, since obviously if the general public regards the persons of cer-

28. Census of India, 1931, Vol.l, Part 1, App. 1, p.472. (Reprinted in Hutton 1961: 194.) 29. Dushkin 1957:88. 30. Phd.

Designation of Scheduled Castes and Tribes

129

tain groups as so distasteful that concerted action is resorted to in order to

keep them away, persons of those groups do suffer a serious disability.>*

But the “right to use public conveniences” might be one of the most difficult and fruitless tests to apply, since roads were generally open to all, practices regarding schools and wells were subject to much local variation and came in a number of gradations, and finally, because

practice might diverge sharply from “rights.”

The existence of multiple (and ambiguous)

criteria permitted re-

gional variation by giving the Census superintendents scope for discretion. Each provincial superintendent was instructed to draw up his own list, taking these tests into account, “and to reckon as depressed only

those castes who

handicap scheme.”

definitely suffer. from

serious social and political

on account of their degraded position in the Hindu social The final reckoning, then, did not apply Hutton’s tests pure

and simple. Excluded were some castes which were theoretically untouchable but who, by virtue of prosperity and vigorous organization, “have built up for themselves a strong position which ebviates the need of any special measures for their social, political or religious protection.”°3 Conversely, some groups were included even though theoretically they were not untouchables.** The total of Depressed

Classes reported by the census was 39.1 millions in British India and

11.1 millions in the States and Agencies—a total of 50.2 millions.35 The insufficiency of any single test to identify the untouchables provoked a variety of responses on the Franchise (Lothian) Committee.

Dr. Ambedkar, the leading untouchable spokesman, points out that the aim is to demarcate those who suffer from the contempt and aversion of higher caste Hindus: “It is a fatal mistake to suppose that differences in tests of untouchability indicate differences in the conditions of the untouchables.” The crucial common element is the odium and avoidance of the higher castes. Once this element is present “the whole

of the class of untouchables so ascertained must be taken into account

for the purpose of representation without any further distinction between rich and poor, advanced and backward, educated and uneduca-

ted....”37 Both of these points were rejected by the dissenters, S. B. 31. Hutton 1961: 195. 33. Ibid., 197.

32. Ibid., 196-97.

34. Dushkin (1957:91-95) observes that in Orissa, Bihar, and Bengal local census

superintendents utilized “ad hoc’ criteria, in addition to those given by Hutton, to

include Hinduized tribals, nomadic and criminal tribes. Their numbers were negligible

except in Bengal where these “ad hoc” castes made up over a quarter of the population

on the final list. 35. Hutton 1961: 199. 36. Indian Franchise Committee 1932:1, 203.

37. Id., 210.

130

Identifying the Beneficiaries

Tambe, C. Y. Chintamani, and R. R. Bakhale. They find that outside of Madras, Bombay, and the Central provinces, untouchables “do not

form a distinct and separate element in the population.” In the other provinces “untouchability . . .is the adjunct not ofa person but of the occupation he occupies.”’>® The majority of the Committee, adhering to denial of access to temples and causing pollution by touch or approach

as “‘generally accepted tests of untouchability”® reiterated its conviction that the depressed classes should include only untouchables and

not “those Hindus who are only economically poor and in other ways backward but are not regarded as untouchables.”*° But in the face of conflicting estimates it confessed its inability to reach any conclusions

about the untouchable population in the north and east.*?

The groups in this category were finally listed (i.e. scheduled) in 1936 for purposes of giving effect to the provisions for special electoral

representation in the Government of India Act, 1935.4? The list finally

promulgated did not correspond exactly with any of the earlier estimates: it hewed closely to the low figures for Assam, the high figures for the United Provinces and Bihar and Orissa, and took a middle course in

Bengal.*? In the event, the list reflected definitions of untouchability

with an admixture of economic and educational tests and considerations of local politics. Lelah Dushkin observes that

the authorities who made up the final list made quite a departure from the original strict criterion of untouchability . . .many castes listed in the North would not have been included in the South . However, whatever the criteria

were, all of the castes included in the final list were socially the lowest Hindu castes in their particular provinces and they were listed partly, at least, on

that basis.*5

The

list contained

over 40 million persons. The

1941

Census en-

umerated 48.8 million members of the listed groups (very close to the 50 million figure current at the inception of the debate a quarter-century

earlier). They constituted over 19% of the Hindu population and 12.6% of the total population of undivided India. (See table 14.) 38. Id., 220 40. Id., 109. 41.

E.g., for U.P.

39. Id., 110. the Committee

:

had estimates of 0.58 million, 6.8 million, 9.2

million, and 11.2 million (the latter accompanied by 1.4 million “untouchable” but depressed (id., 114-16). 42. The Government of India (Scheduled Castes) Order, 1936.

43. For an analysis of the discrepancies between Hutton’s list and the Order’s for

north and eastern India, see M. Singh 1947. 44. M. Singh 1947 gives an estimate of the groups included in the Order which were not

“untouchable”

according

to the Census:U.P.,

(3.2%); Bengal, 2,386,515 (26.2%).

75,936

(0.7%);

45. Dushkin 1957:95.

Bihar,

146,397

Designation of Scheduled Castes and Tribes TABLE

14

131

.

PopuLaTION OF SCHEDULED Castes IN 1941 SC Pop. (mils.)

BRITISH INDIA

39.92

BOMBAY

1.855

MADRAS

BENGAL UTTAR PRADESH PUNJAB

4.340

ORISSA

1,238

NWFP.

SIND

AJMER-MERWARA

TOTALS

STATES AND AGENCIES. COCHIN HYDERABAD

7.378 11.717 1,248 5.578 3.051 0.676

Hindu Pop.

% Total Pop.

20.90

8.068

BIHAR

CENTRAL. PROVINCES ASSAM

|%

13.49

18.85

16.35

29.44 25.57 16.53

12.23 21.39 4.38

11.20

16.37 18.12

16.74 23.57 16.04

8.89

11.49 14.18

12.37 41,34 6.62

0.191

15.58

4.22

39.80

20.87

13.45

8.892

13.86

9.78

0.141 2.928

15.73 21.99

9.92 17.92

MYSORE

1.405

21.03

19.16

TRAVANCORE,

0.395

11.15

6.50

19.15

12.62

RAJPUTANA

Toraxs (throughout

India?

Source:

Notes:

B.

tee

48.81

see

see

Census of India, 1941, I. Part 1:98.

*Greater than sub-totals, owing to omission of minor units. >Some units omitted.

SCHEDULED CASTES UNDER THE CONSTITUTION

The Constitution makers eschewed any attempt at a connotative definition of untouchability and adopted the same sort of solution as

had the British. Lists were to be compiled for each area of those castes

which were regarded as needful and deserving of such treatment. Rather than provide any standard of selection for these groups beyond the general understanding that they were the “untouchables,” the

132

Identifying the Beneficiaries

Constitution provided only a procedure for designating them. The President is empowered to specify, after consulting with the Governor of a state, those “castes, races or tribes or parts of or groups within castes, races and tribes which shall for purposes of this Constitution be deemed to be Scheduled Castes in relation to that State. . . .”46 Once promulgated, this list can be changed only by Act of Parliament.” A Scheduled Castes Order was promulgated by the President in 1950 which basically re-enacted the 1936 list.4® The major additions were four Sikh castes (discussed below in chapter 9,§E) and the provision for

the first time of lists for areas which had none previously, (notably Rajasthan, Gwalior, and Madhya Bharat). There were a few changes in 1951, but these were in the nature of adjustments and elimination of anomalies, rather than any basic

change in policy.49 The

1951 census revealed that the Scheduled

Castes contained over 52 million members.5° More discrepancies, errors, and oversights came to light, and in 1956 the lists were revised

again.5! The principal changes were the addition of about a million

persons each in Rajasthan and in U.P. and the inclusion now ofall Sikh untouchables. The revised list contained 55.3 millions (using 1951 census figures). In the 1971 census, the Scheduled Castes contained 80 millions (14.6% of the population).5? (See table 15.) The changes in the list have been more in the nature of removing anomalies and correcting oversights: moving groups from the Scheduled Caste to Tribe lists, sorting out discrepancies and errors, rationalizing administration.53 There have been no important policy departures (other than inclusion of the Sikhs). Indeed, there has been no further attempt to formulate any criterion for inclusion. In his first report the 46. Art. 341(2). 47. Art. 342(2). 48. Constitution (Scheduled Castes) Order, 1950. 49. Constitution (Scheduled Castes) (Part C States) Order, 1951. 50. The original 1951 Census figure of 51.3 millions understated the total for a number of reasons, including the omission of untouchables who had returned themselves under generic names (Harijan, Scheduled Caste, etc.) or under “no caste” (Ad-Dharmis, etc.). In 1953, the census authorities made adjustments which brought the total to 52.1 million (RCSCST 1954: 154-55, 332-33).

51. The Scheduled Castes and Scheduled Tribes Order (Amendment) Act, 1956 (Act

63 of 1956). The Backward Classes Commission, which was formed in 1953 and reported in 1955, had been asked to recommend alterations in the Scheduled Castes and

Scheduled Tribes lists. 52. These numbers increased slightly when the list was amended in 1976. See n. 74, below. 53. For some examples of administrative confusion see, e.g. RCSCST 1952: 27 (overlapping separate lists of “Depressed Classes” and “Scheduled Castes” in Mysore); RCSCST 1958-59:I, 9 (separate and different Scheduled Caste Lists issued by two government departments in Kerala).

Ott

9e'et

409

Ose

wes

a60

“auay wants rou SHUN Jot apnjout S110] APOC ANN (11) VT] Med “1 S99g “1 261 “EIU Jo snsUIZ) :—> {$0 ‘11 29-1961 LSOSOY + UIL6-9661 LSOSOU 'V

oes ou

wot

v

Tao

(1s61) uon

v

rol Ove

os

-eindog——-endog

201n0g

10N :224n0g

avinng YSSIuO

JAOSAN SVAGVW

WVSSV

AVaWOE

avavino

Hsddvad VAHCVW vivaay WIWHSVY 3 AWKVE VaLHSVAVHVA

VAVLIVNUVY] AVN TINV.L]

VNVAUVH avinod

SIVLO]

ess

wor sel

alg

ces

v9 ot

c10°9

ts

6E

990 ble ool'el css

290 669%

19% sel +e

see aor £90 REZ

stl

HOSES

06t'E 6297

68ST 166°S

ter

UVHIG

+91 BL OSE £91 el GL 08 Lost

ass 186s

wet KL VN

BBE

Hsdavad VaHONV

Solt

cee 508

ls So“

216 0z1 910

OEl's

le

Ve £9

9160 0S69

EOS Gul

ws! tee we

20e'S

ele

06

v

Loo soe tt

LUFSL OSE

wee 69 9 vee

BLL

16%

+t

v

SalMOLIMAAL NOIND IWONAG LSAaM HSIQVad UVLLIN NVHisvivu

eel

908

i9dt

wor

9 le +b E88 LLS0

1961

oo

sive

tel Sot

£99

rl

ove

oes Lee

wl

Bor

9L7 ae

SEOs

ell

£l6>

v

a

42H 0

cel 0

a

ose

a9

Os'ee

a

von —-endog. von -endog

uonw

won

30% OS

0561 01 Burprcooy

Ss2PIQ 1561 pur

tmoy =

IS

ondog

J0%IS

9661 01 Surpacooy

EL

-ejndog

fmol PRO

96'F1

2

2

uon

°

uo —-eindog

von —-endog

(161) uone,

won -endog

os

‘snsus.) 1961

raoy, — -eindog =

O%IS

ondog

Ol,

os

yeoy

J°% DS

‘snsua:) 1261

1Z61 ‘1961 ‘9C6I ‘1S61-OG6I ‘SIOVINZOWAY GNV (SNOITIIN NI) NOLLV1NdOd SALSV:) GAINGaHIg

Sl FTavL

134

Identifying the Beneficiaries

Commissioner for Scheduled Castes and Tribes observed that Hutton’s 193] census criteria “appear to hold good for purposes of specifying the Scheduled Castes.”’54 Although there was no indication that they were applied any more literally than in 1931 or 1936, no one seemed troubled. In the late 1950s the Commissioner remarked that the criteria for classifying a caste as Scheduled were “fairly definite.”’55 Formally, the power to make and change the lists lay with the President (and later with Parliament), but it appeais that the dominant role was played by the executive, center and state. The lists remained in charge of the Ministry of Home Affairs except for an interval from 1964 to 1973 when this task was housed in the erstwhile Department of Social Security (later Social Welfare).56 Although the role of the state is formally consultive, designation seems in fact to be a two-stage process in which center or state proposes and attempts to get consent of the other. Where the state is unwilling, listings favored by the center can be held up for years. On the other hand, where states have proposed changes in the list, especially inclusions, these have been rejected for

lack of any supporting data.5”

The selection of Scheduled Castes, then, has proceeded without be-

nefit of a single connotative definition. The selection has proceeded primarily on the basis of “untouchability”—measured by the incidence of social disabilities—but this criterion has been combined in varying degrees with economic, occupational, educational, and (as we shall see) residential and religious tests.

The resulting list, then, designates all of those groups who in the view

of Parliament require the special protections provided by the Constitution: it defines who may stand for reserved seats and enjoy benefits and reservations

for the Scheduled

Castes.

But it does not necessarily

include every person or group that might be considered “untouchables” by any conceivable definition. It omits some groups which historically suffered disabilities (e.g. Ezhuvas) or which would be untouchables in terms of the 1931 census tests. And it excludes nonHindus (other than Sikhs) who would clearly seem to be untouchables within the judicial test of “origin in a group considered beyond the pale

of the caste system.”5® There is, thus, no single inclusive list of all

groups considered untouchable, just as there is no single criterion for identifying them. 54. 55. 56. 57.

RCSCST RCSCST RCSCST RCSCST

1951: 10-11. 1957-58:1, 7. Cf. Ministry of Home Affairs 1956:1. 1963-64:1, 13; 1972-73:1, 251. . 1960-61 :1, 22. Department of Social Security 1965:55 ff.

58. See Devarajiah v. Padmanaa, A.1.R. 1958. Mys. 84 at 85, discussed below at nn. 105

and 109.

Designation of Scheduled Castes and Tribes

135

In spite of the absence of a definitive criterion (or perhaps because of it), the lists have remained stable for more than thirty years. Some ups continue to petition for inclusion. The Commissioner observes that “there has been a tendency of some castes and communities to use every opportunity for pressing their claims for inclusion in the list of Scheduled Castes/Scheduled Tribes.”5? But the process of getting on is so cumbersome that few have succeeded. If the pre-Independence figures of 12.3% are adjusted to allow for the failure of some areas to have lists at all and for the removal of predominantly non-Hindu areas by partition, the 1971 figure of 14.6% (the same as in 1961) represents very little change in overall proportion. There have been no significant inclusions (other than the Sikhs). The changes that have occurred have

been rationalizations and extensions of existing criteria to overlooked groups and areas previously omitted, rather than any change in criteria. Nor have there been any significant exclusions with the exceptions of the “‘voluntary” departure of the converts to Buddhism. There has been no litigation about inclusion or exclusion of any group challenging the “untouchability” criterion or its application—but there has been litigation on the religious and residence tests. With whatever admixture of other factors, the Scheduled Castes have

been chosen primarily by the criterion of low social and ritual standing.

In the selection of Other Backward Classes, the use of social and ritual

status as a criterion of the backwardness of the group has been severely restricted by the courts. But such criteria have been predominant in the selection of Scheduled Castes, who have been chosen precisely on the ground of their low social and ritual status in the traditional social hierarchy. The justification for employing low status and attendant disabilities as a criterion for preferences is that these groups are generally lowest in income, education, health, cultural resources, and economic

opportunities. Generally, more than a mere correlation is implied.

These castes are, in the prevailing view, backward because they were

subject to disabilities which impeded advancement by individual effort and denied them opportunities for group social mobility.* This emphasis has endured over the years. In the debate preceding passage of the 1976 amendments to the Scheduled Castes Order the 59. RCSCST

1957-58: I, 7. Petitions for inclusion are still made. See the representa-

tions of South Indian barbers and washerinen to the Joint Parliamentary Committee, reported in The Mail, 16 September 1968. In a few instances, castes have petitioned for removal from the list on the ground that inclusion was degrading. See e.g., RCSCST

1952:66.

60. See chap. 8, §B, below.

61. M. N. Srinivas (1956:377) observes that “the barrier of untouchability” cuts off untouchables from the opportunities for group social mobility that do exist within the

caste system.

136

Identifying the Beneficiaries

view surfaced that “it is basically the economic and social backwardness of a caste which should entitle it to be categorized as a Scheduled Caste.”®? But the Home Minister was emphatic in reiterating that the concept of Scheduled Castes is one of backwardness stemming from untouchability. “[I]t is not neglect, it is not mere poverty, it is not mere backwardness that entitles a man to come under the Scheduled Castes.” However, to the extent that the legislative prohibition of disabilities and changing social practice result in a lessening of disabilities, and to the extent that preferential treatment increases educational attainments and economic opportunities, the correlation between low ritual standing and economic and social backwardness will be weakened. If “untouchability” is a useful criterion for identifying those groups which are at the bottom

in terms of economic,

educational, and cultural

resources and opportunities, the very success of redistributive measures might lead to a discrepancy between ritual standing and the other indicia of backwardness. It would seem that Parliament’s power to designate Scheduled Castes is sufficiently broad to permit the use of standards which would decrease the role of ritual standing and disabilities and

increase the emphasis on educational, economic, and

cultural criteria. There is room for eventually changing the mix of ritual and other factors in the selection of Scheduled Castes. There has been some discussion about changing the mix—at

least to the extent of

applying economic and educational tests to de-schedule those groups which have allegedly been so benefitted that the correlation no longer holds. In June 1965 a committee was appointed by the Government to advise on revision of the existing lists of Scheduled Castes and Scheduled

Secretary.

Tribes, under

The

the chairmanship of B.N.

Lokur Committee

1965.6 It found untouchability,

that in view

reported

promptly

Lokur,

the Law

on August

of the constitutional

25,

abolition of

[I]t would indeed be inappropriate . . . toapply the sole test of untouchability in preparing the list of Scheduled Castes . . . Nevertheless, having regard

to the historical background, we have, in revising the list, adopted the test of

extreme social, educational and economic backwardness of castes, arising

out of the traditional custom of untouchability.®

62. Shri Santosh Kumar, RSD Vol. 97, No. 18: Col. 35 (3 Sep. 1976). 63. Shri K. Bhrahmananda Reddy, LSD (5th Series), Vol. 64, No.16: Col. 125 (2

Sep. 1976).

4. Department of Social Security 1965. The CSCST suggests that the Basavalingappa

case’s requirement that all synonyms and subgroups be listed (see chap. 9, §A, below) was the occasion for the appointment of the Lokur Committee (RCSCST 1965-66 :65).

65. Department of Social Security 1965:6.

.

Designation of Scheduled Castes and Tribes

137

The committee noted with satisfaction that untouchability was “fast

disappearing, particularly in cities and towns, and, even where it is practised, it is considerably diluted in form.”® They observed that “[t]he line of demarcation between high castes and low castes, which was fairly

clear in the past, has tended to become blurred.” §7 Asserting that ‘‘a

lion’s share” of the various benefits and preferences were “appropriated by the numerically larger and politically well organized communities” on the list, and noting the deleterious effects of creation of “vested interests” in the listing, they concluded that “the time has come when the question of descheduling of relatively advanced communities

should receive serious and urgent consideration.”

The Committee devoted most of its attention to many technical changes in the lists, including a number of minor exclusions and inclusions. On major de-scheduling, though, they curiously omitted to utilize any of the fresh and abundant 1961 census data. Instead, they cautiously put forward a list of communities adjudged to be “relatively forward” by “several persons who appeared before us, including eminent social workers.”’7° In addition to 14 tribes, this list included 28 Scheduled Castes listings, ranging from the insignificant to such numerous groups as the Chamars (Bihar, U.P., and Punjab) and the related Jatav (Madhya Pradesh); the Dhobis (West Bengal; Orissa, Bihar, and U.P.); the Mahars (Maharashtra and Madhya Pradesh); the Malas (Andhra); and the Namasudras' and Rajbanshis of West

Bengal. In a kind of dim reflection of the disputes of the early 1930s, the proposals would have left the southern Scheduled Castes virtually intact, but practically halved the Scheduled Caste population in the north.

In addition,

Scheduled Tribes.7!

it would

have eliminated

about

a fifth of the

66. Id., at 6. 67. Id., at 6. 68. Id., at 8. 69. Id., at 10. 70. Id., at 11. Dushkin (1972:220 ff.) exposes the thinness of the Coramittee's

assertions about the advancement of the groups it proposes to de-schedule. Only a handful had literacy rates higher than the average for Scheduled Castes in their states; over half had literacy rates lower than the all-India rate for Scheduled Castes. 71. Estimated percentage of Scheduled Castes and Scheduled Tribes in each state who would have been de-scheduled if the provisions of the Lokur Commission had been implemented Andhra Pradesh Assam Bihar Gujarat Kerala

Madhya Pradesh

Madras

Maharashtra

Scheduled Castes

34.8 32.01 34.07 11.35 451

Scheduled Tribes

_ 40.48 _-

66.69

57.98

35.12 (Table continued overleaf)

20.32

0.04

_-

138

Identifying the Beneficiaries

The report was received with angry resistance by Scheduled Casve spokesmen. At a tumultuous meeting with the Law Minister (then in charge of the Social Security portfolio), Scheduled Caste Members of Parliament agreed with the few inclusions but would not countenance

any exclusions from the list. The Minister was reported to have retreated. Scheduled Caste leaders and government officials concurred

that the report was dead.72 After the 1967 elections, the question was raised again—now with the added dimension of the impending debate on whether to extend the

reserved seats, due to expire in 1970. In the meantime, Congress (the

principal political beneficiary of reserved seats) had become considerably more dependent on the adherence of the occupants of the reserved

seats, who now supplied its much narrower plurality in the Lok Sabha.

The Congress plurality was 38 seats (279/520); Congress held 72 of 114

reserved seats.”

On August 12, 1967, a bill to amend the lists was introduced by the

Government; the bill’s proposed list included every one of the groups

whose possible deletion had been suggested by the Lokur Committee. A Joint Parliamentary Committee established to review the lists submit-

ted its report in November 1969. Among the amendments proposed by

the Joint Committee were the inclusion in Scheduled Castes of women who

married

Scheduled

Caste

men,

the exclusion

from

Scheduled

Tribes of converts to Christianity and Islam, and the elimination of all area restrictions on Scheduled Caste status. The conversion proposal provoked great controversy, and neither it nor the removal of area

restrictions were accepted by the Government, which proposed a host

Mysore

(Table continued from previous page)

Nagaland Orissa

Punjab

8.60

98.91

38.56

—

9.6

Uttar Pradesh

62.22

All India

33.88

West Bengal

-

_

31.81

4.95 _

_

20.29

Percentage of SC recommended to be de-scheduled in four southern states: 13.34. Percentage of SC recommended to be de-scheduled in the rest of India: 40.44. Source :Computed from Census 1966. 72. One cynical and well-placed observer concluded that the Lokur Report was never

meant to be taken seriously, since the Government would not remove electorally potent communities like Chamars, Mahars, and Namasudras from the list a year before the general elections, but found it convenient to remind them that there was such a possibility and how much they had to be grateful for. Interview, B. P. Maurya, December 1965. Whatever the motivation of the political leaders, there is no doubt that the

committee members were serious and well-intentioned. 73. Chandidas, et. al 1968:553-643.

Designation of Scheduled Castes and Tribes

139

of amendments to the Joint Committee’s version when the bill was taken up in November 1970. Mired in controversy, the bill lapsed with the dissolution of the Fourth Lok Sabha in December 1970. A new version, brought forward during Emergency Rule and finally in August 1976, eliminated most intra-state area restrictions, adding about 2.5 million to the Scheduled Castes population and over 3 million to the Scheduled Tribes.’ Apart from this, the Act maintained the status quo: there were no significant inclusions or exclusions. It did not address any of the touchy issues of personal status (of wives, converts, or migrants) which had come to surround the application of

the lists.

Caste remained the primary basis for designating the Scheduled Castes. Caste plays a dual role in the selection of Scheduled Castes: castes in the sense of social units are typically the groups which are listed, and it is caste in the sense of rank or status in the social-religious hierarchy which is the criterion for choosing these groups. But caste is not the only factor in the selection: two other factors are explicitly employed—territory and religion. Scheduled Castes are designated by state and sometimes by districts or regions within states. The same caste may be scheduled in one state but not in an adjoining state, or one district but not in an adjoining district. The power of the President to specify Scheduled Castes by districts 74. Scheduled Castes and Scheduled Tribes (Amendment) Act, 1976 (Act 106 of

1976). Only 64 of 1,126 intra-state area specifications were retained. The additions to the Scheduled Caste and Tribes population can be seen in the following table: Increases (and Decreases) in Scheduled Caste and State

Andhra Pradesh

Tribe Population due to 1976 Act (in lakhs) Scheduled Castes

+0.41

Bihar

+ 4.35

Gujarat

Scheduled Tribes

+ 5.68

+ 0.64

+ 0.23

+ 4.27 + 2.30

+0.31 — 0.76

Maharashtra Orissa

+151 — 0.04

+ 8.87 + 0.03

Tamil Nadu

+ 0.22

West Bengal

+ 0.84

+0.70

+ 24.76 + 0.04

+ 31.57 + 0.76

Himachal Pradesh

+ 0.38

Karnataka Kerala

Madhya Pradesh

+ 2.98

Rajasthan

+140

Uttar Pradesh

+ 5.46

Total Source: RCSCST

1975—77:1, 16.

+ 14.28

+0.09

+ 1.38

140

Identifying the Beneficiaries

has been upheld on the ground that the social and educational backwardness of a group may vary in degree in different areas.75 Specification by states has been held to be constitutionally required. A postal clerk residing and working in Orissa was a member of the Konda Kapus, a group listed as a Scheduled Tribe in neighboring Andhra Pradesh but not in Orissa. After being appointed to a higher post against a reservation for Scheduled Tribes, he was reverted on the ground that Konda Kapu was not a Scheduled Tribe in Orissa where he was a permanent resident. He argued that he should be accounted a Scheduled tribe anywhere for purposes of central government employment, since central legislation clearly outlawed state residence requirements for centra} government jobs.” The High Court, turning to the constitutional provisions for designating Scheduled tribes (Art. 342[1]), finds that the phrase “shall be deemed to be Scheduled Tribes in relation to that State” means that in order to get the benefit of being a member ofa Scheduled Caste or a Scheduled Tribe in the matter of public employment, the person claiming it should be a member of such caste or tribe in relation to the particular area or state where he is residing

and where he seeks employment.”” Thus the court takes a restrictive reading of residence rules to be a constitutional requirement. On numerous occasions, officials and political leaders have taken the view that only a constitutional amendment could abate these restrictions.”® However,

the area limits have been troublesome. There is some

doubt as to whether they correspond with any exactness to the different levels of deprivation and need. (Indeed, the decision ‘not to collect

- caste-wise census data for any groups other than those already listed as

Scheduled Castes or Tribes makes this exceedingly difficult to determine.) More troubling still are the difficulties presented by cases of spatially mobile members of these groups. The Scheduled Castes Order says rather obscurely that castes shall “be deemed to be Scheduled Castes so far as regards members thereof resident in the localities specified in relation to them respectively.” .. .79 In his first report in 1951 the Commissioner of Scheduled Castes and Scheduled Tribes 75. Bhaiya Lal v. Harikrishnan Singh, A.1.R. 1965 S.C. 1557, 1560. 76. Public Employment (Requirement as to Residence) Act, 1957 (Act 44 of 1957).

of Posts and Telegraphs, A.1.R. 1969 Or. 220, at 221. 71. K. Appa Rao v. Director

78. E.g., the Lokur Committee (Departmet of Social Security 1965: 14); the Commis-

sioner for Scheduled Castes and Scheduled Tribes (RCSCST 1971-2/1972-3: 242-43) and the Home Minister [LSD (5th Series), Vol. 64 No. 16:Col. 166 (3 Sep. 1976)]. 79. Para. 2. Like much else, this was carried over from the 1936 order, evidently

without any thought about its purpose. Since the principal “benefit” envisaged in 1936 was special electoral provisions, the residence requirement was in the nature of provision for determining electorates, but wholly inappropriate now that the lists are used for a variety of welfare schemes.

Designation of Scheduled Castes and Tribes

141

observed that the provision was ambiguous. Taking the example of the Khatik caste, which is a Scheduled Caste in Punjab but not in neighboring Uttar Pradesh, he propounded the following questions: When,

if at all, does a U.P. Khatik who moves to Punjab become a

member of a Scheduled Caste in relation to Punjab? Is a Punjab Khatik who moves to U.P. no longera member of a Scheduled Caste in relation to Punjab or in relation to concessions granted by the central govern-

ment? In a strikingly unhelpful interpretation the Ministry of Law opined that “each case will have to be decided on its facts.” But since “residence at the time in question” was the “determining factor,” they inclined to the view that the Punjab Khatik would not be a Scheduled Caste while residing in U.P.81 The Commissioner’s example contains a series of puzzles which we might usefully sort out and label. First, there is the out-migration problem.

Suppose an X leaves a state where Xs are a Scheduled Caste and goes to

a state where they are not so listed. Such persons are excluded from any

benefits by the prevailing reading of the Order. Thus a Dusadh who

moved his residence from Bihar (where Dusadh was a Scheduled Caste) to Madhya Pradesh (where it was not) was not a member of the

Scheduled Castes for purposes of making a lower deposit when filing a

nomination paper.®? The same rule applies to intra-state territorial

limitations. A railway worker resident outside the taluks listed for his tribe was held properly denied eligibility for reservation in

promotion.®3

There is the converse problem of in-migration. Suppose a member of group X migrates from an area in which the Xs are not scheduled to an area in which they are. Here the residence test stressed by courts and

officials would seem to argue for inclusion. But the one court that has addressed it decided the other way. ‘Suryabashi” was listed as a Scheduled Caste in Jabalpur District. But a Suryabashi migrant to Jabalpur who hailed from an area where Suryabashis were not Scheduled was held ineligible for benefits on the ground that “only Suryabashis residing in Jabalpur district . . .were declared to be Scheduled Castes under the Presidential order.”® Apparently the court is looking beyond residence to some notion of origin in the designated groups, a test

that proved insufficient in the out-migration cases. These issues are decided entirely by textual interpretation without any consideration of policies which might provide some guidance here. 80. RCSCST 1951:8. 81. Id., at 8-9. 82. Jyoti Bhushan v. Bodh Ram Muritram, 1974 M.P.L.J. 565.

83. Wasudeo v. Union of India, 1973 Mh. L.J. 994. Residence is equated with “‘perma-

nent abode” rather than place of origin or ancestral home (id., at 998). 84. Lalwani v. Collector, A.1.R. 1976 M.P. 44, 45.

142

Identifying the Beneficiaries

For example, in-migration involves the enlargement of the number of beneficiaries, whereas out-migration does not. Again, extension of benefits to in-migrants might induce mobility that was not otherwise economically justifiable; but extension of benefits to out-migrants might be necessary in order to prevent the inhibiting of economically justified moves. The logic of these cases points to an even more grotesque result in the case of lateral movement. Suppose a member of group X migrates from state A where the group is scheduled to state B where it is also scheduled. According to the residence argument of the out-migration cases, he is no longer entitled to any benefits as an X of state A. Nor,’ according to the origins argument of the in-migration case, is he entitled

to any benefits as an X of state B. One suspects that administrative

failure to give full effect to the logic of the cases, perhaps augmented by some dissimulation on the part of the migrants, has tempered the effect of these rules and accounts for the infrequency with which these issues have come to court. That migrant members would lose their Scheduled Caste benefits remained

troublesome.

In

1959, the Commissioner

had raised the

matter again,®5 this time in connection with the case of a Dhobi whose family origins were in U.P. (where Dhobi is a Scheduled Caste) but who was born, brought up, and employed in Bombay (where Dhobi is not a Scheduled Caste). He claimed that his ancestral property, house, etc., were in U.P. and that he had no property or house in Bombay, but lived there only for the sake of his employment. This case was referred to the Law Minister, who again read the order so that he was excluded—a result that the Commissioner found anomalous, in view of the fact that

[s]uch persons may actually continue to suffer from all the disabilities

resultant from the practice of untouchability as they may have to maintain all ties with their relations in the home State.®

He recommended that migrant Scheduled Castes and Tribes should be eligible for benefits from the central government and their home state at least for a generation.®” No relief for inter-state migrants has been forthcoming. In 1965 the Lokur Committee recommended the removal of inter-

state area limitations on the ground that they inhibit mobility.°* The Scheduled Castes and Scheduled Tribes Orders (Amendment) 85. RCSCST 1958-59: 11

Bill,

86. Id.

87. Id.; the Backward Classes Commission had earlier recommended that migrating

members should be treated as Other Backward Classes (1: 155). In 1963 the Commis-

sioner noted that the Government of India had still not taken any decision on this issue,

(1962-63:1, 12).

88. Department of Social Security, 1965: 13.

Designation of Scheduled Castes and Tribes 143 1967, proposed that when a caste is Scheduled for one area of a state, members from that area shall continue to be included so long as they reside anywhere in that state® The 1976 amendment of the lists did not contain this provision, but the need for it was largely dispelled by the abandonment of almost all intra-state area restrictions. The conferring of scheduled status on a group is a monopoly of the

Centre. The 1950 Scheduled Tribes Order list for Maharashtra included an item for Halba from six specified taluks in three districts. In 1967 the Government of Maharashtra ordered that the Halba Koshtis

of the whole of Vidarbha region be treated as belonging to the

Scheduled Tribes. A Halba railway worker from outside the listed taluks was deleted from a promotion panel on the ground that he was not a member of Scheduled Tribe. His challenge was repulsed by the Bombay High Court, which held that the state’s pronouncement could have no effect on his status as a Scheduled Tribe in connection with any of the affairs of the central government. The court left open the question whether the state might treat him as a Scheduled Tribe in regard to state services.9° This issue has never been fully addressed. While'it might be desirable to allow the state to employ its better-informed judgement about local conditions, there is of course the danger that state use of this power could dilute benefits to those deemed deserving beneficiaries by national policy. The most prominent instance of this— Maharashtra’s treatment of the Buddhists on a par with Scheduled Castes—has not

been challenged in court. Religion was introduced as a qualification into the first Scheduled

Castes Order in 1936, which provided that no Indian Christian (nor, in

Bengal, those professing Buddhism or a tribal religion) should be deemed a member of a Scheduled Caste.91 Earlier, it was often recognized that there were comparable depressed groups among Christians and Muslims. But in the disputes leading up to the listing of Scheduled Castes, it was agreed that Muslims and Christians should be excluded.92 This execution was readily understandable, for the major

purpose of the list was to provide for electoral representation, and Christians and Muslims were the beneficiaries of special electoral treatment as minorities.%> In spite of the constitutional ban on religious 89. Lok Sabha Bill No.119 of 1967, p.7. 90. Wasudeo v. Union of India, 1973 Mh. L,J. 994.

91. The Government of India (Scheduled Castes) Order, 1936, §3.

92. Thus Hutton’s note to the Indian Franchise Committee observes that it was decided to omit Christians and Muslims (Indian Franchise Committee 1932:1, 111). 93. For the composition of the legislative bodies under the Government of India Act, 1935, see the First and Fifth Schedules to that Act.

144

Identifying the Beneficiaries

discrimination, the elimination of separate representation tor religious minorities and the change in purpose of the list from electoral to administration of welfare, the religious qualification (or, more properly, disqualification) was retained after Independence. The President’s 1950 Order provides that “‘no person professing a religion different from Hinduism shall be deemed a member of a Scheduled Caste.”* An exception was made for Sikh members of four castes. In 1956 it was broadened to include all Sikh untouchables.95 The religious test for Scheduled Castes is employed, not as a positive test for selecting appropriate groups for inclusion, but as a disqualification of individuals and groups who otherwise meet the criteria, thereby

inevitably discouraging conversion. There is reason to think that this was at least part of its purpose.® It does not operate as an encourage-

ment of Hindu orthodoxy, for the legal definition of Hinduism is so

broad that few individuals are likely to have difficulty with this test other than those who explicitly convert to a non-Hindu religion. In Punjabrao v. Meshram, the Supreme Court rejected the argument that Buddhists were included in the meaning of Hindu.%” The courts have upheld the exclusion of non-Hindus without reaching the broader question of religious discrimination. Nor have the courts addressed the factual question of the effect of conversion in dissipating the conditions that lead groups to be listed as Scheduled Castes. Persistent legislative attempts to include the Buddhists, either by dropping the religious qualification or by specifying that “Hindu” be read to include them, have been unsuccessful—as has a recent attempt

to extend the religious qualification to the Scheduled Tribes. The interplay of religion with caste and tribe membership is discussed in chapter 9. Since we shall be examining the impact of the courts on the working of the policy of “compensatory discrimination,” it is worth noting that the role of the courts in the process of designating the Scheduled Castes has been a very minor one. The search fora uniform criterion that could be employed throughout India to distinguish untouchables from other Hindus proceeded without any assistance from the courts. Although Scheduled

Castes were selected on the basis of the disabilities they

suffered, the jurisprudence of disabilities was too spotty and too en-

meshed with local variations.9® Although untouchables could be read94. Constitution (Scheduled Castes) Order, 1950, §2.

95. Scheduled Castes and Scheduled Tribes Order (Amendment) Act, 1956 (Act 63

of 1956).

96. The problem of converts in taken up in chap.9. 97. A.LR. 1965 S.C. 1179 (discussed in detail below, chap. 9, §E.) 98. See Galanter 1972.

Designation of Scheduled Castes and Tribes ily identified

with

the lower end of the vama

145

scale, the traditional

jurisprudence of varna standing was of little assistance in identifying the groups which were supposedly untouchable. Modern ‘“‘untouchability” bore little correspondence to the chandala category of classical law. In reference to determinations of customary rights, courts had sometimes employed varna categories. Untouchables had sometimes, particularly in South India, been referred to as a fifth vama, below the Sudras.9

But in other places they were regarded as Sudras, albeit “unclean”

ones.?© For purposes of applying Hindu personal law, the courts had never attempted to distinguish untouchables from Sudras; all Hindus

other than the twice-born were lumped together as Sudras.!°! Even where untouchables were popularly regarded as Sudras, they could not be equated with them, since there were non-untouchable groups belonging to this category. Thus the tests used for distinguishing Sudras from the twice-born could untouchability.

not be used as a satisfactory measure of

Once the lists were promulgated, the courts contributed little to their

refinement.1© Although on at least one occasion, judicial demands for

precision helped to provide an opportunity for re-examination, courts have contributed nothing to the critical understanding of the lists for the criteria that underlie them. Although it might seem that identification of “untouchable” groups

would proceed from the definition of “untouchability” in order to select those who suffer from it, in fact the opposite process has taken place. The contemporary legal treatment of untouchability has derived from the listing of these groups (which in turn are chosen, as we saw, with at least one eye to an ill-defined notion of untouchability). The term

“untouchability” had no technical legal meaning before the Constitution abolished it.1 The few judicial encounters with the concept have not succeeded in defining it in a way that would be useful in identifying untouchable groups. The courts have indicated that untouchability does not include all instances in which a person is treated as ritually unclean and a source of pollution. It does not include such temporary and expiable states of 99. See e.g., Sankaralinga Nadan v. Raja Rajeshwari Dorai, 35 1.A.C. 176 (1908).

100. See e.g., Atmaram v. King-Emperor, A.1.R. 1924 Nag. 121.

101. See e.g., Muthusami v. Masilmani, Poongavanammal, A.1.R. 1934 Mad. 323.

1.L.R.

33 Mad.

342 (1909);

Manickam v.

102. A somewhat exaggerated expression of judicial impotence in this area is found in the assertion of then Cmef Justice Ray in State of Kerala v. N. M. Thomas, A.1.R. 1976 S.C.

490, 501, that “‘no court can come toa finding that any caste or any tribe is a Scheduled Caste or a Scheduled Tribe.” Cf. the cases discussed in chap. 9, §A, below. 103. See ch. 9, n. 16, below.

104. On the history of the terminology, see chap. 2, nn.21 and 46, above.

146

Identifying the Beneficiaries

pollution as those suffered by, e.g., women at childbirth, menstruating women, mourners, persons with contagious diseases, persons who eat forbidden food or violate prescribed states of cleanliness.?°5 Nor does it include that “untouchability” which arises from incidents of personal history.1 Nor does it refer to situational or relative impurity, such as that between ordinary worshipper and priest or temple attendant.’ It does not include every instance in which a person is stigmatized as

unclean, polluting, or inferior because of his origin or membership in a particular group—i.e., where he is subjected to invidious treatment because of difference in religion or membership of a lower or different caste.! The “untouchability” with which Article 17 is concerned is that which “refers to those regarded as untouchables in the course of historical development” and which is related to the relegation of persons “beyond the pale of the caste system . . .on the ground of birth in certain classes,” Thus untouchability is confined to disabilities imposed upon groups commonly regarded as “‘untouchable,” and its meaning is to be determined by reference to those who have traditionally been considered untouchables. But it is, as we have seen, no easier to define untouch-

ables than to define untouchability. “Beyond the pale of the caste system”

105. The Mysore High Court, the first to address itself to the problem, pointed out that the word “untouchability” appears in Article 17 between inverted commas and inferred that “‘the subject matter of that Article is not untouchability in a literal or

grammatical sense but the practice ofit as it had developed historically in this country”

(Devarajiah v. Padmanna, A.1.R. 1958 Mys. 84 at 85). Cf. the remarks of Professor K. T.

Shah, VII CAD 668, and Mr. Nazirrudin Ahmad, VII CAD 665. 106. Thus the ban on untouchability does not make excommunication from caste

unconstitutional, although it makes a person “for all practical purposes untouchable.”

Hadibandhu v. Banamali, A.1.R.

1960 Or. 33. (But cf. the argument of Sinha, CJ.,

dissenting in Saifuddin Saheb v. State of Bombay, A.1.R. 1962 S.C. 853 at 866 : forbidding

excommunication was only carrying out “the strict injunction of Art. 17.”” The majority disagreed, finding that the power to excommunicate may in some circumstances be the constitutionally protected right of a religious sect.) 107. Parameswaran Moothathu v. Vasudeva Kurup, 1.L.R. 1960 Ker. 73. Consonant with this understanding, temple-entry legislation puts untouchables on a parity with other worshippers, but does not attempt to erase distinctions between worshippers and priests or attendants. All temple-entry legislation confers rights to enter and worship “in the same manner and to the same extent” as is permissible for other persons professing the same religion, etc. On several occasions the Supreme Court has implied that the ambit of the temple-entry power would not extend to erasing the distinction between ordinary worshippers and temple functionaries (Sastri Yagnapurushadasji 0.

Muldas Bhundardas Vaishya, A.1.R. 1966 S.C. 1119) or sectarian initiates (Sri Venkataramana Devaru v. State of Mysore, A.1.R. 1958 S.C. 255). 108. An attempt to explicitly attach this meaning by amending Art. 17 to read “‘no one shall, on account of his religion or caste, be treated or regarded as an ‘untouchable™”

was defeated in the Constituent Assembly

(VII CAD 665, 669).

109. Devarajiah v. Padmanna, A.1.R. 1958 Mys. 84 at 85.

Designation of Scheduled Castes and Tribes

147

is a misleading and unworkable formulation. Even the lowest castes are within the traditional system of reciprocal rights and duties, though they had more of the latter and fewer of the former. Their disabilities and prerogatives are articulated to those of other castes. Presumably, the court that devised this standard meant outside the four vamas of the classical lawbooks. But, as we have seen, the varna test is not workable

either—although it continues to crop up in discussions of untouch-

ability.42° Thus the courts have not developed any learning on “un-

touchability” that could be employed in refining the criteria of selection of the Scheduled Castes. However, thereis the possibility that as the correlation between low ritual status and economic and educational condition breaks down,

some of the learning developed by the courts in addressing the problems of identifying the Other Backward Classes (see chapter 8 below) will be utilized in the Scheduled Castes and Tribes area. C.

THESCHEDULED TRIBES

There has been much less dispute about the identification of the Scheduled Tribes (ST) than about the Scheduled Castes or the Other Backward Classes. The general notion is that this category should include those groups distinguished by “tribal characteristics” and by their spatial and cultural isolation from the bulk of the population.144 The British attempted to protect these “aboriginal” people by placing areas in which they were concentrated outside of ordinary administration to permit a policy of insulating them from exploitative or demoralizing contact with more sophisticated outsiders. These enclaves, (called Excluded Areas and Partially Excluded Areas in the Government of India Act, 1935) were placed outside the operation of general law and were the object of broad executive powers to provide special protective

regulation.!!? Some provinces had undertaken a policy of protective

110. Thus Justice Krishna Iyer in his passionate defense of special protection for Harijans identifies them as “this mixed bag of tribes, races, groups, communities and

non-castes outside the four-fold Hindu division.” (State of Kerala v. N. M. Thomas, A.L.R. 1976 S.C. 490 at 535). 111. In the early 20th century tribals were sometimes included within the Depressed

Classes, but by 1930 they were carefully distinguished. Weitbrecht-Stanton (1920) counts 18 million aboriginals (excluding Burma). The first serious attempt to identify the “‘primitive tribes” was in the 1931 Census, which counted 24, 613, 848

India 1931, Vol. I, Pt. 1, p. 503).

(Census of

112. Government of India Act, 1935, §91, 92. These enclaves were called “Backward

Areas” in the Government of India Act, 1919 §52-A (2), and before that Scheduled

Districts (see Scheduled Districts Act, Act XIV of 1874). For a review of earlier British

Policy, see SASTC I:21-31; Ghurye 1963, chap. 4.

148 Identifying the Beneficiaries treatment of tribals outside these areas.*13 In the 1935 Act, for the first

time, provision was made from some representation for “Backward Tribes” in the reformed Provincial Legislatures," and a list of Backward Tribes was promulgated in 1936 for all of the provinces except Punjab and Bengal.'5 The 1941 census counted 25.4 million tribal people (6.58% of the total population).16 The Constitution carried forward this dual treatment of tribal areas

and tribal people in a set of provisions for their protection and advancement far more elaborate and detailed than the provisions for the

other categories of backward classes. Along with reservations in legisla-

tures and services comparable to those for the Scheduled Castes, there are provisions for direct central control over administration,’’” for direct central financial responsibility,!1® for annual reports by the Governors

to the

President,!9

for extensive executive power over

Scheduled Areas to insulate them from the application of inappropriate law and to fashion protective regulation (especially regarding allotment and transfer of land and money-lending),'”° and for the formation of Tribal Advisory Councils to participate in the formulation of policy.21 The formal mechanism of designating the Scheduled tribes is identical with that for Scheduled Castes. The President may specify, after consulting with the governors, the “tribes and tribal communities or parts of or groups within tribes or tribal communities” to be the Scheduled tribes for each state.'?2 Once promulgated, the list can be varied only by Act of Parliament.!3 The Scheduled Areas are te be ordered | by the President, who retains the power to alter or de-schedule them. 113. See. e.g., Bombay 1930 (Starte Committee] :3-7. 114. The quantity was only token:a total of 24 of 1,535 seats in the provincial legislatures were reserved. 115. The Government of India (Provincial Legislative Assemblies) Order,

1936,

Thirteenth Schedule. Reprinted in Rajagopala Aiyanger 1937. The identity of the list for Bombay with the Starte Committee’s list suggests that these lists corresponded to already existing state lists, but I have been unable to check this for the other provinces. 116. Census 1943: 101. The total included 19.7 millions in British India (5.65% of the population) and 8.7 millions in the states and agencies (9.61% of the population).

117. Art. 339 (2) (Scheduled Tribes); Fifth Schedule §3 (Scheduled Areas). 118. Art. 275 (1). 119. Fifth Schedule, §3. 120. Fifth Schedule, §4.

12]. Fifth Schedule, §5. The discussion here pertains to the Scheduled Areas outside

Assam. The Scheduled Areas in Assam are dealt with separately in the Sixth Schedule,

which provides for autonomous districts and regions with elected councils which enjoy extensive law-making powers over land, forests, agriculture, social and political life, administration of justice, education, revenue, licensing, and moneylending.

122. Art. 341 (1). 123. Art. 342 (2).

124. Art. 244, Fifth Schedule, §6.

Designation of Scheduled Castes and Tribes

149

In 1950, the President promulgated the list of Scheduled Areas!?5 and a list of Scheduled Tribes, !#° apparently by making some additions

to the 1935 list of Backward Tribes.!27 (Some list of the ST must have

been in use earlier, for the central government’s scheme of postmatriculation scholarship included ST beginning with the academic year 1948-49.)!28 Some groups that had been listed as Scheduled

Castes in Bengal were now moved to the ST list.129 The 1951 census counted 19.1 millions in these groups (5.30% of the total population).

Some tribes had been omitted; about a million members of these “left

out” tribes were temporarily accommodated in the Other Backward Classes category! until 1956, when the Scheduled Tribes Order was amended,'3? adding 2.5 million tribals, mostly in Rajasthan and Madhya Pradesh, bringing the total to 22.5 millions (6.93% of the total population). In 1971 these groups numbered over 38 millions (6.93% of the total population). In 1976 over 3 million persons were added, bringing the total to almost 40 millions. 132 (See table 16.) Although the lists contain groups throughout the country, the tribal pulation is heavily concentrated in central, eastern, and northeastern India.93 Unlike the relatively dispersed Scheduled Castes population, about 90% of the tribals are found in nine states. Three states had less

than 1% Scheduled Tribes in 1961, while Orissa and Madhya Pradesh had more than 20%. More than one-third (10.6 millions) of the

Scheduled Tribes population resided in the Scheduled Areas (where they formed less than two-thirds of the population). (See table 17.) More

125. Scheduled Areas (Part A) Order, 1950; Scheduled Areas (Part B) Order, 1950. 126. Constitution (Scheduled Tribes) Order, 1950; Constitution(Scheduled Tribes) (PartC States) Order, 1951.

127. See Department of Social Security 1965:6. 128. Ministry of Education and Scientific Research 1957:2. This report suggests that the decision to extend the scheme to Scheduled Tribes was made in 1947. 129. Dushkin 1957: 159. 130. RCSCST 1954: 238 ff. 131. Scheduled Castes and Scheduled Tribes Order (Amendment) Act, 1956 (Act 63

of 1956). 132. Virtually all the sizable groups with a plausible claim to inclusion in the Scheduled Castes list are included. There are no prospects ofany major additions to the

list. However, there is a controversy abaut a possible further sizable addition to the Scheduled Tribes list. For years preceding the 1976 revision of the lists there was an effort to secure the inclusion of some 2 to 4 million tea-garden laborers of tribal origin in Assam, a project that threatened unwelcome political consequences in that state and attracted the intense opposition of the state government (LSD [Fifth Series}, Vol. 64, No. 17: Col. 58 and passim [2 Sep. 1976]; Department of Social Security 1965: 18 ff.).

133. For two decades the Government Scheduled Tribes category in spite of the the requisite characteristics. The refusal Tribes “even though the conditions of life

of Uttar Pradesh presence of groups of U.P. to classify and surroundings

refused to countenance a that appeared to have all any groups as Scheduled of these tribals are almost

150

Identifying the Beneficiaries

TABLE

16

ScHEepuLeD Tripes PoPULATION 1950 & 1951 Orders Source

A

State ANDHRA ASSAM NAGALAND BIHAR BOMBAY GUJARAT MAHARASHTRA JAMMU & KASHMIR KERALA MADHYA PRADESH MADRAS MYSORE ORISSA PUNJAB ~ HARYANA

RAJASTHAN UTTAR PRADESH WEST BENGAL Torats

1956 Order A

A

(mils)

%

(mils)

0.766 1.735

8-245 19.18

«1.149 L761

3.827 3.671

986 7.60

3.880 3.743

A %

1961 Census

1971 Census

B

B

c

(mils) = % =

(mils)

“19.147

5.30

22511

%

3.681.328 3.681.657 3.79 19.48 2.068 17.42 1.92012.84 343 93.03 457 88.46 «10.00 4.204 = 9.02, 4.93278 7.76 2754 13.34 3.7384 13.97 2.397 603-2954 5.85.

No Scheduled Tribes 074 0.540.134 = 0.990.207 123 0.270 3.859 14.79 4.844 = 18.58 6.678. «20.62 8.387 0.060 0.20 0.13 © 0450.52.74 O31 0.04 = 0.230.080 O41 0.192 810.231 2.967 20.25 3.009 20.55 4.223 24.02, 5.071 0.002 0.01 002 0.020.014 0.06 0348 218748 No Scheduled Tribes 1386 «5.271.366

c

UD

2.309

5.96

2063

6.23

29.846

-

1.26 0.14 0.78, 0.78 23.11 -

1.023.125 £198 5.912.532

12 2 5.70

680

6.93

38.015

.

Sources: A: RCSCST 1956-57 : II, 15; B: RCSCST 1961-62: II, 74; C: Census of India, 1971, Series I, Part I-A (ii): xxx-xxxi. Note: The totals may be slightly higher than the columns indicate, owing to the inclusion of Scheduled Tribes in the Union Territories.

than halfof the Scheduled Tribes lived in parliamentary constituencies where they formed a majority of the population. The Scheduled Tribes were defined partly by habitat and geographic

isolation, but even more on the basis of social, religious, linguistic, and cultural distinctiveness— their “tribal characteristics.” Just where the line between “tribals” and “non-tribals” should be drawn has not

always been free from doubt.'55 We shall see that there are still identical with tribals in neighboring states” was described as “unusual” and “unsound” by the Backward Classes Commission (I BCC 155-56). The matter engaged the Government of India for a long time without any action by the state government

(RCSCST

1965-66:66), but finally in June 1967 a notification was issued by the

President declaring five communities as Scheduled Tribes in U.P. (RCSCST 196667:27). 134. See chap. 3, text at notes 28 ff. 135. This has been entwined with a controversy over the extent to which tribals were “Hindus.” For a taste of the long-standing academic controversy over this, compare

Designation of Scheduled Castes and Tribes TABLE

151

17

Scuepuep Argas PoPpuLation STATE ANDHRA PRADESH

BIHAR GUJARAT

MAHARASHTRA JAMMU & KASHMIR KERALA MADHYA PRADESH MADRAS MYSORE ORISSA PUNJAB

RAJASTHAN ‘UTTAR PRADESH WEST BENGAL Tortats

1961 Total

1961 ST

1961 ST

inSA

1961 ST in SA

SA Pop.

Pop. in SA

as% TotalSA

as % Total ST

(mils.) .

(mils.) 0.438

% a

% 33.08

5.257 1.987

1.165 No Scheduled No Scheduled 3.395 No Scheduled No Scheduled

2.929 1.306

55,71 65.60,

69.67 4742

63.52

32.15

2.392 0.014

60.50 68.90

56.64 100.00

10.647

61.28"

0.856 Areas Areas 2.147 Areas Areas

63.26,

0.877 0.564 No Scheduled Tribes No Scheduled Areas

6431

3.950 0.020

16.659

35.71

24.42

35.70

Source: RCSCST 1965-66 : 186. Note: “Andhra Pradesh not included in total.

problems of overlap with the caste order (discussed in chapter 9, §§ C,D, below) and controversy about whether a specific group is

more approporiately classed as a Scheduled Tribe or a Scheduled Caste. In any event, not all tribal peoples were listed, but only those

Ghurye 1963:19 (‘the only proper description of these people is that they are the imperfectly integrated classes of Hindu society ....[T]hey are in reality Backward

Hindus”)

with von Furer-Haimendorf 1967. On the theoretical complexity of the

tribe-caste distinction, see Bailey 1961.

136. A recent and fascinating instance surfaced in the Lok Sabha debate over the

revision of the Scheduled Castes and Tribes Order in 1976. Shri Dasaratha Deb

that the Laskar community in Tripura be treated as a Scheduled Caste rather than as a Scheduled Tribe because of their lack of affinity with the tribes ofTripura and especially because their treatment as tribals has enabled them to play a “dirty game” of using their marginal status to circumvent the restrictions on transfer of tribal lands to non-tribals. According to his account, as tribals, tribal land could be transferred to them without prior sanction of the government; they then used their Bengali surnames to transfer the land to non-tribals without arousing the suspicion of the registration authorities. Apparently this group are the descendants of migrants from Bengal who ‘were accorded tribal status by the former Maharaja. They are not separately listed in the Tripura List as either a Scheduled Caste or a Tribe (nor were they previously). Shri

152

Identifying the Beneficiaries

whose primitiveness, backwardness, or isolation made them deserving

of special treatment. 137 No religious test was specified. A recent attempt

to exclude converts failed to gain acceptance.38

In 1951 the Commissioner noted that no uniform test for distinguish-

ing the Scheduled Tribes had been formulated."99 He circularized the

state governments for their views. After surveying the divergent res-

ponses, which included racial, religious, linguistic, and cultural tests as well as tésts of political, economic, and social organization, and degree of contact, isolation, or assimilation/distinctiveness, the Com-

missioner proposed as the common elements: tribal origin, primitive way

of life,

remote

habitation,

and

general

backwardness

in

all

respects.140 Whatever theoretical imperfections may underlie the lists, nothing

takes on fixity so readily as a rough approximation made to solve an immediate and intractable problem. A few years later the criteria for

Scheduled Tribes seemed wholly unproblematic. The Home Minister

in 1956 referred to the ‘“‘well-defined and easily ascertained” criteria for

distinguishing them.** By the time the Dhebar Commission reported in

1961, it felt no need to devote any attention to the problem of

identifying the tribes.'4? The Lokur Committee in 1965 reported that, in revising the lists, we

have

looked

for

indications

of primitive

traits, distinctive

culture,

geographic isolation, shyness of contact with the community at large and backwardness; we have considered that tribes whose members have by and large mixed up with the general population are not eligible to be in the list of

Scheduled Tribes. ‘43

The provisions for Scheduled Tribes are distinct not only in their scope but in their intent, for preservation of their separate integrity, rather than complete assimilation, is one element in policy toward Dasaratha Deb’s Amendment to exclude them from the item for “Tripura” by adding “except Laskar termed as Deshi Tripura by ex-ruler’s Order” was defeated (LSD [Fifth Series], Vol. 64, No. 17:Cols. 45, 156, 166 [2 Sep. 1976])

137. 138. 139. 140. 141.

Department of Social Security 1965:47. See discussion in chap. 9, § E. RCSCST 1951: 11. RCSCST 1951: 109-11. Ministry of Home Affairs 1956: 1. Cf. Majumdar’s (1960:219) 1954 reference to

the relative case organization.”

of their classification

by “self-evident and

homogeneous

tribal

142. SASTC.

143. Department of Social Security 1965:7. The first part of this language is utilized

almost verbatim by the Home Minister more than 10 years later in the debate about revision of the lists (LSD [Fifth Series], Vol. 64, No. 17: col. 124 [2 Sep. 1976}).

Designation of Scheduled Castes and Tribes

153

them. In the case of the Scheduled Castes, the Other Backward Classes,

and the Denotified Tribes; it is hoped that the disabilities and disadvantages that separate them from their compatriots will eventually be overcome by the preferential treatment. That is, the aim of the policy is to eliminate their distinctiveness by dispelling all of the differences that set them apart from other Indians. The aim of Scheduled Tribes’ policy is more complex—to balance improvement of their condition and a degree of assimilation with preservation of their distinctiveness and a measure of autonomy.* Special treatment of the Tribes will, it is hoped, foster development “along their own lines.” Because of this difference in purpose (and because of their geographical separation), the objections that have been made against the use of communal criteria in other categories of backward classes have not been made with the samé force in regard to the Scheduled Tribes. Although Scheduled Tribes share with Scheduled Castes a variety of reservations—in legislatures, government service, and educational institutions— the principal schemes for the Tribes have been protective and developmental (economic, educational, health, and communica-

tions). Tribal aspirants have not pressed hard on the reservations; many have gone unfilled. With the exception of a few election cases raising problems of tribal membership,’*5 tribes have been little involved in litigation about compensatory discrimination. They have had little effect on the law, which is being developed in cases involving the Scheduled Castes and the Other Backward Classes. 144. Just what the mix of assimilation and distinctiveness should be has long been a

matter ot dispute. Cf., ¢.g., the Dhebar Report [SASTC] and Ghurye’s (1963) critique. 145. Discussed below in chap. 9, §C, D.

6

The Other Backward Classes: The Process

of Designation

A. THE PRE-HISTORY OF THE BACKWARD CLASSES CATEGORY THE OTHER BACKWARD Classes for whom preferential treatment are authorized are not defined in the Constitution, nor is any exclusive method or agency for their designation provided. For all the uncertainty surrounding the term “Scheduled Castes” (and its predecessor ‘“‘Depressed Classes”), its central purpose is clear: to identify the victims of “‘untouchability.” At the time of Independence, the term

“Backward Classes” had a less fixed and definite reference. The term

had been around for some time, but it had a variety of referents, it had

shifted rapidly in meaning, and it had come to mean different things in different places. For purposes of tracing these varied meanings in the following discussion, we may portray some of the variety of meanings in schematic form in figure 1. : In 1917 the Maharaja of Kolhapur recounted to Mr. Montagu that he had “‘been taking very keen interest . . .in uplifting the Backward

Classes and especially the untouchables”? [Denotation 3?]. Soon after,

the term appears in the terms of reference of the Southborough Committee, which was to advise on measures to secure representation of “minorities, of special interests, or of backward

classes.”? But the

Committee did not mention any such groups in its report, other than

Depressed Classes [Denotation 1?]. The Joint Select Committee of the

British Parliament which reviewed the Southborough Report mentioned in passing the importance they attached to “the educational advance-

1. The Fort St. George Gazette No. 40 of November 5, 1895, mientions grants in aid to

schools for a list of “Backward Classes” which includes most of the “untouchable” castes of Madras Presidency. Irschick (1977) notes that the term was used as far back as 1880 to describe a list of groups, also called illiterate or indigent classes, entitled to allowances for study in elementary schools. See Saraswathi 1974 : 107 ff. 2. Latthe 1924 : II, 574. 3. Indian Constitutional Reforms 1919: 1.

155

Other Backward Classes FIGURE

1!

Highest Castes -

y

_

intermediate

one

y

Communities

YY

Wy

WY

Untouchables ; Crmial Vibes, Nomads,

Y

etc

7/ 7

YY

yy

Y

YU

Yyy

YY Yyf

oe

j

“YY Yi Yy Y

Yj

Yy

Y Uy

ty

VARIOUS DENOTATIONS OF THE “‘BACKWARD CLASSES”

So

2

So°

Se

1. 2. 3.

Asa synonym for Depressed Classes, untouchables, Scheduled Castes.

As comprising the untouchable, aboriginal and hill tribes, criminal tribes, etc.

As comprising all the communities deserving special treatment, namely those included

in

(2) above

and

in addition

the

lower

strata of non-touchable

communities. As comprising all non-tribal communities deserving special treatment. As comprising all communities deserving special treatment except the untouchables. As comprising the lower strata of non-untouchable communities. As comprising all communities above the untouchables but below the most “advanced” communities. As comprising the non-touchable communities who were “backward” in comparison to the highest castes. As comprising all communities dther than the highest or most advanced. As comprising all persons who meet given non-communal tests of backwardness (e.g., low income).

Note: this figure portrays varying usages, of which examples are given in the text. Exact specifications may differ along other dimensions as well—e.g., whether non-Hindus are included or whether prosperous families within these groups are excluded.

156

Identifying the Beneficiaries

ment of the depressed and backward classes.”* There is no indication that the term was meant to include anyone besides the depressed classes

[Denotation 1? 2?].

“Backward Classes” first acquired a technical meaning in the Princely State of Mysore. In 1918 the Mysore Government appointed a committee to enquire into the question of encouraging members of the “backward communities” in public service.5 In 1921 preferential recruitment of “‘backward communities” was instituted, and they were

defined as ‘all communities other than Brahmins, who are not now

adequately represented in the public service.” [Denotation 9]. The Reforms Enquiry Commission (1924) did not find occasion to use the term, but the Ministry Report refers to its use as a synonym for the Depressed Classes (untouchables) and in contradistinction to “non-Brahmins.’’? [Denotation 1].

Although “Non-Brahmin” and “Maratha” were much more fre-

quently used in the setting of the Bombay “non-Brahmin” movement of the 1920s, the Backward Classes rubric was employed both popularly and officially in a broad meaning, somewhat akin to that in Mysore.® In

1925 a Government Resolution defined Backward Classes as all except

Brahmins, Prabhus, Marwaris, Parsis, Banias, and Christians? [Deno-

tation 9?]. Reservations in government service were provided for this group. The Hartog Committee glossary:

(1928) defined Backward Classes in their

Castes or classes which are educationally backward. They include the depressed classes, aboriginals, hill tribes, and criminal tribes. 1° [Denotation 2}

In 1929 the Indian Central Committee distinguished the problem of the “backward classes,’ among whom may be counted aboriginals, criminal tribes and others among the less advanced of the inhabitants of British India”?! [Denotation 5]. Apparently the Committee not only excludes

the Depressed Classes

but includes,

in addition to tribals,

some strata of the caste population, for they mention an estimate of 16 4. Reprinted in P. Mukerji 1920 : 528.

5. A 1921 G.O. referring to the 1911 G.O. is reprinted in Mysore Backward Classes

Committee 1961 : 57-61.

6. Id., at 58. An attached education memorandum employs the term “backward classes” (id., at 61), but it is not clear that it was exactly synonymous.

7. Reforms Enquiry Committee 1925: 147.

y. cited 10. 11.

8. Omvedt 1976: 184, 186, 188.

Government of Bombay Finance Department Resolution No. 2610 of Feb. 5, 1925, by Omvedt 1976 : 343. Indian Statutory Commission (Hartog Committee) 1929:399. Indian Central Committee 1929 : 45.

Other Backward Classes

157

million backward classes in U.P. (excluding the untouchables).'? But the category is not used as expansively as the Mysore or earlier Bombay

usage, for it did not include the Madras non-Brahmins or Mahrattas in Bombay, who were discussed separately.13 The term could not have been a familiar one, for the Report finds it necessary to distinguish Backward Classes from untouchables several times. In a separate note, Mr.

M. C. Rajah, an untouchable spokesman,

one

tion 1].

mentions Backward

only as a confusing synonym for Depressed Classes’ [Denota-

In 1930 the Starte Committee in Bombay devoted careful consideration to the question of nomenclature. They noted that in 1924 the term Depressed Classes had been accorded a wider meaning to include aboriginal tribes and the Criminal Tribes and some other wandering and

backward Castes . . .[which had] resulted in much confusion of thought in

this Presidency, as in ordinary usage the phrase Depressed Classes is taken as meaning the untouchables . . .whereas they do not form half of this new and enlarged grouping of Depressed Classes.5

The Committee recommended that “Depressed Classes” should be used in the sense of untouchables, a usage which “will coincide with existing common practice.”*© They proposed that the wider group

should be called ‘“‘Backward Classes” [Denotation 3], which should be

subdivided into Depressed Classes (i.e., untouchables); Aboriginals

and Hill Tribes; Other Backward Classes (including wandering tribes).

They note that the groups then currently called Backward Classes [Denotation 8] should be renamed “intermediate classes.”!? In addition to 36 Depressed Classes (approximate 1921 population 1.475 millions) and 24 Aboriginal and Hill Tribes (approximate 1921 population 1.323 millions), they list 95 Other Backward Classes (approximate 1921 population 1.041 millions)*® [Denotation 6].

The Simon Commission, though it refers to “intermediate castes” and takes note of the non-Brahmin movement, makes no mention of

Backward Classes.!9 But shortly after, in the early 1930s, we find a flowering of the term. In the hearings held by the Indian Franchise Committee in the United Provinces, the term recurs in the testimony in a number

of significations. Thus a note by Dr. S. S. Nehru,

I.C.S.,

presents a list of Depressed Classes of which portions are designated 12. 14. 16. 18.

Id. Id., at 365. Id., at. [d., appendix I. The

13, Id., at 47. 15. Bombay 1930 [Starte Committee}: 8. 17, Ia.,av9. total Backward Classes population was 3.840 millions, or

14.4% of the 1921 population of the presidency.

19. Indian Statutory Commission 1930 : 1 35-37.

158

Identifying the Beneficiaries

“Vagrant Tribes” and ‘Backward Classes”?° [Denotation 6]. Presumably these are the groups who least fit the “untouchability” criterion. However,

the memorandum

of Pandit Sheokaran Nath Misra

(Deputy Collector of Fyzabad) suggests that “depressed classes should include untouchables as well as backward classes.” He equated Depressed Classes roughly with all Sudras excepting Kayasthas?4 [Denotation

4].

The

United

Provinces

Hindu

Backward

Classes

League

(founded in 1929) submitted a memorandum which suggested that the

term “Depressed” carried a connotation ‘“‘of untouchability in the sense

of causing pollution by touch as in the case of Madras and Bombay”

and that many communities were reluctant to identify themselves as depressed. The League suggested the term “ ‘Hindu’ ‘Backward’ ” asa more suitable nomenclature.?2 The list of 115 castes submitted included all candidates for the untouchable category as well as a stratum above [Denotation 4]. “All of the [listed] communities belong to nonDwijas or degenerate or Sudra classes of the Hindus.” They were described as low socially, educationally, and economically and were said to number over 60% of the population. This inclusive usage was adopted elsewhere. Travancore in 1937 abandoned the Depressed Classes nomenclature and substituted the

term “Backward Communities” to include all educationally and economically backward communities% [Denotation 3]. However, in

Madras and elsewhere the term “Backward Classes” was used to refer to the strata above the untouchables.75 The Madras Provincial Backward Classes League, consisting of the less forward non-Brahmin communities, was founded in 1934 for the purpose of securing separate treatment from “the forward non-Brahmin communities.” According 20. Indian Franchise Committee 1932 : IV, 734. 21. Id., 738.

22. Id., 826. In an interview in 1966, Mr. S.D. Singh Chaurasia of Lucknow, who was associated with the League and subsequently became a leading spokesman of Backward Classes and served on the Backward Classes Commission, recalled that he coined

the term in 1930 or 1931. Since 1974 Mr. Singh Chaurasia has been a member of the Rajya Sabha. 23. Id., 826. 24. Anonymous 1935-37 : 364. The official announcement was that communities not then classified as depressed would be included and that advanced groups among the

depressed would be reclassified. For a similar inclusive usage, note the announcement

in December 1936 that the Government of the Central Provinces constituted a committee to promote the education of the Backward Classes. Anonymous 1935-37:273.

25. A writer in the Times of India (July 1936) claimed that the estimates of Depressed

Classes population were highly inflated by the inclusion of “backward class Hindus,

such as dhobis, barbers, potters and others who are not untouchables.” See response of S. R. Venkataraman, reprinted in Anonymous 1935-37: 161. Cf. the March 1937 mani-

festo on the Christian Duty to Depressed Classes and Backward Classes, reported in Anonymous 1935-37 : 354.

Other Backward Classes

159

to the list they presented to the Madras Government in 1944, the Backward Classes comprised more than a hundred communities with about 25 million people (more than two-thirds of the non-Brahmin category and 50% of the total population of the Presidency)? [Denotation

6]. In November

1947, separate

reservations

in the Madras

services were provided for these “Backward Hindus.”’?7 In the meantime, the Objectives Resolution of the Constituent Assembly, moved by Jawaharlal Nehru on December 13, 1946, had resolved to provide adequate safeguards for ‘minorities, backward and ae

areas, and depressed and other backward classes”?8 [Denotation

5? 6?). Thus, the term had never acquired a definite meaning at the all-India level. There had been no attempt to define it or employ it on the national level, and there were no nationwide backward classes organizations or spokesmen. The term had definite meanings in local contexts, although these differed somewhat. After the listing of Scheduled Castes, the usage as a synonym for untouchables [Denotation 1] drops away. Two major species of usage emerge: (1) as the more inclusive group of all those who need special treatment [Denotations 3, 9]; (2) asa stratum

higher than the untouchables but nonetheless depressed [Denotations 6, 7]. This double usage continues today: the former in the usage of backward classes in the wide sense (including Scheduled Castes and Scheduled Tribes); the latter in the usage as equivalent to “Other Backward Classes.” It is with the latter that we are now concerned here. B. THE CONSTITUTIONAL DESIGN By the time of the Constituent Assembly the usage of the term “‘Backward Classes” to refer to some larger or smaller portion of the population deserving of special treatment was familiar in many parts of the country. In the Assembly, delegates from the north expressed puzzlement at the provisions for “Backward Classes,” (in what came to be Article 16 (4). It struck them as vague, and some thought it was meant merely as a synonym for the Scheduled Castes,?® while others were 26.G.O. 190 Public Department (Services) (Jan. 20, 1944). Saraswathi 1974: 121-22.

27. G.O. 3437 Public Department (Services) (Nov. 21. 1947). Saraswati 1974 : 124.

This Backward Hindus list at this point included 145 communities, who formed nearly

one-half the population of the Presidency. 28. Quoted in Banerjee 1948 : I, 313. 29, Pandit Hirday Nath Kunzru (U.P.) at VII CAD 680; Shri Ari Bahadur Gurung (West Bengal) at VII CAD 685; Shri R.M. Nalavade (Bombay) at VII CAD 686. Dr. Dharam Prakash (U.P.) proposed that Depressed Classes or Scheduled Castes

should be used in place of Backward Classes, since the former terms had a definite meaning (VII CAD 687).

160 Identifying the Beneficiaries concerned

that

it might

mean

:nore.*° But

representatives

from

Madras, Mysore, and Bombay assured their colleagues that Backward

Classes was a distinct term with a technical meaning. Examples were given of Mysore where Backward Classes included all but Brahmins,>1

of Madras where it referred to a stratum of non-untouchable Hindu castes,32 and of Bombay where it included not only Scheduled Castes and Tribes but others who are economically, educationally, and so-

cially backward.*3 A representative from Bihar (which had the most active Backward Classes movement in the north) explained that Backward Classes were a section of society between the highest castes.and the Scheduled Castes: ‘“‘the third [section] occupying the middle position . . .and consisting of a large portion of our people is what may be termed

the Backward

Class. ... No doubt they are not treated as

untouchables.”> Mr. K. M. Munshi assured the house that the term would include the Scheduled Castes, when doubts were expressed, and explained that it as ‘“‘necessary to find a generic term.” “It signifies people—touchable or untouchable, belonging to this community or that, who are so backward that special protection is required in the services.”35 Although one speaker pointed out that classes were not necessarily castes and that literacy might be the test of backwardness,?° it was generally envisioned that the backward classes would be communities. When asked what is a backward community, Dr. Ambedkar, defending

the draft, explained: ‘‘we have left it to be determined by each local

Government. A backward community is a community which is back-

ward in the opinion of the Government.”37 However, the matter was not as clear as this suggests. The most prophetic statement was Mr. T. T. Krishnamachari’s prediction that the clause would be “a paradise for lawyers.”38

It was anticipated, then, that the backward classes uther than the

30. Shri H. J. Khandekar (C.P. & Berar) preferred the use of Scheduled Castes lest

“the people of other castes would claim to be backward ....” (VII CAD Hukum Singh (E.P.) expressed uncertainty as to whether the term Scheduled Castes and religious minorities (VII CAD 694). 31. Shri T. Channiah (Mysore) at VIIL-CAD 689-90. 32. Mhd. Ismail Sahib (Madras) found the Madras usage objectionable not extend to “the backward classes of minority communities” (VII CAD

33. Dr. K. M. Munshi at VIT CAD 697. 34. Shri Chandrika Ram (Bihar) VII CAD 687.

36. T. T. Krishnamachari at VII CAD 699.

692.) Sardar covered the in that it did 692).

35. VII CAD 697.

37. VIICAD 702. In the light of the later analysis and the text itself, this seems rather misleading, for the “opinion” of the government is determinative only of the group’s

“under-representation,” not of its backwardness. See chap. 8, §D, below.

38. VIT CAD 699.

Other Backward Classes

161

Scheduled Castes and Tribes were to be designated at the local level. The delegation to local authoriti-s undoubtedly reflected an acceptance

(at least temporarily) of the divergence of existing practices, a desire to preserve flexibility, and an awareness of the difficulties of prescribing universally applicable tests of backwardness in view of the varying local

conditions. It may also have been presumed that “Backward Classes” were sufficiently potent politically to look out for their own interests on a

local level, and unlike Scheduled Castes and Tribes, central control of their designation was not required to insure the inclusion of the deserv-

ing. However, the central government was not entirely excluded from the process. The President is instructed to appoint a Backward Classes Commission

to investigate the conditions of socially and educationally backward classes . . and the difficulties under which they labour and to make recommendations as to the steps that should be taken by the Union or any State to

remove such difficulties and to improve their condition . . . .39

No special central machinery for supervision of programs for backward

classes is provided in the Constitution, but Article 338 (3) provides that

the Commissioner of Scheduled Castes and Tribes shall include in his duties such other groups as the President may specify on receipt of the report of the Backward Classes Commission. It is not clear from the text that this presidential specification was meant to be exclusive. Even before the Constitution came into force there was a great expansion in the employment of the Backward Classes category. Several states created such a category for the first time,*° and of those who 39. Art. 340. While the Constituent Assembly was at work, A.V. Thakkar, a member

of that body and one of India’s leading social workers, explained that the Draft visualized

a commission ‘“‘to go through the whole of the country . . .and find out which castes of Hindus, Muslims, etc., are really backward according to certain standards, educa-

tional, social, economic etc.” (Hindustan Times, Dec. 26, 1948, reprinted in Jagadisan and

Shyamla! 1949 : 338). That what was contemplated was reservation for backward “communities” is clear from the recommendations of the 1948-49 University Education

Commission 1950-51 : I, 53 that “the needs of justice to the members of the Scheduled

Castes and the communities declared to be backward by the government . . .can be met

by reserving a certain proportion of the seats . . .for qualified students of these communities. ...””

4U. E.g., the Bihar Government made provisions in 1947 for Other Backward Classes

in post-matriculation studies and, on the basis of the Ministry of Education’s list,

propounded its list of Backward Classes in 1951. (Interview with Deva Charan Singh,

one of the founders of the All-India Backward Classes Federation, and later Chairman

of the Bihar Legislative Council, on 30 March 1966.) It was estimated that these groups made up 60% of the state’s population. Educational concessions for Other Backward Classes began in U.P. in 1948. The list of56 castes was estimated to make up 65% of the

population (interview with Mr. S. D. Singh Chaurasia on 17 March 1966).

162

Identifying the Beneficiaries

already had it, several expanded the benefits conferred upon the Backward Classes. The central government was pressed to extend its scheme of post-matriculation scholarships to the Other Backward Classes,*? and when it did so, it compiled lists of Backward Classs in each state for this purpose.‘? Backward Classes organizations emerged;*? a national federation was formed.“ Apart from the central scholarships, provisions for Other Backward Classes roughly followed a regional pattern as indicated by table 18. In South India and in Bombay, provisions for Other Backward Classes flourished. In the north, there were mostly a scatter of educational concessions, and many states did not have any lists at all (other than

those compiled for purposes of the central post-matriculation scholarship scheme). 41. Deshmukh (see n. 44 below) reports that his “first concrete attempt” took the

form of personal letters to Prime Minister Nehru and Maulana Azad in March, 1948,

“requesting that some amount should be set apart for the award of scholarships to the Other Backward Classes also as was done for the Scheduled Castes and Scheduled Tribes. This request ...was accepted in principle by Government and the Other Backward Classes were added to the list for award of scholarships for the first time in 1948-49. The State Governments were then asked to prepare schedules of these castes, ... As soon as the Central Government took this decision . . .1 was nominated as

a member of the Scheduled Castes and Scheduled Tribes’ Scholarships Board . . . .” (AllIndia Backward Classes Federation 1955 : 2). Post-matriculation scholarships were provided for Other Backward Classes begin-

ning in 1949-50, with an expenditure of 2.46 lakhs, less than half of that for the Scheduled Castes. By 1953-54, the expenditure had increased tenfold to 26.51 lakhs, approximately equal to that for Scheduled Castes (RCSCST 1954 : 366). 42. The central government's lists may be found at RCSCST 1951 : 163 ff. Although I have not made an exhaustive comparison, it seems that these lists follow very closely the lists of those states which then had Backward Classes lists. Thus, for example, the

Mysore list includes not only Muslims, Christians, and Jains, but all Hindu groups other than Brahmins. The Madras and Bombay lists, however, are more restricted and do not include the higher non-Brahmin castes.

43. The Bihar State Backward Classes Federation was founded in 1947 (interview with Deva Charan Singh, as above in n. 40). 44, Majumdar (1955 : 474-75) in 1954 counted 88 organizations working for the Backward Classes in 15 states, of which 74 represented individual communities, and 14, Backward Classes in general on a local or state basis. To some extent, at least, the national Backward Classes movement represented a coming together of remnants of the non-Brahmin movements of the South and Bombay with the less successful Backward

Class movements of the north (especially U.P. and Bihar). Their re!ative success is reflected in the disparate benefits afforded by the states in table 18. The chiefspokesman for the Other Backward Classes at the national level was,Dr. Punjabrao Deshmukh of Madhya Pradesh (Minister of Agriculture from 1952 to 1957 and 1958 to 1962). He

reported that when he came into the Constituent Assembly he “mustered courage to take up the problem [of Backward Classes] and place it on an all-India plane. .. ."" The All-India Backward Classes Federation came into existence on January 26, 1950, the day that the Constitution came into force (All-India Backward Classes Federation 1955: 2).

Other Backward Classes TABLE

163 *

18

Concessions FOR OTHER Backwarp C asses, 1951-1952

Education: Reserva. Other fee concess- tions in Welfare ions, stipends, Educational Insts, STATE

facilities,

Recruitment to Government Posts

Expenditures, 1951-52 (lakhs)

Expenditures, 1958-54

.

ANDHRA ASSAM BIHAR BOMBAY MADHYA PRADESH MADRAS

x x x O x

° ° x ° x

° ° x ° x

o ° x ° x

18.04 6.22 24.95 0 13.04

4.36 249 10,26 50.15 0.7 41.25

PUNJAB UTTAR PRADESH WEST BENGAL HYDERABAD MADHYA BHARAT MYSORE P.EP.S.U. RAJASTHAN SAURASHTRA TRAV ANCORECOCHIN

° x x °

° ° o °

o ° ° x

° ° ° °

0 4.75 36.27

x ° x o

? ° ° o

x o ° °

x ° o x*

? 0 0.29 0

0 6.29 5.86 170 0 9.14 0 we 64

x

>

?

x

?

a7l

x °

° o

° oO

° °

0 0.05

0 0.05

° x O ° °

x rey ° ° °

o ° ° ° oO

° ° ° oO fo}

0

°

oO

o

x

°

x?

°

ORISSA

AJMER BHOPAL

BILASPUR

COORG DELHI HIMACHAL PRADESH KUTCH MANIPUR TRIPURA VINDHYA PRADESH CENTRAL GOVERNMENT Source: Note:

Oo

o

Oo

°

oO

°

oO

°

0

?

ve 3.47 ve

0

te

0.06 we 37

0

This table is a composite of information found in RCSCST 1952:202-5 with

introductory chapters X, XI. The 1953-54 expenditures are taken from I

BCC 146, * OBC lumped together with SC and ST.

164

Identifying the Beneficiaries

It remained unclear just how many people were included in the Backward Classes category.*5 The never-completed 1951 census enumeration, based in part on the existing state lists, counted 67 millions

(18.9% of the total population). The Planning Commission in July 1951 estimated that the Other Backward Classes were approximately 20% of the population,*” but in late 1952 noted that “these other backward classes . . .are believed to number about [54.6 millions].”’** Other observers predicted a smaller list,4? but the President in his address to the inaugural meeting of the Backward Classes Commission is reported to have mentioged the figure of 70 millions.5° A pair of Supreme Court decisions in April 1951 shattered the legal foundation of the system of communal quotas which prevailed in South India. State of Madras v. Champakam Dorairajan struck down the Madras reservations in educational institutions and by implication barred all preferential treatment outside the area of government employment.S! 45. The Backward Classes category was found to provide a convenient administrative depository for groups which were felt to deserve some special treatment, but which were not included or includable under other headings. Two and a half million “left out” tribal peoples, who had mistakenly been omitted from the list of Scheduled Tribes were included until the Schedule was amended in 1956 (RCSCST 1953 : 196). The denotified tribes (formerly Criminal Tribes) were at first listed separately for budget purposes but

were later assimilated to the Other Backward Classes for hudgetary purposes. Backward Classes lists typically included converts from Scheduled Castes to non-Hindu teligions, and several states used this category to provide some concessions to sections of their Muslim population. This is especially true of Bihar and Hyderabad. (See the lists in RCSCST 1951.) 46. The 1951 Census, unlike previous Censuses, had decided not to collect caste data,

except for Backward Backward authorities

“Special Groups,” including not only Scheduled Castes and Tribes, but Classes specified by the state governments. However, the statistics on Classes were not fully tabulated and were never published. The Census provided the Backward Classes Commission with two figures for “Backward

Classes : their enumerated population” (67 million, or 18.9% of the total population)

and their estimated 1951 population of each caste (73 million, or 20.5% of the total population) (III BCC 9). The similarity of totals conceals great discrepancies at the state

level ((see table 19). Although Professor Shah (IIT BCC 9) indicates that the census

totals are based on the existing state lists, there seem to be some cases in which this is not so—e.g., Mysore, where the census enumeration is 19.3% of the population, while the state list was closer to the Commission’s figure of 65%.

47. Planning Commission 1951 : 231.

48. Planning Commission 1953 : 243. 49. E.g., a writer in the Times of India anticipated that the Backward Classes Commission, soon to be formed, was expected to enumerate about 20 million in the Backward

Classes—that

December 3, 1952).

is, less than one-third of the census total (Times of India,

50. Dr. Deshmukh at All-India Backward Classes Federation 1954 : 5.

51. A.LR. 1951 S.C. 226. Both judgments were handed down on April 9, 195T, but

they could not have been wholly unanticipated, for the Madras High Court decided

similarly in the Champakam case on July 27, 1950 (A.IL.R. 1951. Mad.

120). A select

Other Backward Classes

165

Venkataramana v. State of Madras struck down the Madras quotas in

government posts for all groups other than the Scheduled Castes and “Backward Hindus,” confining potential recipients to those who could qualify as the “backward classes” mentioned in Article 16(4).5? These decisions caused a political furore in South India and occasioned the prompt addition of Article 15(4) to the Constitution.5>

The debate over the amendment centered on the desirability of providing educational preferences to the Backward Classes, and it revolved in part around the question of who were the backward classes.5* Although the Scheduled Castes and Tribes were equally discomfited by Champakam Dorairajan and presumably equally concerned with educational preferences, there. was little mention of them in Parliament. The question worrying the house was the identity of Backward Classes and

the possible abuse of the new provision by advanced groups. In an attempt to confine the possible meaning of Backward Classes, the wording of Article 15 (4) was keyed to that of Article 340,55 which provided that the not yet established Backward Classes Commission should list the “socially and educationally backward classes of citizens.”’ Some speakers drew assurance from the notion that the listing by Committee to report on proposed amendments was established on 16 May 1951. With the exceptions noted below, the debate on the Amendment took place between 29 May 1951 and 2 June 1951. 52. A.LR. 1951 S.C. 229.

53. In the debate on the amendment, Prime Minister Nehru remarked: “The House

knows very well and there is no need for trying to hush it up, that this particular matter in this particular shape arose because of certain happenings in Madras” (Parliamentary

Debates, Vol. XII-XIII (Part II), ool. 9615). While others concurred in locating the thrust for the amendment in Madras, one member pointed out: “it is not only the

Madras Government that is concerned with this but the whole of South India—the States of Mysore, Travancore-Cochin and even Bombay” (Shri Shankaraiya, id., at col. 9000, 17 May 1951). Dr. Deshmukh saw the regional origin as a question of time-lag: the

problem was not confined to Madras, but was bound to arise elsewhere as soon as the Backward Classes became more aware and assertive (id., at 9775). 54. The bill was referred toa select committee after some discussion on May 16, 1951. Further debate on Article 15(4) took place on May 18, 29, 30, 31, and June 1 and 2. The clause inserting Art. 15(4) was passed on June 1; the entire bill on June 2. This

amendment was one of the three major changes made by the Constitution (First

Amendment) Act, 1951. The amendment both of Art. 19 (freedom of speech) and of Art.

31 (acquisition of property) received far more of Parliament'’s time and attention. Of the three major provisions, Art. 15(4) encountered the least opposition. The clause was passed 243 to 5. (The Art. 19 amendment passed 228 to 19; Article 31, 239 to 7.) 55. The original draft of Art. 15(4) would have added to Art. 15(3), which authorized

special provision for women and children, the words “or for the educational, economic

or social advancement of any backward class of citizens” (id., at 8929, 17 May 1951).

Prime Minister Nehru explained that the Select Committee chose the final wording “because they [socially and educationally] occur in Article 340 and we wanted to bring them bodily from there” (id., at 9830).

166

Identifying the Beneficiaries

the Commission and subsequent presidential specification would be determinative.5® Others assumed that the identity of the Backward Classes would remain a question for the state governments and counselled trust in their good faith.57 Examination of the debate leaves it abundantly clear that the Backward Classes, by whomever designated and according to whatever tests they were chosen, were expected to be a list of castes or communities. Dr. Ambedkar, then Law Minister, forthrightly observed that the amendment was needed precisely because “what are called backward classes are . . .nothing else but a collection of certain castes.”5® There was considerable concern that the provision should not permit com-

munal quotas to be enjoyed by more advanced groups. While there was

discussion of the economic backwardness of the groups who deserved preferences, it was not merely the poor that the drafters and speakers had in mind.°9 (Indeed, if they had, an amendment would hardly have been necessary.) Some speakers argued that preferences should be directed to the economically poor, but the predominant concern was to provide some special treatment to offset and remedy specifically those social inequalities of caste and community which were seen as underlying and compounding economic differences. 56. While some members liked the final phrasing because they thought it confined

Backward Classes to those to be specified by the President under Art. 340 (Pandit

Thakur Das Bhargava, id., at 9719; Shri M. A. Ayyangar, id., at 9817), others objected that they were not so confined (Sardar Hukum Singh, id., at 9823; Dr. S. P. Mookerjee,

id., at 9824). An amendment to make explicit this limitation to the groups specified under Art. 340 was not accepted by the government and was defeated by the House. (id., at 9832-33). 57. Seth Govind Das, id., at 10051; Shri Venkataraman, id., at 10081.

58. Id., at 9006 (18 May 1951).

59. Prime Mininister Nehru observed that “we have to deal with the situation where for a variety of causes for which the present generation is not to blame, the past has the

responsibility, there are groups, classes, individuals, communities, if you like, who are

backward. They are backward in many ways—economically, socially, educationally—

sometimes they are not backward in one of these respects and yet backward in another.

The fact is therefore that if we wish to encourage them in regard to these matters, we have to do something special for them” (id., 9616). In spite of his reluctance to talk about

caste, it is clear that what was intended were not measures to erase all inequalities, but

specifically those which were associated with traditional social structure. “[W Je want to

put an end to. . .all those infinite divisions that have grown up in our social life. . .we may call them by any name you like, the caste system or religious divisions, etc.There are of course economic divisions but we realize them and we try to deal with

them. .. .But in the structure that has grown up . . .with its vast number of fissures or divisions. ...” The Prime Minister’s extraordinary reticence about using the word “‘caste” was not shared by all of his fellow members. Cf. the more straightforward remarks of Reverend d’Souza, id., at 9689-90; Dr. Deshmukh, id., at 9775-76.

60. Thus,

Professor K. T. Shah, the strongest advocate of an individualized

Other Backward Classes

167

The Venkataramana case had indicated that the “Backward Classes” included only those caste groups of whose backwardness there was some assurance. In spite of some unflattering observations about that

judgment, there was nothing in the proceedings to indicate that Parlia-

ment sought to overturn it.©1 Indeed, it was seen as the foundation for judicial protection against unwarranted use of the power to confer preferences.®? Upon passage of the amendment, Madras revised its former system of communal quotas to accord with the new requirements; the situation in other states remained unchanged. There wa’ no further litigation over the Backward Classes for eight years. C. THE FAILURE TO CENTRALIZE POLICY The Backward Classes Commission was established in 1953 and directed to determine the criteria to be adopted in considering whether any sections of the

people . .(in addition to the Scheduled Castes and Tribes...) should be

treated as socially and educationally backward classes; and, in accordance

with such criteria to prepare a list of such classes . . . .

approach, avowed that the backwardness to be remedied was economic (id., 9821) and proposed to eliminate the word “classes” and to add “economically” to the qualifiers of the term “backward classes” (id., 9815). Prime Minister Nehru, explaining his unwillingness to accept any of the amendments, indicated that he had no objection to adding “economically” but that to do so would put the language at variance with that of Art. 340. He then observed: “But if I added ‘economically’ I would at the same time not make it a kind of cumulative thing but would say that a person who is lacking in any of these things should be helped. ‘Socially’ is a much wider word including many things and certainly including economically” (id., 9830).

61. Although Dr. Ambedkar attributes the necessity for the amendment to both

judgments collectively, it is difficult to discern that he has a specific objection to the

Venkataramana case (id.,9006, 18 May 1951). Shri Venkatraman (Madras) closed the debate with the observation that the present amendment puts the position in regard to edycation ona par with that in government service and expressed satisfaction with the Supreme Court’s holding in that area (id., at 10081).

62. Shri Vérkataraman assured the house that there was no need for ministeriak assurances; for “there is a decision of the Supreme court. ...You have adequaté protection, even as the law now stands ...to prevent any abuse of this clause” (id.,

10081). 63. Madras listed 155 communties as the Backward Classes. See e.g., Government of

Madras, Public (Service) Department G.O. No. 2687 (dated 28 September 1953.) Nearly two decades later, the Tamil Nadu Backward Classes Commission estimated

that 105 communities that made up the then current list were 51.3% of the population

of the state (Tamil Nadu Backward Classes Commission 1971 : I, 176).

64. The charge to prepare a

list is, of course, not specified in the constitutional

provision for such a committee (Art. 340), although some listing seems to be anticipated

by Art. 338(3).

168 Identifying the Beneficiaries The directions express an expectation of centrally-established uniform standards and a central master list of Backward Classes. AccordTABLE

19

EstImMaTEs OF PopULATION OF OTHER BACKWARD CLASSES,

Ear ty

1950s

1951 Census

STATE

Total EnumeraPopulation tedPop.of (lakhs) OBC

ASSAM BIHAR BOMBAY

.

18.84 62.76 44.90

208 156 125

570.16 146.46 126.41 632.16 248.10

197.16 41.73 NA 4151 13.59

313

H6 28.5 NA 66 55

56.04"

186.55 79.54 90.75

94.13 13.68 17.53

505 172 19.3

RAJASTHAN 152.91 SAURASHTRA 41.37 TRAVANCORE-COCHIN 92.80 AJMER 6.93 BHOPAL 8.36 HIMACHAL PRADESH & BILASPUR 11.09 COORG 2.29 DELHI 17.44 KUTCH 5.68 MANIPUR 5.78 TRIPURA 6.39 VINDHYA PRADESH — 35.75 ANDAMANS & NICOBARS .30 INDIA 3568.29

32.16 15.78 4.16 0.35 1.25

21.0 38.1 45 5. 144

oe 0.69 NA. oor

30.3 NA. 02

MADRAS (.2ANDHRA) ORISSA PUNJAB UTTAR PRADESH WEST BENGAL HYDERABAD MADHYA BHARAT MYSORE P.E.P.S.U.

Source: Note:

212.48

34.94

6.48

.

0.30 7.40 674.39

Backward Classes ‘Commission

Percent- Estimated Percent- Estimated Percent ageol Pop.of == ageof_ = Pop.of = ageof’ Enumto OBC Est.to OBC Est.to Total Pop. Total Pop. Total Pop.

90.44 402.26 359.56

MADHYA PRADESH —

;

:

47 20.7 ” 189

26.15% 66.90 16.13,

289 166 45

28.66 153.21 110.10

317 38.1 30.6

63.95 15.85 18¢ 41.51 18)

46.3 108 62 66 0.7

126.81 13.56 25.56 269.10 22.66

22.2 93 20.2 42.6 9.1

122.85 13,25, 20.62

659 16.7 27

137.66 19.37 59.64

73.8 244 65.7

34.90" 11.06 2.85 0.35, 1.35

22.4 267 3.1 51 13.5

3431 12.16 9.12 2.98 2.98

224 204 9.8 429 35.2

2387 068 24t 1.08 4.12 ve

214 2.7 13.8 190 we AS we ve 205

3.15 064 3.18 201 0.35 0.69 13.56 te (1135.10

28.4 28 18.2 354 61 10.1 385 ve 318

3.79%

:

730.01

26.4

108

79.03

442

37.2

12.7

HI BCC 14-15. (All figures in lakhs). Represents estimated 1951 population of educationally backward classes, (The source is not entirely

clear regarding Assam.)

65. The expectation that there would emerge an official central list of Backward

Classes was expressed by the Commissioner for Scheduled Castes and Scheduled Tribes, who noted in his report for 1953 that “‘at present there are no ‘Backward Classes’ as such, which may have been officially recognized by the Government of India. This would be done after the recommendations o/ the Backward Classes Commission are available” (RCSCST 1953: 196). See also Planning Commission 1953 : 243.

Other Backward Classes

169

ingly, the Commission, after two years of work, presented a list of 2,399

backward

groups

and

recommended

various

measures

for their

economic, educational, social, cultural, and political advancement.

It was estimated that these groups totalled more than

116 million

members (about 32% of the total population of India).®7 (See table 19.)

This does not include women as a separate group, although the Commission recommended that all women in India made up a Backward

Class. Nor does it include Scheduled Castes and Scheduled Tribes,

who in 1951 made up over 14% and 6% respectively of the total population. It had been generally anticipated, if not universally approved,” 66. During the Commission’s deliberations, estimates of the Backward Classes had

risen as high as 70% of the total population (III BCC 9). It was proposed and seriously

considered to list “advanced” groups with all of the remainder to constitute the

Backward Classes (I BCC 48). This is discussed at length in Shri Chaurasia’s minute of

dissent (III BCC 41 ff). 67. This population figure is for only 913 of the 2,399 Backward Communities (Ministry of Home Affairs 1956 : 2). The Commission had no population figures for the

others, but it may be presumed that few were very numerous. 68. | BCC 31-32. 69. ‘Fhis expectation was evident in the Parliamentary debate over Art.

15(4),

discussed above in nn. 53-62. It was very clearly expressed in 1954 by N. D. Majumdar, a civil servant, writing in a government publication on social welfare : “The

backwardness with which the commission is expected to deal is much more special. For one thing, the commission cannot recognize individual backwardness, however widespread. For its purpose, the backwardness must be collective. ...Even this is not enough. The group must not be a purely economic group or even a social group of the modern type. For example, industrial labour. . . . The commission cannot concern itself

with these groups because it cannot trench on the ground of labour or agrarian welfare. In order to come under the scope of the commission's investigations, a social group has to be a genetic or hereditary group, and exhibit disabilities and backwardness as 4 group confined to lowliness by birth. ...{T]he commission is expected to deal with

certain undesirable consequences of the evolution of Hindu society, and developments of the same kind in the non-Hindu societies which have been influenced by the Hindu

social pattern” (Planning Commission 1960 : 219). An eminent anthropologist (who had served as Gandhi’s secretary and was later to serve as Commissioner for Scheduled Castes and Scheduled Tribes) writing in 1954 also visualized the specific purpose of the: Backward Classes Commission as the devising of measures to eliminate not all inequalities but specifically those associated with caste. “It is . . the desire and will of the Indian nation to do away with the hierarchy of caste and of its consequent social

discrimination, and prepare the ground for full social equality” (Bose 1967 [1954] : 182). He suggests that “castes whose water is not acceptable to Brahmins, and who are at the same time very poorly represented in school registers or in Income Tax and Union

Board [tax] lists in proportion to their population, should be considered ‘backward’ in the sense of the Commission’s enquiry” (Bose 1967 : 188).

70. Nehru is reported as saying at the inauguration of the Backward Classes Commis-

sion (on March 18, 1953) that he disliked the term Backward Classes, and that it was

basically wrong to label any section as backward even if they were so, particularly when’ 90% of Indians were poor and backward (1 BCC 3). This seems to mark shift from his 1951 position in the debate over Art. 15(4). (See n. 59 above.)

170

Identifying the Beneficiaries

that the ‘‘classes” designated by the Commission would be castes or communities. While indicating its desire to avoid perpetuating the evils of caste and its eagerness to avoid caste, the Commission “found it difficult to avoid caste in the present prevailing conditions.”7? The Commission felt it was “ not only correct but inevitable” to interpret its terms of reference “as mainly relating to social hierarchy based on caste.”7? As general criteria of backwardness the Commission listed trade and occupation, security of employment, educational attainments, representation in government service, and most emphatically, position in the social hierarchy.” It used these standards, not directly to isolate categories of backward persons, but to isolate backward communities. The units to which these tests were applied were for the most part caste and sub-caste groups. In identifying the backward, the

Commission used caste in two ways: first, it used caste groups as the

units or ‘“‘classes” to be classified; and second, it used the position or

standing of these groups in the social hierarchy as the principal criterion for determining their backwardness. In classifying communities, the Commission aimed to take into consideration

the social position which a community occupies in the caste hierarchy, the percentage of literacy and its general educational advancement; and its representation

in Government

service or in the industrial sphere. The

economic backwardness also had to be kept in view . . .as also the recent trends

in

its

advancement

as

a

result

of

measures . . .during the past one or two decades.”*

various

[Government]

The Commission was deluged by communities claiming to be backward.75 But it found that the paucity of data fully matched the immensity of its task.” The state governments were found not to have relevant statistics, “administrators and census officers pleaded their inability to supply the relevant material,” and figures supplied by the communities themselves were chiefly guesswork.77 The decision to de-emphasize caste in the 1951 census operations had left the Commission without any figures on the literacy, income, or occupation of the various communities.” The Commission had no facilities for generat-

71. 1BCC 41, 73. Id., 46.

72. Id., 42.

74. dd., 47.

75. The Commission received a total of 3,344 memoranda and interviewed a total of 5, 636 persons (I BCC 217). 76. In addition to a compilation of a list of Backward Classes, the Commission was subsequently charged with recommending revisions of the lists of Scheduled Castes and Scheduled Tribes. 77.1 BCC 7. 78. 1 BCC 11. Apparently the absence of caste data was the deliberate policy of

Sardar Patel, the Home Minister until 1950, who rejected caste tabulation as a device to

Other Backward Classes

171

ing data themselves. In the face of these obstacles, the Commission plowed bravely on: In the absence of reliable facts and figures, the only course open to us was to

rely on the statistics available from the various Governments and the previous census reports, and to go by the general impressions of Government officers, leaders of public opinion and social workers.”

In some cases there was no data at all and “‘the decision had to be taken on the strength of the name of the community only, on the principle of giving the benefit of doubt.”®° The Commission did not undertake to document the application of its tests to the communities on its list, for which it provided only names, traditional occupations, and estimated

population.®?

In addition to listing the Backward Classes, the Commission was instructed to investigate the conditions of all such socially and educationally backward classes and difficulties under which they labour:

and make recommendations (i)as to the steps that should be taken by the union or any state to remove such difficulties or to improve their conditions;

(ii) as to the grants that should be made. . . .®?

True to its charge, the Commission recommended a vast array of schemes for the protection and advancement of the backward, includconfirm the British theory that India was a caste-ridden country and as an expedient

“to meet the needs of administrative measures dependent upon caste division” (from a 1950 address to the Census Conference, quoted as I BCC 9). Cf. III BCC 18. 79. Id., at 8. Cf. | BCC 47, where the Commission notes that the existing backward

classes lists of the states and the Ministry of Education “formed the basis of our information.” 80. IIT BCC 9. Cf. I BCC 49.

81. Hardgrave 1969: I4Iff. provides a revealing account of the Nadars’ successful efforts to be listed by the Commission as a Backward Class. This numerous Tamil caste,

whose traditional occupation was toddy-tapping, had produced educated and prosperous strata by dint of strenuous efforts at self-improvement. The Nadar Mahajana Sangam submitted to the Commission that “barring a few individuals who can be counted in numbers, the community is illiterate, economically poor and socially boycotted.” A Nadar partisan was informed by the Secretary of the Commission that it would be ‘impossible to include ‘Nadars’ since many witnesses have told the Commission that the ‘Nadars’ are not all backward.” The Commission’s solution was to include in its list “Shanan,” a derogatory appellation for the caste which the Nadars had long fought to discourage. (The Madras Government later adopted the same device, and after a

period of some confusion, finally ordered that “Nadars” be treated as synonymous with “Shanan.”) 82. 1 BCC 2.

172

Identifying the Beneficianes

ing a number of major changes in rural life: redistribution of land,

protection of tenants, help to the small agriculturalists (credit, price supports, irrigation, etc.).83 They recommended the creation ofa separate ministry for Backward Classes’ Welfare. They proposed reservations tor Backward Classes in government service of at least 25% in Class I, 33%%

in Class II, and 40%

in Classes III and IV.85 In

addition there were various aids to the education of these groups and a reservation of 70% in medical, scientific, and technical colleges.®¢

In a last minute volte face, the chairman virtually repudiated the Commission’s work, having concluded that “it would have been better if we could determine the criteria of backwardness on principles other than caste.’’®” He finds the caste test repugnant to democracy and

inimical to the creation of “‘a casteless and classless society” (in the then fashionable phrase) by perpetuating and encouraging caste division. It is not entirely clear what he would put in its place. He recommends that backwardness be measured by residential, economic, educational, and

cultural criteria.88 Apparently he repudiates not only caste standing as the test of backwardness but also the use of caste units. At several places

in his covering letter he suggests that only individuals and families should be the units whose backwardness is ascertained.®9 Elsewhere,

however, he recommends policies which seem incompatible with this position (e.g., that where an income test in employed, members of backward communities should be given priority). The chairman’s last-minute desertion foreshadowed (and perhaps augmented) the negative reception that awaited the report.?! The spectacle of numerous groups vying to display their backwardness, the feeling that caste classifications were divisive and unfitting, the casual83. 1 BCC 511F

84. 1 BCC 143 ff.

85. I BCC 140. The Commission indicated that it was inclined to recommend reservations in proportion to population, where educational attainments permitted, but adopted the minimum figures to leave sufficient scope for highly qualified candidates. 86. I BCC 125. This recommendation was meant to apply until accommodation was available for all students qualified for admission. 87. I BCC xiv.

88. I BCC xiv-xv.

89. I BCC iv, xiv.

90. I BCC viii. Cf. his suggestion that the most backward communities should be treated separately to prevent all benefits from going to the most advanced among the backward (I BCC xxi ff).

91. In addition to the Chairman’s repudiation, three other members of the Commission, including the Secretary, filed minutes of dissent objecting to the caste basis of

classification (Minutes of Dissent of Dr. Anup Singh, Shri Arunangshu De, Shri P.G.

Shah). Two other members dissented on the ground that the Commission had not gone far enough : Shri T. Mariappa merely objected to the failure to include urban Lingayats and Vokkaligas in Mysore; Shri S. D. Singh Chaurasia set forth a detailed proposal for

equating Backward Classes with Sudras. The various Minutes of Dissent make up Volume III of the Commission’s Report.

Other Backward Classes

173

ness of the Commission’s application of its criteria, the vastness of the number they found backward, and the expansiveness of the preferences they proposed exposed their work to widespread criticism.9? The Commission’s report was laid on the table of both houses of Parliament on September 3, 1956, accompanied by a withering critique from the Minister of Home Affairs, expressing disappointment with its

criteria and

its conclusions.??

The emphasis

on caste, the Minister

asserted, displayed the “dangers of separatism.” Not only was the caste basis unfait to the backward outside these communities, but the

caste system was undeniably “the greatest hindrance in the way of our progress toward an egalitarian society, and the recognition of specified castes as backward may serve to maintain and perpetuate the existing distinctions on the basis of caste.”95 The Commission’s standards other than caste were ‘obviously vague.” The very expansiveness of the Commission’s list undermined its usefulness, for if everyone “‘barring a few exceptions, has thus to be regarded as backward, the really needy would be swamped by the multitude and hardly receive any special attention. . . .”97 Thus the Commission failed to find “positive and workable criteria.”%* “[F]urther investigations will have to be

undertaken so that the deficiencies that have been noticed in the

findings of the Commission are made good and the problem is solved with due regard to the requirements of Art. 340... .”% The state governments were requested to undertake ad hoc surveys to determine the numbers of Backward Classes and in the meantime to “give all reasonable facilities” to the Backward Classes ‘in accordance with their existing lists and also to such others who in their opinion deserve to be considered as socially and educationally backward in the existing

circumstances.”!°0 Thus the matter went back to the states; the

Commission’s report remained on the table, and in spite of occasional agitations, was not taken up by Parliament until 1965. When the replies from the state governments “were not found helpful,” the central government requested the Office of the Registrar General to conduct ad hoc surveys to determine suitable criteria.) It was hoped that the Registrar General could determine occupational tests of backwardness. On the basis of a three-state sample survey, an 92. Opposition to the way in which the Commission proceeded might have been

anticipated in view of the Government's decision to ignore caste in the 195] Census (see n. 78 above) and from Prime Minister Nehru’s remarks at the Commission’s inauguration (see n. 70 above).

93. Ministry of Home Affairs 1956 : 3-4. 95. Id. 96. Id. 97. 1d., at 4. 100. Id., at 4-5. 101. RCSCST 1957-58 : 1, 9.

94. Id., at 3. 98. Id.

99. Id.

174

Identifying the Beneficiaries

occupational test with an income ceiling was tentatively proposed.1©2

The Home Ministry’s suggestion that the states adopt such an occupational criterion enjoyed a mixed reception.! However, this effort collapsed when the Deputy Registrar General reported that it would be impossible to draw up a precise and complete list of occupations whose practitioners were socially and economically backward.1% Those who entertained hopes of generous preferential treatment for the Other Backward Classes (OBC) found some encouragement in the late 1950s. State expenditures on schemes for the OBC increased (particularly in education,-where they were substantial),1°5 although the expansion of the central government’s scheme for postmatriculation scholarships slowed down after 1955-56. Apparently at some point in 1957 the Government of India instructed all state governments to extend the benefits of various welfare schemes to the OBC,

using the Ministry of Education lists, pending finalization of a list of OBC. Furthermore, it requested the state governments to fill unused

vacancies in seats reserved in educational institutions for Scheduled Castes and Tribes with students from the OBC,!® a measure found

very gratifying by the All-India Backward Classes Federation (AIBCF), which observed that “‘at long last the voice of the Federation is being effectively heard.””?7

102. RCSCST_ 1958-59: I, 11-12. The Census authorities indicated that “it mightbe

possible to draw up a list of socially and educationally backward occupations on the

basis of: (a) any non-agricultural occupation in any State. . in which 50% or more of the persons belong to the Scheduled Castes or Scheduled Tribes; or (b) any nonagricultural occupation in which literacy percentage of the persons depending thereon is less than 50% of the general literacy in the State.” It was suggested that this test be used in conjunction with an income limit of Rs. 1,000 per family. 103. RCSCST 1958-59 : I, 12. 104. Letter of the Ministry of Home Affairs to Chief Secretaries of all State Government/Union Territories, 14 August, 1961, reprinted at RCSCST 1960-61 : II 366.

105. A number of states gave fee concessions to Other Backward Classes (RCSCST 1956-57: IT, 65). 106. Beginning i in 1957, the Ministry of Home Affairs and the Education Ministry

recommended

that vacancies in reservations for Scheduled Castes and Scheduled

Tribes should be filled with Other Backward Classes candidates provided they met minimum qualifications. Letters from Ministry of Home Affairs No. 10/41/57-

dated SCT(IV), dated 30 July 1957; Letter No. 10/32/57-SCT(IV) , 11 June 1958. This provision was withdrawn in letter No. 28/6/61-SCT(I), dated 4 Jan. 1962, after the

government had decided not to draw up an all-India list of backward classes. The latter two letters are reproduced in Planning Commission 1965 : 234-35. 107. All-India Backward Classes Federation 1958 : 6-7. In December 1958 the Federation’s executive committee reiterated its thanks to the Government and. its appeal that the same be done for unfilled reservations in government posts (All-India

Backward Classes Federation 1959 : Appendix vi). In an interview-with Mr. Jai Narayan Singh Yadava (10 April 1966) he estimated that under this arrangement Backward Classes candidates obtained roughly 6% of places in the affected institutions.

Other Backward Classes

175

By the beginning of the 1960s the tide was running strongly against definition of the Backward Classes by community. Opposition within the government! was augmented by criticism from academics and much of the national press, who voiced a common suspicion of the caste

criterion.’ For the first time since 1951, a court intervened to strike

down a scheme for Backward Classes in a decision widely acclaimed as

a blow

at casteism.1!°

Revulsion

from

communal

criteria was

reinforced by reports of their abuse. In a situation where many thoughtful persons were increasingly concerned about the dangerous potentialities of social cleavages,'"! the alleged divisive tendencies of the communal

criterion seemed

integration.11?

a serious threat to national unity and

108. In 1959 the Study Team on Social Welfare and Welfare of the Backward Classes

recommended the abolition of the OBC Category and its replacement with economic criteria for selecting beneficiaries for government schemes. The recommendation is

contained in the Introduction (dated July 1959), but the Committee’s position is hardly clear in the body of its report (1959 : 7; cf. id., 127).

109. Indian is time criteria

See e.g., Srinivas’s 1957 presidential address to the anthropology section of the Science Congress. After a dispassionate review of developments, he suggests “it to give serious thought to evolving ‘neutral’ indices of backwardness. .. .The of literacy, landownership and income in cash or grain should be able to

subsume all cases of backwardness.” But other less disinterested critics perceived the

development of a ‘vested interest in backwardness” and all sorts of dire effects on national integration and efficiency. See, e.g., Mehta 1963. (There were few to argue the other side. For a rare example, see Subbiah 1963.) For a convenient review with many references to the scholarly and polemical literature of this period, see Barnabas and Mehta 1965. 110. Ramakrishna Singh ov. State of Mysore, A.1.R. 1960 Mys. 338. The Court’s judgment (discussed below passim in chaps. 7 and 8) on 18 September 1959 was acclaimed as a blow at casteism. E.g., the Times of India’s Mysore Newsletter for September 23, 1959,

began: “The Mysore High Court has done what no politician of this State would dare to do. It has put a brake on the race to seek educational privileges by all and sundry.” The

next day the decision was lauded in an editorial which observed: “The only rational test for backwardness is a person’s economic circumstances and not his caste... . The

Mysore and similar orders elsewhere show how caste-ridden the Congress is despite the anticasteist protestations of its leaders.” 111. In 1960 Selig Harrison’s book appeared with its gloomy reflections on the disintegrative forces of language, region, and caste and its serious questioning of whether India could remain united. Although he points out that caste tests of backwardness consolidate caste consciousness, he does not assign them any major disintegra-

tive role: “[1'Jhe constitutional guarantees only institutionalize a group awareness that

would in any case exist to a great extent. Since the upper castes often refuse to recognize the achievement of an individual member of a low caste who may, for example, acquire

education, the individual invariably concludes that mobility on any significant scale must be a group phenomenon” (Harrison 1960: 104). 112. E.g., in 1961 the Ministry of Education’s Committee on Emotional Intergration

received over 1,200 replies in a poll of highly-placed educators. Over 70% were convinced that divisive forces had increased since Independence. When asked to

176

Identifying the Beneficiaries

While a “casteless and classless society” remained the avowed aim of the Congress and a wide section of the intelligentsia, there had been a subtle shift in notions of how this aim was to be pursued. A decade before, it was widely thought that special redistributive measures were required specifically to offset inequalities associated with caste, even while general development programs addressed other aspects of inequality. The notion of caste differentials as themselves a significant form of inequality deserving of special governmental attention to eliminate their effects gave way to a notion that the salient differences were

economic; specific redistributive measures directed at caste differences were not necessary, since overall development would raise the general

level.13 Recognition of caste differences in order to offset their effects

was

replaced

by an enhanced

reluctance to recognize them at all;

indeed, any recognition of such differences was seen as itself a violation of egalitarian principles and productive inequality. In May 1961 the cabinet decided that no national list of Other Backward Classes should be drawn up, and the states were informed that in the view of the Government of India “it was better to apply economic

tests than to go by caste.”?"* At the end of May a Conference of Chief Ministers to consider matters relating to National Integration “agreed that economic backwardness rather than community or caste would provide an appropriate criterion for giving aid to individuals in matters

of education including professional and technical training.”"15 In August 1961 the Home Ministry informed the state governments of the Centre’s decision not to list Backward Classes. In the Ministry’s view,

the very expansiveness of such proposed enumerations as that of the Backward Classes Commission militated against them: “If the bulk of

identify the “disintegrative forces,” the highest number chose “‘casteism/communalism” (62% )—-significantly, more than chose religious bigotry, etc. (39%) or regionalism, etc. (44%) (Ministry of Education 1962 : 189). Cf. the Times of India’s view

that if the use of preferences were not to defeat the establishment of a “casteless and classless society” the government must amend the Consitution to eliminate the permissibility of communal unis in distributing preferences (Times of India, August 30, 1961). 113. Thus the Estimates Committee of the Lok Sabha (Forty-eighth Report) proposed

that the “weaker sections of society” should be defined by economic criteria as well as

edycational and social backwardness. “Progressive emancipation from economic back-

wardness should help the people belonging to Scheduled Castes and Scheduled Tribes to overcome their social backwardness” (quoted at Study Group on the Welfare of the

Weaker Sections of the Village Community 1961 : I, 12). The shift in emphasis is

apparent in the October 1961 report of the Study Group (chaired by Jayaprakash Narayan). The “weaker sections” are identified as the economically backward— practically all of the village community. Caste inequality is a contributing cause of economic backwardness which is to be addressed by remedial measures framed along lines ofa means test (1961: [, chap. 2). 114. Ministry of Home Affairs 1962a : 38

115. dd.

Other Backward Classes

177

the country’s millions were to be regarded as coming within the cate-

gory of Backward Classes, no useful purpose could be served by separate

enumerations of such classes.” Furthermore, the caste criterion was

objectionable: “‘the remedies suggested on the basis of caste would be worse than the evil of backwardness itself.”!1° But subsequent efforts to discover usable criteria on economic lines “did not yield any useful results.” Whereas in 1956 the Home Ministry had acknowledged an obligation to compile a list in accordance with the requirements of Article 340, it now pointed out that the Constitution did not require the Centre

to draw up a list. Since, even if it were to do so, “it will still be open to every state government to draw up its own lists, any all-India list would

have no practical utility.”1!7 More important, the “crying need” of the day was social cohesion and emotional integration. There was the grave danger that different treatment of the “backward” would foster divisive tendencies and would undermine efforts for general economic uplift and the reduction of disparities between different classes. The states were urged to emphasize the expansion of welfare and educational benefits to all of the poor, employing economic rather than communal categories.118 The withdrawal of the central government from involvement in preferences for the Other Backward Classes was confirmed by the omission

of any provision for them in the central sector of the Third Five Year Plan.!? The amount contributed by the central government for postmatriculation scholarships for OBC, which had increased steadily since 1949, was frozen at the 1958-59 level.12° And beginning in 1963, the state lists of OBC used in administering this scheme were abandoned in favor of a test of family income.?21 116. Letter of Minister of Home Affairs to Chief Secretaries of All State Govern-

ments/Union Territories, 14 August, 1961, reprinted at RCSCST 1960-61 : IT, 366.

117. Id.

118. Jd. These views were endorsed by the Ministry of Education’s Committee on Emotional Integration (1962 : 45) which concluded that “the time has now come, in our opinion, when increasingly assistance should be based on economic criteria. In some

states powerful groups have exploited ‘backwardness’ to their own advantage and to the detriment of society as a whole. ... It also results in the long run in making the

backward classes less self-reliant than they should be.” 119. The Other Backward Classes category had become, from the Centre’s point of view, not a stratum of the population but a catch-all category for groups other than the Scheduled Castes and Tribes felt to be deserving of goverament help. In 1960 the Commissioner for Scheduled Castes and Scheduled Tribes noted that Denotified Communities, Nomadic Tribes, and Neo-Buddhists were ‘‘decidedly the groups to get Government help as Other Backward Classes during the Third Five Year Plan period” (RCSCST 159-60: 1,9). 120. RCSCST 1959--60:238. 121. RCSCST 1964-65: 155, 158, 159.

178 Identifying the Beneficiaries The central government’s campaign for economic criteria in the

states was given added impetus by the first (since 1951) intervention of

the Supreme Court into the matter of who are the Backward Classes. In September 1962 the Supreme Court struck down the Mysore Backward Classes list, whose defects included exclusive reliance on caste standing as a measure of backwardness, adding the onus of constitutional dis-

repute to the, caste criterion.122 The Court’s judgment, which warmly commended economic tests, was widely acclaimed and widely interpreted (mistakenly, as we shall see in chapter 7) as outlawing entirely the use of caste tests. This case marked the emergence of the judiciary as the institution within which the problem of who are the Backward Classes was most carefully and coherently addressed. In 1965, when the Report of the Backward Classes Commission was finally discussed in Parliament, the central government’s spokesman firmly reiterated its opposition to communal criteria. Caste criteria were not only administratively unworkable, but were contrary to the “first principle of social justice” in their unfairness to the other poor. They were contrary to the Constitution, would perpetuate caste, and would create in the recipients both vested interests and a sense of helplessness. The

Centre endorsed economic criteria, but refrained from

enforcing it on the states, preferring ‘“‘the path of persuasion.” Eight states were said to have adopted the economic criterion.1% Backward Classes organizations continued to campaign for greater benefits and-for a revival of central responsibility and interest in the Backward Classes.'?5 In particular, they petitioned for a restoration 122. Balaji v. State of Mysore, A.1.R. 1963 S.C. 649.

123. LSD (3rd Series), Vol.48 No. 16, cols. 3973-3976 (25 Nov. 1965). 124. The list included two states (Gujarat and Maharashtra) that had retained the caste test outside the scholarship area; one state (Punjab) that not only retained the

caste test outside the scholarship area, but employed a

list of communities in the

scholarship area as well, albeit in conjunction with an income test; one state that had no

schemes of its own for Other Backward Classes (Orissa); and two states that had none outside of scholarships (Assam, West Bengal). The only states with economic tests and

a high level of benefits for Other Backward Classes were Mysore and Andhra Pradesh. The latter reverted to a caste test less than a year later, and the former in 1977.

125. Backward Classes groups themselves differ in their definitions of the Backward Classes and their population estimates, but they tend to be rather expansive—often including all religious minorities as well as Hindu groups. Thus speakers at the 1966 conference of the All-India Backward Classes Federation constantly used the figure of 85% of the population—this was intended to include the Scheduled Castes and Tribes, Muslims, Buddhists, and Christians as well as the Other Backward Classes who were

said to make up 40% of the total population. (Author’s notes). Welcoming the particip-

ants in a 1969 Seminar on Backward Classes and the Fourth Five Year Plan (organised by the All-India Backward Classes Federation), Mr. V. P. Singh mentioned the figure of 60—70 million Backward Classes. The chief guest, A. N. Jha, Lieutenant Governor of

Delhi, used the 85% figure while warning the delegates of the futility of attempting to

define Backward Classes (Backward Classes Review 1(3):18).

Other Backward Classes

179

of the caste basis, for implementation of the long-ignored report of the Backward Classes Commission, and for creation of a ministry to attend to the problems of the Backward Classes.'2° Organizations of particular communities and some composite Backward Class organizations at the local level (especially in the south and in Bihar) continue to flourish, but many of the local composite organizations have become moribund.’?’ After a period of desuetude, the national AIBCF had a brief revival in the late 1960s. D. THE VARIED STATE PATTERNS

As the central government’s role in designation of Backward Classes moved from an attempt to prepare a single nationwide list to suggesting

standards

to the states, courts involved in litigation over Backward

Classes lists made it clear that the Constitution did not confer on these 126. E.g., the resolutions of the All-India Backward Classes Conference held in New

Delhi in March 1966: “The Federation is of the firm opinion that even though ultimately a class of people are to be judged by the economic well-being, in the transition period when large sections suffer from social disabilities in addition to economic poverty it would not be in the national interests to determine backwardness in terms of economic criteria alone. Social backwardness—as laid down in the Constitution—can only be

determined in terms of castes and communities to which the stigma applies as a whole and till the society becomes casteless it only injures the interests of the ‘socially backward’ to determine backwardness in terms of economic criteria only.” The Federation supported generous help for the “poorer sections of socially advanced classes” but asserted that while economic tests might be appropriate for these, “for the Backward

Classes, the criterion should be ‘Social.’ ” Cf. Sathi (n.d.) who expresses the view,

widely shared in Backward Classes circles, that Art. 340 represents a definite (and unfulfilled) commitment toa specific stratum of communities. M. S. A. Rao (1968: 781)

reports that at the Conference of the Yadava Mahasabha in 1968, implementation ofthe

BCC Report and ‘the revival of the caste criterion were among the most prominent demands. 127. All of these composite groups are oriented to the “caste” basis, and there is some

indication that they have declined in direct correspondence to the slackening of the flow of benefits along these lines. The desuetude of the All-India Backward Classes Federa-

tion can be roughly measured by the fact that the last ofa series of nicely printed reports begun in 1955 appeared in 1961. (There was a quickening of activity after the 1967 elections. The Federation began publishing a publication called the Backward Classes Review, whose first issue appeared in December

1968.) The Mysore State Backward

Classes Welfare Association, established in 1960 (apparently in response to initiatives from the Federation) and devoted mainly to organizing and representing the “more backward classes” (i.c., not the Vokkaligas and Lingayats), became defunct after Mysore abandoned the caste list (interview with Mr. Venkataswamy, Secretary, in

Bangalore, 1966). Organizations of particular communities have proved much more. durable. Thus, ironically, the government’s policy against communal criteria has discouraged the more broadly based composite organizations and left the field to organizations that are communal in the literal sense. I encountered no organized groups of recipients of benefits distributed along economic lines.

180

Identifying the Beneficiaries

central proposals any conclusiveness in the identification of the Back-

ward Classes.128 While the State might make special provision for “any” Backward Class, it is under no constitutional obligation to make

such provision for every class designated as backward by the Backward Classes Comisssion or by any other agency. Indeed, the State may not rely on the findings of the Backward Classes Commission to establish conclusively that a given class is backward.'29 Conversely, preferences may be given to a group which does not appear on the Commission’s list. State power to provide preferential treatment to the backward is not limited to those listed by the Commission or any other agency. The original expectation that the Commission’s list or some

part of it would be confirmed by presidential specification has gone unfulfilled. This failure does not invalidate other listings of Backward

Classes, for the courts have held that presidential specification would

not exhaustively define the Backward Classes for purposes of preferences under Articles

15(4) and

_16(4).3° Central proposals may, of course,

carry some persuasive weight with the states (or, as guides to “‘reasonableness,” with the courts).151 But it is the “State” in the broad sense

128. But cf. Professor Tripathi’s argument (1972 : 206) that the way to vindicate rights against caste discrimination is to confine the power to make provisions for Backward Classes exclusively to parliament. (Tripathi 1972 : 206). Presumably he would then confine the power to designate Other Backward Classes to the center as

well. On the failure of textual arguments to this effect, see n. 130, below. 129. Jacob Mathew v. State of Kerala, A.1.R. 1964 Ker. 39 at 56. 130. Art. 338(3) served as the cornerstone of an ingenious argument that the constitu-

tional plan for the designation of Backward Classes is analogous to the exclusive central

control over designation of Scheduled Castes and Scheduled Tribes: that Articles 340

and 338 (3) provide for exclusive central designation corresponding to Articles 341 and 342, the only difference being the additional step of the Commission Report. This argument was rejected in Ramakrishna Singh v. State of Mysore, A.1.R. 1960 Mys. 338 at 342. The court, noting the absence of any provision corresponding to Art. 366 (24) or (25), pointed out that there is no indication that the presidential specification of Backward Classes for purposes of the operations of the Commissioner of Scheduled Castes and Scheduled Tribes was to define Backward

constitutional purposes.

Classes exhaustively for all

Another variant of this argument for exclusive central control over designation of Backward Classes put forward in Balaji v. State of Mysore, A.1.R. 1963 S.C. 649, 657-8, relies on the provision in Art. 340 for the appointment of the Backward Classes

Commission. It was argued that “Backward Classes” could be designated only by the President in pursuance of recommendations of the Commission. Noting that Art. 340 (I) refers to measures that might be taken by the states and the Union government, the

Supreme Court found that exclusive presidential power in this area was not contemplated by the Constitution. (This conclusion is reinforced by the fact that the Commis-

sion is not a continuing body with power to revise its list. Nor is there any provision for revision ofa presidential listing of Backward Classes.) This argument (that reservations can only be made for Backward Classes identified by the Commission contemplated by

Art. 340) is advanced again in Pradip Tandon v. State of U.P. A.1.R. 1975 All. 1,6, and encounters a similar rejection. 131. Thus, in the Balaji case, the court took careful note of the central government’s

Other Backward Classes

181

of all. governmental organs that retains the power of designation. 3? Preferences and who is to receive them may be provided by executive (as well as by legislative) action—and they almost always are.34

With the abandonment of central attempts to define the Backward Classes, and the relinquishment of whatever control might accompany

central funds for Other Backward Classes, the matter reverted to the

states. The composition of the OBC, the scope of preterential programs, and the level of benefits continued to vary widely from state to state. Some of the varying state practices for selecting Backward Classes are summarized in table 20. For all its incompleteness and oversimplification, this list gives us a rough profile of who were the Backward Classes in the late 1970s. Later chapters will discuss the standards by which such classes were selected. But at this point it is worth noting a few of the salient features of this profile. Caste and communal units remain the predominant “‘classes” deemed backward. Caste lists range in magnitude from those which include a substantial portion of the state’s population to those constituting a narrow stratum just above the untouchables. Income tests are also employed in many cases, sometimes independently and sometimes in conjunction with communal units. The level of benefits ranges from none through scholarships and fee concessions to an array of reservations in government posts and medical college admissions. There is important regional variation. For convenience we may think of three contiguous groupings. First, there is what we might call the peninsular bloc comprising the four Dravidian states (Andhra Pradesh, Karnataka,

Kerala,

and Tamil Nadu)

and Maharashtra.

In these

states the Backward Classes categories have a long history descending from pre-Independence arrangements; there are a wide range of benefits; and (except in Maharashtra) a major segment of the population— from 38% to 55% —is included, and a major segment of scarce opportunities are reserved for them. In stark contrast is what we might call the eastern-middle band, stretching across India from Assam in the northeast through West Bengal and Orissa, across Madhya Pradesh to Rajasthan (Gujarat recently

moved

from

this

category

to

a

pattern

like

that

of

suggestion (incorporating the recommendations of the All-India Council for Technical

Education) that reservations in higher education should not exceed 25% (or, in exceptional areas, 35%). A.I.R. 1963 S.C. at 656. 132. The “State” in Arts. 15 (4) and 16(4) must be interpreted as defined in Art. 12 as

including governments as well as legislatures and state and local as well as central authorities. 133, This was early established by Ramakrishna Singh v. State of Mysore, A.1.R. 1960

Mys. 338 at 343. The latest assertion of this point is in Triloki Nath Tiku v. State of Jammu and

Kashmir {II], ALR. 1969 S.C. lat 2.

‘Sapqununuos 6 °% suonedn3 *|

Da

——Ajyeuoneonpo pur Ayjeo0y,

8261

tol

L61

2a

TN tN

IN

Da

2299009‘f

ty

IN

TN

seare s00d

TN TN

8

%z iE 0

8261

TN

TN TN

oa

IN

TN

TN

4

(090 1828| AL? TLL 956e2 uy Yonowodd oun Vou

TN

4 Wo 6% %

TN

TN

Sonrunuiwo parsy ZB

wey es awoouy

D8

——_Ayeououory

BL6r

zL61

De

sux

%S

24/008" PH

“ediyssejoyog iN

TN

a

‘TN

TN

2u0N

wWnHSVX vonnvi

Hsaavud

IWHOVAIH

VNVAUVH

ivavino

tL61

%ot

U0

45%

oa

oa

oa

TN

BUON, %01

UN

1 Sse)

ALL 82567:

TN

11

ALP 1 266e> 1 WOHOWOY tuaungnasay wong AL

AW a su

2

: TN DEO/LS/OS HE

——_Alfestuouorg

“A

uvHiE

saqno “Aqyenuue 00021

6461

%E

‘sdiqeajoyss

SE

0061 “Ry 29pun

IN

suoTEEaoUGD 295 tediyermoyrs,

sy21904

TN

oa

Hsadvud VaHONV

aa

%05

Aqreaqusoucsg

ust

Jopun swoour pum TY 000°%1 “Fy LET s82q 1 enue a0ym sontunumusdD BEI Jo THY Jo craquiayy

009'6 “a J9pan awoout ys 2004) 305 Ao S1yau2q .2210,,

SISOd = ‘dOd 40 asa

%8E

40,

auoou Yat TTY

24

aLVLS

TN

‘somrununuse> poisy 26

AVA

TN

VI¥aLLIIO

anowd do aINVN

wa

ov

SLNAWNOO

v

agunos

WIHLO NOISSINGY a93TIOO ‘‘WoIdan

S061 ZLV] ZHL NI SA8SV1E CUVMAOVE WAHLO 40 TIAONY Y

0 ATAVL

WIWIOT) SaSSE] PlLMxIVG NPeN

MYL °1. 0861 YBUIS ‘d 21 EB: 8261 S49PerT] UF

vonowosg quauninusoa pentuy

“g wuunyoo ur are uotuaauo:)|

TN TN

%E

%IE

TN

IE

tN %SI %91

[EL

GI ‘uomsarnsUL 322F13M:'T duounsedag]

| uoneonpg Jo aresonvasigy Ves oq yseptig Bypuy :\ :s207n0g

101g *2 707 Groyy ut [euoRMaNSUOIUN Se UMOP YNLSy

i tat $i jPNg IUeULIO} Og ‘DUBy]9.\4 WY 6L6L “WTW “do

8161 HSddVYd AVLLA

4461

AGN TIN

seasy IIH

zi61

pe

Da

avinna

ue aperoog

Pur Ajpepog

Da oa

280 240

L261

Le6l

wL61

6!

9961 1261 VULHSVaVHVW

Aqyeuoneonps

Aqjeuoneonpe

Da od AljeotMoURrs

6161

VAVLYNYV

vivugy

De

LL61,

BL61

VAVIVHOaW

TN

(ousig yewny

“|,PPe snd)

cot Jo 1st] souunususos

eAuY 20) 281]

[sonununuos

SomtunUwODJo 18] TN

porsy jo siaquiayy,

it

3 og" a 4apun Sunoout jenuue yam Ty 29 8 S215 621 “sontuNUIIHOD 9] JOIST]

Biol:

Da

zt TN

aq Dnunusuoo 86 028°

%tl TN

TN

s]ferulOUar 2y) Jo) uoEALasON wag Ae MG TEOY OF 210 "RODE “dni wep so uonressosos 61 241 Jo uoneoidnp fenzed e snyy pus sassy) pemyreg snf parou sasse]:) paemypeg s|feswouosg ay) pue sdnoud porst] 24) uoosiog YsinSuNsipor sprey 11 FEY UL UdWaTEIS19A0 UE st UNBy sty TEU ajqeqaud $11], S2I0Ni 2 Ww *] {s]2180H,

Buuren jepods

TN

ow la]

TN

tN Ot

TN

%S “pia "sue “sdiysaejoyg TN

%0+

TN

TN

%S

2

a

a

sequsn 2tu0s pur seyppng-aen Joy suo -avsaryo aatsnpoxy

IN

TN

TN

fF i)

rN

TN

TN

o

sadiysreqoyos

%S%

suro}

TN

1

{Ho

wN

TN %SI

TN TN

e%0t — [%SH)

RT ae TN

TN

TN | Weuntuvsa pang

2 a IN}.L°9

184 Identifying the Beneficiaries Maharashtra). In these states there is no significant use of the Other Backward Classes category. The northern tier of states displays an intermediate pattern. Jammu and Kashmir, with its history of communal quotas, resembles the southern pattern. The admixture of geographical criteria there is also found in Punjab and Uttar Pradesh along with use of communal categories. Bihar, like Jammu and Kashmir, approximates the peninsular pattern. The Backward Classes are selected on a communal basis and make upa sizable portion of the population, but the benefits have not been as extensive as in the south. Extension of benefits to reservations of government posts in 1978 set ofa political crisis. What the central government tried in 1965 to portray as a trend toward substitution of economic for communal criteria was even then largely a rhetorical artifact albeit one built around two substantial items—the abandonment of communal units in Andhra and in Mysore." From the vantage point of the late 1970s this “trend” appeared as a transient deflection from the main line of development of Backward Classes lists. That line of development has been one of continuing use of communal units, but with increasing refinement and restraint. Under pressure from the courts, almost all the state governments that made extensive use of the Backward Classes category set up commissions to identify the Backward Classes. From 1965 to 1976, such

commissions

reported

in

Kerala,

Jammu

and

Kashmir,

Andhra

Pradesh, Tamil.Nadu, Kerala again, Karnataka, and Gujarat.¥95

All of these commissions, and the government orders based on their recommendations, use communal units to designate the Backward Classes. But Jammu and Kashmir relies heavily on occupational and territorial groups, and Kerala employs an income cut-off. The selection of communities is more sophisticated: the commissions attempt to assemble (and sometimes generate) evidence about occupation, income, and education as well as status and disabilities. Most of them strive to eliminate the well-off. In some cases, the number of groups

designated is more modest than earlier.456 Benefits are recommended 134. See n. 124 above.

135. Kerala (Kumara Pillai Commission) 1966; Government of Jammu & Kashmir

1969; Andhra Pradesh Backward Classes Commission 1970; Tamil Nadu Backward Classes Commission 1971; Kerala (Backward Classes Reservation Comniission) 1971;

Government of Karnataka 1975; Government of Gujarat 1971. The Bihar report by the Mungeri Lal Commission is described in Blair 1980. Two other reports that are not strictly part of this genre might be mentioned in connection with these: Punjab Welfare Department 1966 and the report of the Chhedi Lal Sathi Commission on the “most

packward” classes in U.P., described in Chhotey Lal v. State of U.P., A.1.R. 1979 All. 135

( Ie Eg, in Andhra Pradesh the number of groups on the list fell from 139 in 1963 to 112 in 1966 to 92 in 1970. See State of A.P. v. Balaram, A.1.R. 1972 S.C. 1375 at 1387 ff.

Other Backward Classes

185,

for a limited span of time, and there is concern about termination and

re-assessment. The best of the recent commissions display an industry

and a sophistication in gathering and analyzing data that is a far cry from the uncritical and suppositious compilations of some of the early reports on backward classes. 157 As we shall see, state governments are

not always receptive to the critical assessments of their commissions. Perhaps some of this state activity was in response to the animadversions of the Centre on communal criteria, but more directly and palpably it was a response to the courts. In some cases it was a response to courts actually striking down schemes; in others, to courts threatening to do so if they were not reformed; in still others it was to the implicit threat of litigation. The states aligned their schemes with the prescriptions of the courts rather than with the pronouncements of the central executive. In part this may be because the Centre offered so little, apart from commending the income test. In retrospect it is surprising that the Centre never attempted to formulate a workable quantitative standard for selecting backward communities—e.g., those groups whose average income and average literacy were less than one-half of the state average.

It may have been felt that practically such a course was foreclosed by the decision to omit castewise data on the censuses of 1951 and 1961,

but the absence of such attempts suggests that the animus against

communal

tests eclipsed

any

consideration

of making, them

more

workable. What emerges from the interaction of state governments, commissions, and courts are lists of communal groups, with some admixture of geographic and income factors, chosen on the basis of low status, low educational attainments, and poverty. We might generalize very tentatively about the magnitude of the Backward Classes category: the lists tend to converge on something like the second and third lowest quintiles of the population—in terms of “traditional” status—assuming the Scheduled Castes and Tribes make up roughly the lowest quintile. Northern and southern states arrive at this position from very different Starting points. In the peninsula (and in Jammu and Kashmir) the commisson process represents a pruning away of the more prosperous and powerful groups from a comprehensive list that approximated a regime of communal quotas. In the northern tier, however, provision for this stratum is added slowly, first in education, then in government

jobs, converging on a somewhat similar position. Through the course of this development the term

“‘backward clas-

ses” has retained a multiplicity of meanings. It is used to describe the

totality of groups entitled to preferential treatment on the basis of their

137. The methods of the commissions are discussed in chap. 8, §F, below.

186 Identifying the Beneficiaries “backwardness” 3—i.e., the Scheduled Castes and Tribes as well as

“Other Backward Classes” — but not those accorded special treatment

because of temporary or situational disadvantage (e.g., disaster victims,

refugees, defense personnel). The term also refers specifically to those

backward groups other than, the Scheduled Castes and Tribes. There

remain fundamental, if rarely articulated, disagreements about who

these groups are. Some would confine this category to the lowly— those

“far below” the mean in welfare and resources, or those whose depriva-

tions are comparable to those of the Scheduled Castes and Tribes;159

others

use

the

term

Backwerd

Classes

to describe

a wide

middle

stratum of Indian society, who require and deserve special help because

they are lagging behind the most advanced groups. **° There is, as we have seen, further disagreement over whether the term refers to the less well-off in all communities or whether it encompasses only those communities that suffer “backwardness” as a group. As we saw in §A, the question of who were the Scheduled Castes was

debated and roughly settled before Independence within the executive and without the participation of the courts. But who are the Backward

Classes is a post-Independence question which the constitutional recognition of the category made one of all-India scope. The Constitution

left the matter with the executive at the state level with an option for the

Centre to unify it. When the executive at the centre first failed and then declined to provide a resolution, the question reverted to the states. In the wake of the Janata victory in the 1977 elections, the Backward

Classes returned

to the national political agenda. Pursuant to its

electoral promise,*! the Janata Government appointed a new Back-

138. In this broad meaning it is more or less equivalent to “weaker sections” in current political usage. (Cf. the reference to “weaker sections of the people” in Art. 46.)

139. See chap. 8, §E.

140. This view is neatly epitomized in the remarks of Shri K. Hanumanthaiya, a

Congress M.P. from Karnataka, during the debate over the 1976 revision of the lists of Scheduled Castes and Tribes. He observed that although he welcomes reservations for

the latter, “[i]t is the middle or the backward classes that have been completely

ignored and they are the people who are suffering the most in this set-up . . .it is the

middle classes which suffer most during the time of inflation. These most neglected middle classes or the backward Kalelkar Commission[’s}

classes ...{should]

be helped in the spirit of the

. ..recommendations.” LSD (Fifth Series), Vol. 64, No. 16,

cols, 76-77 (2 Sep. 1976). 141. The 1977 election manifesto of the Janata party called for an end to caste

distinctions and promised the establishment of an independent and autonomous civil

rights commission “competent to ensure that the minorities, scheduled castes and tribes, and other backward classes do not suffer from discrimination or inequality.” It

promised a radical reduction of disparities by a “policy of special treatment” in favor of the “weaker sections of our society.” In connection with the provision of “‘preferential opportunities for education and self-employment to these sections” the party promised

Other Backward Classes

187

ward Classes Commission in February 1978.14? Northern state governments under Janata control substantially enlarged the preferences for Other Backward Classes, leading to massive violence in Bihar. In its declining days, the Janata caretaker government came close to instituting job reservations for Backward Classes in central government

service. "43

.

During the hiatus of central involvement, what the states did was increasingly subjected to the examination of the courts. It has been the Supreme Court rather than the central government which has been the unifying and limiting influence, and presumably any new central policy will be shaped in light of two decades of judicial predominance in this area. It is to the judicial encounter with the Backward Classes category that we now turn. to “reserve between 25 and 33 per cent of all appointments to Government service for the backward classes, as recommended by the Kelkar [sic] Commission.... The Party will formulate a special program within the framework of the five-year plaris for the substantial advancement of the scheduled castes and tribes and other backward classes and will provide adequate funds for the purpose . . . Special machinery will be set up to implement the programme and assure fulfillment of the relevant Constitutional

guarantees” (from Commerce, Feb. 19, 1977).

142. The new five-member commission was established in December 1978 under the

chairmanship of B. P. Mandal, M.P. and instructed to report by the end of 1979. Its terms of reference, reminiscent of those of the 1953-55 Backward Classes Commission (I

BCC 2) with the addition of a specific inquiry about reservation of government posts, were (i) “to determine the criteria for defining the socially and educationally backward

classes; (ii) to recommend steps to be taken for the advancement of the socially and educationally backward classes of citizen so identified; (iii) to examine the desirability

or otherwise of making provision for the reservation of appointments or posts in favour of such backward classes of citizens which are not inadequately represented in the

services of both the Central and State Government/Union Territory Administrations;

and (iv) present a report setting out the facts as found by them and making such recommendations as they think proper.” When asked under what authority this

Commission was constituted, the Minister of State for Home Affairs replied that it was “constituted under Article 340 of the Constitution” (RSD Vol. 107, No. 23, cols. 171-73

[21 Dec. 1978}).

143. It is reported that a few weeks before the 1980 elections, the coalition govern-

ment headed by Charan Singh proposed to reserve 25% of central government jobs for Backward Classes. This proposal was dropped after the President objected that it violated an agreement that the caretaker government would refrain from taking policy decisions which might amount to electoral initiatives (Overseas Hindustan Times, Dec. 27,

1979).

7

The Backward Classes and the Judiciary: The Selection of Beneficiary Classes

From this. point forward the analysis focusses on the work of the courts. Readers

who are unfamiliar with courts and litigation in India may find it helpful, before

proceeding, to look ahead to chapter 14. A quick acquaintance with the structures and operations of the courts, the system of authority and precedent, litigation

patterns, and the legal profession may be obtained by reading the opening pages of

each of the sections of chapter 14.

A. CASTES AS CLASSES WE TURN Now to the emergence of the doctrinal frame which shaped,

and was in turn shaped by, the policy developments recounted in chapter 6. One of the crucial tasks of this doctrine is to assign a meaning to the term “‘classes” as it appears in the constitutional authorization of preferential treatment. To make sense of the course of development it may be helpful to identify three basic and recurrent views of the constitutional meaning of “‘classes,” views that reflect more comprehensive inclinations about equality and society in India. First, there is what we might label the “modernist” view, which would confer preferential treatment on lines of income, occupation, illiteracy, or other “neutral”

or “secular” characteristics and would eschew the communal groupings whose use is otherwise forbidden in the Constitution. Thus “classes” would be collections of individuals selected for special treatment because they suffered one or another form of disadvantage. The opposite position is taken by proponents of what may be labelled the “historic” view. In this view the constitutional provisions for backward classes aim at overcoming the cumulated disparities of power, wealth, and culture among historic communities. “Classes” then refers only to

Backward Classes and the Judiciary—Selection

189

those historic social formations (caste, and religious groups, perhaps occupations and localities) that were the object of constitutional policy. Finally, there are adherents of an “elastic” position, who regard “classes” as broad enough to include both historic social formations and ad hoc categories. In this view government is neither commanded to use one of these nor excluded from using the other.

These views are introduced here as a mnemonic convenience. In the following account, which is arranged along other lines, we shall en-

counter some

of the arguments

of constitutional history, textual in-

terpretation, and policy consequences that are advanced on behalf of each. Later we shall consider the more comprehensive positions as to the pursuit of equality in Indian society that are implicated in these competing readings of the constitutional provisions. . The Constitution names as the permissible recipients of preferences,

not backward individuals or families, nor backward castes, religious

communities, occupational or regional groups, but backward ‘‘classes” of citizens. These “classes” are restricted neither to economic classes nor to classes in the senses familiar to modern social science.’ The term seems to be used in the broad connotation of any group of persons having certain common characteristics. In particular it would seem to include, though it is not confined to, those classifications otherwise forbidden in Articles 15, 16, and 29(2)—e.g., racial, religious, and caste

groups. This is so because Articles 15(4) and 16(4) are exceptions or provisos to these Articles, limiting the operation of their provisions. If the proviso that “nothing in this article” shall prevent the State from making preferences for backward classes is not broad enough to authorize the use of the forbidden classifications, then it was unnecessary to have any proviso at all, since other classifications would be permissible without it. Furthermore, the history of Articles 15(4)'and 16(4) indicates that they were included for this purpose. The entire debate on Article 16(4) in the Constituent Assembly revolved around the question

of which “communities” were intended to be included.? The addition of Article 15(4) by amendment was for the specific purpose of permitting preferences for caste groups.? In a discussion of the use of caste in designating Backward Classes, there are two related and easily confused but distinguishable questions. First, may castes or communities be used as the units or classes that are designated as backward? Second, may the rank, standing, or prestige of a caste group be used as a measure or criterion ofits backwardness? The 1. Cf. Article 23(2), where class is used in what appears to be an economic sense. 2. VIL CAD 673-702, discussed in chap. 6 above.

3. Discussed in chap. 6, above, and chap. 11, below.

190

Identifying the Beneficiaries

latter points to “caste” as the measuring rod; the former points to “castes” as the things to be measured.

This distinction, emphasized here for purposes of clarity, introduces an unavoidable element of artificiality into our analysis of the cases to the extent that courts, along with the other participants, continually discuss the ‘‘caste test” or “caste criterion” as a unity. As the passages quoted in the ensuing discussion indicate, usually it is eifher caste as a unit or caste as a status that is meant. But there are instances in which both meanings are fused. That the courts have succeeded in clarifying the distinction only partially and after a long conceptual struggle points

to the significance of this fusion;* even a body so detached anc devoted to analytical distinctions as the judiciary has found it difficult to envi-

sion even momentarily a caste group abstracted from its status in a

ranked order of groups.5 The courts, along with government agencies and the Backward Classes Commission, have accepted caste as a permissible unit of classification. In Venkataramana v. State of Madras, the Supreme Court upheld reservations for “backward Hindus,” a list of castes designated as backward by the Madras Government.® The series of cases arising out of Mysore’s expansive list of Backward Classes confirms this. Although the reservations were invalidated as arbitrary in Ramakrishna Singh v. State of Mysore, the court indicated that caste was a permissible basis of classification.” It was argued that “‘class” in Article 15(4) excluded “caste” by implication, since that Article specified “Scheduled Castes” as opposed to “backward classes.” Thus,, it was argued, the prohibition on caste differentia in Article 15 (1) should not be read as qualified by Article 15(4) in toto but only by that part of Article 15(4) which refers to

Scheduled Castes. The Court held that “class” included persons

4. Although the courts have not clarified the distinction, it has not escaped the most

perceptive students of the Supreme Court’s decisions in these areas. Thus L. G.

Havanur (1965 : 60) explains: “The word ‘Caste’ can be used in two senses—one in the

concrete and the other in the abstract. In the concrete sense ‘a caste’ means a group of persons or families. In the abstract sense ‘the caste’ means status or rank or position” (italics in original). Cf. Radhakrishnan 1966.

5. The difficulty many Indians experience in separating reference to a caste group from a notion of standing or rank became more comprehensible to me when I imagined some American analogs. We find it easier to isolate terms describing occupational,

religious, and ethnic groups from notions of relative rank, but it is difficult for an

American to abstract the term “working class” or “upper middle class” from its comparative and relative connotations. Similarly, a knowledgeable baseball fan might find

it difficult to envision a set of teams totally abstracted from their relative standing in the league. 6. A.LR. 1951 S.C. 229. 7. A.LR. 1960 Mys. 338.

8. It should be noted that the text of Art. 16(4) supplies no ground for such a

distinction.

Backward Classes and the Judiciary—Selection

191

grouped together on the basis of their castes, though it would certainly be open to the government to determine the classes on any other basis. In Partha v. State of Mysore the court, upholding this aspect of the state scheme, indicated that the caste “classification will only be open to challenge if it can be shown that the criterion adopted for determining their backwardness is useless as a test of backwardness, so that the

preference would amount to a preference on grounds of caste alone.”?

That is, it was permissible to use caste units, but there had to be some

criterion of backwardness in addition to caste rank or standing. The law set forth in these high court cases has remained essentially unchanged. But a series of notable Supreme Court cases refined it and widely conveyed a notion of substantial change. In Balaji v. State of Mysore the Supreme Court for the first time explicitly dealt with the question of the permissibility of the use of caste as a unit.1° There are a number of places in its opinion where the Court is critical of the use of caste as a criterion of backwardness. But it is quite clear on close inspection that the Court puts forward no constitutional objection to the use of caste or community as a unit of designation, although it

deplores this use on policy grounds. In its examination of the state’s

classification of educationally backward communities, the Court clearly accepts the use of “communities” — including caste groups—as the units whose backwardness is to be determined. It concludes that

“only those communities which are well below the State average [of

literacy] can properly be regarded as educationally backward classes of citizens.”11 The Court takes strenuous objection to the exclusive

reliance of Mysore on caste (i.e., in the sense of rank or standing ) as a criterion of social backwardness and notes that occupations and place of habitation are also relevant to determination of the backwardness of “‘a

community of persons.”

It is clear that communities are permissible units: “It is for the attainment of social and economic justice that Article 15(4) authorizes the making of special provisions for the advancement of the communities there contemplated, even if such provisions may be inconsistent with the fundamental rights guaranteed under Article 15 or 29(2).”*3 The communities that may be used appear to be broader than castes in the narrow sense, but include all communities—for one of the Court’s

objections to emphasis on caste is its inapplicability to non-Hindu

groups."

9. A. LR. 1961 Mys. 220 at 230. 10. A.LR. 1963 S.C. 649. A decade later Balaji was still referred to respectfully as the.

locus classicus of learning on Backward Classes (Janki Prasad Parimoo v. State of Jammu and

Kaskmir, A..R. 1973 S.C. 930 at 937). 13.Jd., 664 (italics added).

11. Id., 660.

14. See chap. 8, §B, below.

12. Id., 659.

192

Identifying the Beneficiaries In short, Balaji seems to leave intact caste or communal units as the

things to be measured, while confining the role of caste as the measuring rod.15 However, the failure to make clear the distinction between

castes as units and caste rank as a measure of backwardness put both

uses of caste in some disrepute and encouraged the notion that ‘“‘caste”’ was in all respects eliminated from the selection of Backward Classes. Asa

result of the Balaji case, Mysore once more revised its scheme for

reservations in technical and medical colleges. It deemed as Backward

Classes all persons following the occupations of cultivator, artisan,

petty businessman, inferior service, or other occupations involving

manual labor and with an annual income of less than Rs. 1,200. It

reserved 30% of the seats for these. Upholding this new order, the Mysore High Court expressed grave doubts about this classification, which it found to be far from satisfactory. It read Balaji to indicate that there could be no satisfactory classification of Backward Class Hindus while ignoring the caste basis. “[I]n addition to the ‘occupation’ and. ‘poverty’ tests, the State should have adopted the ‘caste’ test.’”16 Pointing out the difficulties of the income/occupation test and its detrimental effect on the most socially backward castes, the High Court concluded with the hope that the state would make ‘‘a more appropriate classification lest its bona fides should be questioned.”’!7 This case went up on appeal on another question, and appellant’s counsel urged the Supreme Court to repudiate these observations of the High Court as conflicting with the Balaji case. Surprisingly, the Court complied ina series of dicta that undermined the limited clarity attained by Balaji. In Chitralekha v. State of Mysore, the Supreme Court finds that Balayt

established that (1) “the caste of a group of citizens may be a relevant

circumstance in ascertaining their social backwardness” and (2) “‘it

cannot be the sole or dominant test in that behalf.”'® It then (3) repudiates the High Court’s view that caste is a necessary test of social backwardness:

We would hasten to make clear that caste is only a relevant circumstance in

ascertaining the backwardness of a class and there is nothing in [Balajt] . . .

which precludes the authority concerned from determining the social backwardness of a group of citizens if it can do so without reference to caste. While this court has not excluded caste from ascertaining the backwardness

15. “What was questioned in [Balaji] ...was not ...the taking of a caste ...or a section ofa community as a unit ....But the principle .. .of taking the status ofa community or a caste in the hierarchy of castes as the determining factor of its backwardness ....”” (Havanur 1965: 100-1; italics in original)

16. Viswanath v. Government of Mysore, A.1.R. 1964 Mys. 132 at 139.

17. Id., 140

18. A.LR. 1964 S.C. 1823 at 1833,

Backward Classes and the Judiciary—Selection ofa

193

class of citizens, it has not made it one of the compelling circumstances.

. .. [T]he authority concerned may take caste into consideration . . .but,

if it does not, its order will not be bad on that account. . . .19

So far the Court has only added to Balaji the notion that caste is a permissible but not a mandatory measure of social backwardness. It has not said anything about what are the “classes of citizens” whose backwardness is to be measured — which, as we have seen, Balaji tacitly

permitted to be caste or communal units. Subba Rao, J., then presents some reflections about the use of caste units which compound the ambiguities:

The important factor to be noticed in Article 15(4) is that it does not speak of castes, but only speaks of classes. If the makers of the Constitution intended to take castes also as units of social and educational backwardness, they

would have said so. . . . Though it may be suggested that the wider expres-

sion ‘‘classes” is used . . . as there are communities without castes, if the intention was to equate classes with castes, nothing prevented the makers of

the Constitution to use the expression “Backward Classes or castes.”” The

juxtaposition of the expression “Backward Classes” and “Scheduled Castes” in Article 15(4) also leads to a reasonable inference that the expression “classes” is not synonymous with castes. It may be that for ascertaining whether’a particular citizen or group of citizens belong to a backward class or not, his or their caste may have some relevance, but it cannot be either the sole or the dominant criterion for ascertaining the class

to which he or they belong.?°

He suggests that this restricted use of “caste” is necessary to prevent exploitation of these provisions by well-off sections within groups that are largely backward. If we interpret the expresson ‘“‘classes” as “castes,” the object of the Constitution will be frustrated and the people who do not deserve any adventitious aid may get it to the exclusion of those who really deserve it. This

anomaly will not arise if, without equating caste with class, caste is taken as

only one of the consideration to ascertain whether a person belongs to a backward class or not. On the other hand, if the entire sub-caste, by and large, is backward, it may be included in the Scheduled Castes by following

the appropriate procedure laid down by the Constitution.?!

19. Id., at 1833. 20. Id., at 1833-34. 21. Id., at 1834. The argument that if the whole “‘sub-caste” is backward it should be added to the list of Scheduled Castes is elaborated by Tripathi (1972 : 204), who

portrays the legitimate use of caste as exhausted by the provisions for identification of

the Scheduled Castes: “‘castes which ought, as such, to be regarded as backward have

194

Identifying the Beneficiaries

With a bow to the complexity of the problem, the Court concluded: [What we intend to emphasize is that under no circumstance a “class” can be equated to a “caste” though the caste of an individual or a group of individuals may be considered along with other relevant factors in putting

him in a particular class.?2

The practical policy that the Court tells us it is trying to effect— elimination of the well-off within the backward group— could of course be achieved as readily by requiring income ceilings. But the Court seems to go much further than addressing this problem; it comes close to saying that caste units cannot be used at all—although it does not explain what sort of “group of individuals” it has in mind. Caste in the sense of rank or standing may be part of the measuring rod, caste in the sense of unit cannot be the thing that is measured.> The argument is flagrantly unhistorical. Whatever else classes have

included in Indian jurisprudence, they have clearly included caste and communal units.” Furthermore, the textual argument is unconvincing,

since it would either undermine the well-established meaning of “back-

ward classes” in Article 16(4), which clearly includes the Scheduled

Castes, or lead to the conclusion that the “backward classes” in Article 16(4) are different than those referred to in Article 15(4).

The Chitralekha judgment reinforced the popular reading of Balaji as completely outlawing the caste basis. The national press editorialized with satisfaction about the demise of the caste test of backwardness. already been included in the category of scheduled castes and have been expressly

mentionedin .. .[Article 15 (4)]; therefore the rule of expressio unius est exclusio ulterious must apply, and the expression ‘classes’ in the sub-clause must be construed as contemplating groups other than caste-group.” This piece of interpretation not only

involves departures from the original (and subsequent) understanding of Art. 15 (4), but it

leads to an unseemly textual tangle, for “backward classes” in Art. mean something very different from the “backward classes” referred 22. Id., at 1834. The warning that “caste” is illegitimately used as a cannot be “equated” with class seems to imply that no term narrower

16(4) would then to in Art.15(4). “class” because it than class may be

used, for by being narrower it would not be equatable. But the question is really not what is

equatable with class, but what is included within it! 23. Mudholkar, J., dissenting at some length on the interview question, joins the majority in condemning the observations of the High Court Judge—but for exactly the opposite reasons. He agrees that the presence of Art. 15(4) as an exception to Arts. 15(1) and 29(2) “does not justify the inference that castes have any relevance in determining what are socially and educationally backward communities” (A.I.R. 1965 S.C. at 1843). Apparently he accepts communities as the units, but implies that Arts. 15 and 29 bar any use of caste rank as a measure of backwardness. 24. Havanur (1965) not only points to textual evidence (the heading of Part XVI, the

wording of Art. 16(4)), but provides numerous instances of previous usage from the

Constituent Assembly, the High Courts and Privy Council, official documents, and

political leaders. See also Radhakrishnan 1965.

Backward Classes and the Judiciary—Selection

195

The impact of these decisions in the courts was more complex. The success of the petitioners in Balaji had stimulated litigation in several states where there were extensive reservations for Backward Classes— especially Andhra Pradesh and Kerala. The first court to which Chitralekha’s dicta on caste were cited was the Division Bench of the Kerala High Court in State of Kerala v. Jacob Mathew.?5 A single judge had struck down the state government’s order specifying Ezhuvas and Muslims as Backward Classes on the ground that “the predominant if not the sole test [of social and educational

backwardness] that has weighed in the mind of the State Governmentis

the test of caste or community,” and that this is impermissible. Here,

“no investigation has been made, or data or materials collected, for the

purpose of enabling the State Government to consider . . . which particular groups or sections...have to be treated as backward

classes. . . .”27 On appeal to a Division Bench, respondents urged that Chitralekha (which had been decided in the interim) made it impermiss-

able to use caste units in specifying Backward Classes. The Division Bench declined to read it that way, and taking Chitralekha with Balaji, found nothing which

precludes the conclusion that if the whole or a substantial portion ofa caste is socially or educationally backward, then the name of that caste will be a symbol or a synonym for a class of citizens who are socially and economically backward and thus within the ambit of [Article 15(4)].?%

The Division Bench proceeded to restore the reservations for Ezhuvas, Muslims, and Latin Catholics. This case was admitted on appeal to the Supreme Court, but was withdrawn in the course of oral argument on the understanding that the state was reviewing its policy on educational

reservations. A Commission established for this purpose reported in late 1965. It recommended, and the Government adopted, a scheme of

reservations for the same backward communities, but now qualified by

an income ceiling.2? The “means-cum-community” test was subse-

quently upheld against the challenge of an applicant who satisfied the means test but not the community test.3° Communal units without an income ceiling continued to be used in distributing reserved posts in 25. I.L.R. 1964(2) Ker. 53. 26. A.LR. 1964 Ker. 39 at 57. 27. Id., at 58. 28. 1.L.R. 1964(2) Ker. 53 at 60. Mayakrishnan and Radhakrishnan (1965 : 28) observe that the Division Bench “failed to appreciate” that Chitralekha was irreconcil-

able with Balaji on this point.

29. Kerala (Kumara Pillai Commission) 1966—known as the Kumara Pillai Com-

mission after its chairman, a retired High Court Judge. 30. Laila Chacko v. State of Kerala, A.1.R. 1967 Ker. 124.

196

Identifying the Beneficiaries

government service. This was challenged and reluctantly upheld in Hariharan Pillai v. State of Kerala.34 The Andhra High Court responded to the Balaji and Chitralekha cases ina very different fashion. The long-standing castewise list of Backward

Classes had gone unchallenged until Balaji, Indeed, the government in

1963 added two castes of doubtful backwardness to the list. This was challenged, and a single judge struck down the reservations in educational institutions on the ground that the list was prepared entirely on the basis of caste.3? The state government then dramatically shifted its policy on preferences: it withdrew-the list of Backward Classes, adopted

an economic criterion for educational and other benefits, and elimi-

nated reservations in professional colleges and government employ-

ment. This withdrawal was then challenged by members of the erst-

while Backward Classes, who sought the restoration of the list and of the reservations. In Dasa Rayadu v. Andhra Pradesh Public Service Commission, the High Court denied this on the ground that reservations were permissible but not mandatory.*3 Adopting the Balaji-Chitralekha position that caste is a relevant but not a necessary factor in determining backwardness, the Court then quotes Justice Subba Rao’s broader dicta on the distinction between caste and class and infers that [D]etermination of backward classes on the basis of castes would therefore be not only derogatory of Articles 15 and 16 but would also go quite contrary

to the avowed principles enunciated in the Constitution. . . .>4

It is not clear whether the Court really thinks it is impermissible to use caste units if there is some way other than caste rank of measuring their respective backwardness. In the meantime, political pressure from the erstwhile Backward Classes led to the formation ofa Cabinet Sub-Committee to re-examine the matter. On the basis of this Committee’s finding, the reservations

were restored and the Backward Classes defined much as before ina list

of 119 castes. This list, too, was struck down by the High Court, but this

time the Court’s emphasis was on the absence of adequate findings of backwardness, not on the use of communal units, which were, imp-

licitly, permissible to use.35 Thus,

before the issue returned

to the Supreme Court, we see the

Kerala and Andhra Pradesh judiciaries displaying contrasting styles in handling the litigation stimulated by Balaji.36 The Kerala Court took a 31. A.LR. 1968 Ker. 42. 32. Sukhdev v. Government of Andhra Pradesh, 1966(1) Andh. W.R. 294 (1963). 33. A.LR. 1967 A.P. 353. 34. Id., at 361.

35. Sagar v. State of Andhra Pradesh, A.1.R. 1968 A.P. 165.

36. The only other state in which litigation was stimulated by the Supreme Court’s decision in Balaji was Bihar, where it was more sparse and intermittent. In Nanda Kishor

Backward Classes and the Judiciary—Selection

197

soft line, cautious and deferential to the state government, applying ure for rationalization and elimination of anomalies while assuming responsibility for the proper working of the system of preferences.

The Andhra Court, on the other hand, takes a more detached and

indifferent stance. The difference in judicial response corresponds to the contrasting fashions in which the state governments handled the matter. The Kerala Government was consistent and determined in its policy; the Andhra policy was subject to sharp vacillations, which exposed its character as a product of shifting political forces rather than an expression of basic state policy. In Kerala, the government’s position

was founded on the careful and workmanlike report of the Kumara Pillai Commission, marshalling relevant data and arguments;?7 in

Andhra, the government’s policy was unsupported by data, and the court was left with no choice between swallowing it completely or overturning it completely. In spite of the differences in style and result, neither high court relied upon the animadversions of Chitralekha against caste units. Building cautiously on Balaji, the Kerala Court completely ignored this aspect of Chitralekha. In Andhra the question of caste units was relegated to the background, and other objections to the government’s scheme brought it down. Ina series of cases decided in early 1968, the Supreme Court moved somewhere closer to articulating the distinction between caste units and caste rank. In P. Rajendran v, State of Madras the Court upheld the use of castes as the units by which Madras’s Backward Classes were

designated:

{I]t must not be forgotten that a caste is alsoa class of citizens and if the caste as a whole is socially and educationally backward, reservation can be made

in favor of such a caste on the ground that it is a socially and educationally

backward class within the meaning of Article 15(4).3%

Sharma v. State of Bihar, A.1.R. 1965 Pat. 372, a 10% reservation for Backward Classes in medical college admissions was held invalid on the ground that the determination of backwardness was, contrary to Balaji, solely on the basis of caste. Reservations for Backward Classes remained in force, however. See Umesh Chandra Sinha v. Singh, ALR.

1968 Pat. 8. In late 1967, a Full Bench had no difficulty in finding the principle of Balaji

inapplicable in a case where a notoriously “backward community is described conve-

niently by its caste name.” Chait Ram v. Sikander, A.I.R. 1968 Pat. 337 at 339; Hridaya

Narain Singh v. Mohd. Sharif, A.1.R. 1968 Pat. 296. 37, The

Kerala Court had the additional advantage that the number of groups

involved was small and encompassable. The judges themselves had a sense of who the

groups were and what their situation was. See chap. 8, §F, below. 38. A.I.R. 1968 S.C. 1012 at 1014—15, Ghouse (1969) finds P. Rajendran “irreconcil-

able” with Balaji and Chitralekha and calls for its repudiation. It appears, however, that he confounds the remarks of Chitralekha against caste units with the Balayi holding that

caste standing is unacceptable as an exclusive test of backwardness. From the point of

198

Identifying the Beneficiaries

It is clear that what the Court is upholding here is the caste as the unit, not caste standing as the measure of backwardness. Listing by caste, the

Court continues, “does not necessarily mean that caste was the sole

consideration ” in finding these groups backward.>? In State of Andhra Pradesh v. Sagar, the Court confirmed its awareness

of the distinction, but succeeded in blurring it once more. It found that

in the context of Article 15(4), “class”

means a homogeneous section of the people grouped together

because of certain likenesses or common traits and who are identifiable by some common attributes such as status, rank, occupation, residence in a

locality, race, religion and the like. In determining whether a particular

section

forms a

class,

caste

cannot

be excluded

altogether.

But

in the

determination of a class a test solely based upon the caste or community cannot also be accepted.‘°

Shah, J.,’s qualification that “determination of a class” cannot be

“solely based upon the caste” is obscure and puzzling. Apparently, the question of what is a class— to which caste as a unit is relevant—is not yet severed from the question of the standards to be employed in measuring backwardness—to which caste as a status or rank is

relevant.

The position is clarified by the decision, if not by the language, in Triloki Nath Tiku v. State of Jammu and Kashmir, which came before the Supreme Court twice. The state provided reservations for “Muslims of Kashmir” and “Hindus of Jammu.” In 1966 the Supreme Court remanded the case to have the High Court determine whether the communities named were backward and ordered it to submit a report on their population, backwardness, and representation in the services.‘ In 1968, when the Supreme Court took up the matter again,*? it was careful to specify that Backward Classes were not synonymous with castes or communities. Members of an entire caste or community

may in the social, economic, and educational scale of values at a given time

be backward and may on that account be treated a backward class, but that is not because they are members of a caste or community, but because they

form a class.*3

view that regards such a distinction as unwelcome, P. Rajendran marked a “setback” to the “new trend” towards vindicating Art. 15(1) rights (Tripathi 1972 : 202).

39. Id., at 1015. 40. A.LLR. 1968 S.C. 1379 at 1382. 41. A.LR. 19675S.C. 1283. The judgment was written by Subba Rao, J., who did not see

fit to advert to his statements in Chitralekha about the impermissibility of communal units.

42. ALR.

19695.C. 1.

43. Id., at 3.

Backward Classes and the Judiciary—Selection

199

After defining class as a “homogeneous section” as in the Sagar case, quoted above, the Court cautions that

for the purpose of Article 16(4) in determining whether a section forms a class, a test based solely on caste, community, race, religion, sex, descent,

place of birth or residence cannot be adopted, because it would directly offend the Constitution. . . .*

Justice Shah’s puzzling proviso appears here in a more elaborate

from, listing all of the classifications forbidden in Article 16 (2). The

Court has just acknowledged that a community or a caste may constitute a Backward Class. So its observation that caste (and the other

forbidden criteria) cannot be the exclusive test of what is a class appears, then, to amount to no more than saying that the group taken as

a class must share traits of backwardness as well as traits giving it an identity as a class.*5 The proviso adds little to the discussion of what is a class, though it may have implications as to the permissble tests of backwardness. As we have seen, “caste’’ has different connotations in

the context of the questions “‘what is a class” and ‘what is backwardness.” In relation to the former it means a unit or community; in relation to the latter it means a status or rank. The Court may be

implying that caste in the sense of status cannot be the exclusive test used

unit).

to determine the backwardness of the class (i.e., the caste as a

These formulae from Sagar have become a judicial commonplace,

recited at length in all discussions of backward classes. But there has been an interesting development in the way in which the language is

interpreted and applied. In Periakaruppan v. State of Tamil Nadu, Justice Hegde (as he then was) added a straightforward reading: “‘a caste has

always been recognized asa class,” and cites P. Rajendran as standing for

the proposition that caste units are permissible so long as they are

shown to be socially and educationally backward.*® In State of Andhra 44. Id., at 3.

45. Superficially, the Court appears to mean that, quite apart from the question of

backwardness, caste (or the other forbidden classifications) cannot be employed to

identify a class of potential recipients, unless some additional differentia is also applied. That it does not mean this is suggested by the Court's acknowledgement that it is discussing class “for the purpose of Article 16(4)”—that is, a Backward Class. The Court imports into the question of what is a class the standard of “‘no exclusive reliance

on caste” developed in answering the question of what is backwardness because it does

not distinguish the two questions. Furthermore, the additional differentia interpreta-

tion is textually suspect, since Art. 16(4) is an exception to Art. 16(2), which forbids

classification on grounds of “‘caste . . only.” Ifa caste cannot be a class unless there is an additional differentia, Art. 16(4) would be redundant. 46. A.I.R. 1971 S.C. 2303 at 2310.

200

Identifying the Beneficiaries

Pradesh v. Balaram, a two-judge bench reiterates that caste units are per-

missible: “one thing is clear that if an entire caste is, as a fact, found to

be socially and educationally backward, their inclusion in the list of Backward Classes by their caste name is not violative of Article 15(4).”47 Furthermore, the problem of over-inclusion is not fatal to such a classification: a caste may be a backward class notwithstanding the presence in it of a “few individuals [who are] . . both’socially . and

educationally above the general average.’’48 In Janki Prasad Parimoo v. State of Jammu and Kashmir, a Constitution Bench of the Supreme Court cites the Sagar “homogeneous section” language in the course of applying tests of backwardness to groups selected on the basis of residence, present occupation, grandfather’s

occupation, caste, and other characteristics.*? Like its predecessors, the

Court has no difficulty with caste units: the only problem with these is the absence of adequate material on 4 of the 23 caste groups named. But other groupings give rise to difficulty—i.e., the state’s listing of small cultivators and holders of small pensioners as backward classes is troubling to the Court. The Court finds it erroneous to place “economic consideration above considerations which go to show whether a particular class is socially and educationally backward.”5° Neither small cultivators nor small

pensioners is a class in the sense of “a homogeneous social section . . .

with common traits and identifiable by . . . common attributes.”’5+ Instead, these are “artificial groups’? created by the state for the

purpose of receiving these benefits. The nub of the Court’s difficulty is

revealed in its discomfort with the notion that one cultivator or pensioner might qualify as backward while his brother with a slightly large holding or a slightly higher pension would not. It strikes the Court as anomalous that those with similar early circumstances and formative experiences would end up on different sides of the backward line. Although the Court does not explain its unease, what seems implicit is the notion that what is supposed to be remedied by provision for 47. 48. 49. 51.

A.LR. Id., at A.LR. dd. at

1972 S.C. 1375 at 1396. 1395; cf. Chait Ram v. Sikander, A.1.R. 1968 Pat. 337 at 339. 1973 S.C. 930. 50. Id., at 941. 941. 52. Id., at 942.

53. Id., at 941. The notion that the units which are “socially backward” should not

cut across kinship lines surfaces elsewhere. In Pradip Tandon r. State, A.1.R. 1975 All. 1,7, the Full Bench observes that where one branch ofa family lives in the city and the other remains in the rural area, both branches remain the same “as far as social habits and

norms . . .prevailing in the community to which the family belong. . . . Then, to say that

the branch which resides in urban centre is socially forward while the branch that

resides in rural centre is socially backward would be contradictory.” In Shetty v. Stateof Mysore, A.1.R. 1969 Mys. 48, 51, the Court observes that it is not possible to envisage a situation in which one member ofa family is treated as backward and another is not.

Backward Classes and the Judiciary—Selection

201

backward classes is not incidents of individual life history but the ac-

cumulated disadvantages of palpable social groups. Presumably, these groups may be castes, or territorial groups, or the followers of an occupation, but they must have some shared traits which can be measured in terms of social and educational backwardness. One wonders whether a group chosen by the income test commended by Balaji could pass muster with the Janki Prasad Court. This line of development is broken by some judicial observations in State of U.P. v. Pradip Tandon. In the course of addressing the measure of backwardness, the Court displays the full potential for confusion that is entailed by failure to distinguish the measurement question from the units question and to explicate the dual usage of caste. The Court asserts that “caste cannot be made one of the criteria for determining social and educational backwardness. . . . The socially and educationally backward classes of citizens are groups other than groups based on caste.”55 In a particularly puzzling redaction of the Sagar “thomogeneous section” language, the Court pronounces: The homogeneity of the class of citizens is social and educational backward-

ness. Neither caste nor religion nor place of birth will be the uniform element of common attributes to make them a class of citizens.5¢

This phrasing seems the product of confusion rather than a desire to

abandon existing learning on the matter. In any event, these observations are pure dicta as to caste units, since the scheme before the Court did not involve any communal units. The same standards that apply to caste units also apply to the use of religious communities as units. Preferences for religious groups have

been struck down in a number of cases‘’ and upheld in others®® without

54. State of U.P. 0, Pradip Tandon, A.1.R. 1975 S.C. 563. 55. Id., at 567 56. Id., at 567. 57. In Venkatramana v. State of Madras, A.1.R. 1951 S.C. 229, the Supreme Court

tejected the inclusion of Muslims and Christians as Backward Classes but was silent as to the permi: lity of using religious criteria. In State of Jammu and

Kashmir v. Jagar Nath, A.I.R. 1958 J. & K. 14, affg A.LR. 1958 J. & K. i, a cabinet

order authorizing direct appointment of Muslims to certain communal disparity was held void, since there was no preferred group as backward. Reserved seats for Muslims tians on the Madras Corporation Council were held invalid in

of Madras (High Court at Madras,

Aug.

posts to remove the designation of the and Indian ChrisA.R.V. Achar v. State

25, 1952, aff'd on other grounds, A.LR.

1954 Mad. 563). Ramakrishna Singh v. State of Mysore, A.1.R. 1960 Mys. 338, invalidated a scheme of reservations in professional colleges which included reservations

for

Sikhs,

Muslims,

Jains,

and

Indian

Christians.

However,

the

Court implied that the Government was free to use religious as well as caste criteria in designating Backward Classes (id., 345). 58. In Keseva Iyengar v. State of Mysore, A.1.R. 1956 Mys. 20. religious groups were allowed without discussion. (This case is now seriously discredited on other

202

Identifying the Beneficiaries

reaching the question of whether an entire religious community might

be designated as backward. But from Balaji on, the Supreme Court

implicitly treats the religious group on a parity with the caste group for this purpose. In Balaji the Supreme Court, while rejecting the inclusion

of the Muslim community, indicates that preferences may be made for

properly selected “‘communities.”’® Indeed, since one argument for the rejection of the “caste” criterion is its inapplicability to non-Hindu groups like Christians, Muslims, and Jains, it would appear that these religious communities could be the units or classes of backward citizens in the same fashion as caste groups. In Trikoli Nath Tiku v. State of Jammu

and Kashmir [II] the Supreme Court struck down reservations for “Muslims

of Kashmir” and “Hindus of Jammu”

on, inter alia, the

ground that the selection “could not be deemed to have been made on the backwardness of [these] classes.”®! Clearly, the Court reads “class”

broadly enough to include religious communities. Members of an entire caste or community may be “treated as a backward class” because of their shared characteristics.? Although caste and community are permissible units, their use in determining the Backward Classes stands on a different footing than their use in designating Scheduled Castes. In the case of the Scheduled Castes, the caste unit is explicitly authorized by the Constitution, and its use is confined to the President and Parliament.® In the case of Backward Classes, it enjoys no such explicit constitutional sanction nor

is its use similarly confined to the highest central authorities. Backward Classes for the purpose ofa particular measure may be defined not only by central and state legislatures but by administrative departments and possibly even local authorities**—agencies whose accountability is more parochial and who are more likely to share, or at least be responsive to, local prejudices and pressures. Scheduled Castes, for all the vagaries of the definition process, are composed of a limited number of definable groups, already in existence and forming a minority of the total population.

But

Backward

Classes make

up no such definite

group, for it is a category indefinitely expandable, which might include

a majority as well as a minority. The distribution and social position grounds. See chap. 8, §F, and chap 12, §B, below.) In State of Kerala v. Jacob Mathew,

LL.R. 1964 (2) Ker. 53, Muslims and Latin Catholics were included among Backward Classes without discussion of this point (rev’g on other grounds A.I.R.

1964 Ker. 39, where these groups were disallowed). In Hariharan Pillai v. State of Kerala, A.1.R. 1968 Ker. 42, Muslims, Latin Catholics, and Backward Christians

were included among the Backward Classes, while the Court observed “classification should not be made on the basis of religion or caste” (id., 50)

59. A.LR. 1963 S.C. at 661. 61. A.LR. 1969S.C. 1 at 3. 63. Art. 341.

that

60. id., at 659. 62. Id. at 3.

64. See chap. 6, above.

65. See, ¢.g., the situation in Mysore where Backward

Classes included

the

Backward Classes and the fudiciary—Selection

203

of Scheduled Castes makes it unlikely that they could exercise a preponderant influence on the central government, but it is not improbable

that Backward Classes could exercise such influence at the local or the state level. Different dangers are involved and different safeguards are required. Thus, judicial reluctance to review the exercise by the highest central authorities of their exclusive power to designate Scheduled Castes implies no similar deference to local agencies designating Backward Classes. : In designating the Scheduled Castes not only may the President and Parliament use caste as a unit, but it was generally anticipated that they would use caste (in the sense of rank or standing) as the major criterion

of inclusion. Scheduled Castes are composed of “untouchables,” who traditionally suffered disabilities and restricted opportunities precisely on the grounds of membership in a particular caste. Since caste affiliation is crucially relevant to the determination of untouchability, the caste unit is accepted as an appropriate way to identify individuals who suffer that species of backwardness. But when dealing with Other

Backward Classes, it is not clear that caste (in the sense of standing) in

and of itself represents the same kind of barrier and source of disabilities;

therefore,

it is not clear that the caste unit has the same

relevance in identifying the backward.§7 Similarly, the use of religious criteria in designating Scheduled Castes is not necessarily an appropriate precedent for the use of religious units in selecting Backward Classes. The power of the State to define Backward Classes is not necessarily as unfettered as the power exclusively granted to central authorities by the broad wording of Article 341. The relevance of religion in defining untouchability is apparent entire population except for Scheduled Castes, Scheduled Tribes, and Brahmins and constituted a majority of the total population (RCSCST 1958-59 : 12), Scheduled Castes would seem to have little to gain from expansion of their numbers and the consequent dilution of benefits (the quantum of which is outside of their control). But a situation involving tangible incentive to existing Backward Classes to expand the category of Backward Classes is less difficult to imagine.

66. A striking example of this is provided by the situation in Mysore where

the state government

restored the populous Lingayat community

to the list of

Backward Classes after the Nagan Gowda Committee, established by the state government to identify the Backward Classes, had excluded them from its list (RCSCST 1960-61 : 1, 318-19). 67. Cf. the observation of M. N. Srinivas (1956 : 377) that “the barrier of untouchabil-

ity” cuts off untouchables from the opportunities for group social mobility that do exist within the caste system. 68. The President, it has been held, may use religious criteria in designating member-

ship in Scheduled Castes. The exclusion from Scheduled Castes of persons professing a religion differing from Hinduism has been upheld as a valid exercise of his power to specify parts within castes as Scheduled Castes. Fora full discussion, see chap. 9, § E.

204

Identifying the Beneficiaries

(though its application in marginal cases may be questionable); there is

no self-evident reason to accord it such relevance in defining other

species of backwardness. Although a sharper distinction between the question of units and the question of backwardness would permit greater clarity, it is by now well

established that (1) caste and communal units may be used as classes whose backwardness is to be established; (2) caste or communal rank or status may be one of the tests or measures of backwardness by which these groups are selected; (3) caste or communal rank or status may not be the sole or exclusive measure of backwardness. B. THE PERMISSIBILITY OF NON-COMMUNAL CLASSES

The most persevering adherents of the “modernist” view that communal units are impermissible might regard this erratic but basically consistent course of judicial development as merely a temporary deviation from the true constitutional position. But by the mid-1970s the expectation had receded that governments and courts might imminently adopt an austerely modernist view, eschewing the use of traditional ascriptive units in favor of classes based on income, occupaton,

literacy, or other “neutral” tests applied to individuals rather than communities.”° The courts, in disappointing this appealing modernist

urge to de-recognize communal divisions, have been responsive to the

pull of a competing theme in the constitutional commitment to equality—the commitment to pursue equality in the face of the institutionalized group inequality of Indian society. The Constitution authorizes, as Balaji puts it, preferences for “classes of citizens and not individual citizens as such.”?! Does backward “classes” imply anything more than backward citizens—i.e., citizens

69. In his Telang Endowment Lectures, Professor Tripathi (1972 : 204) asserts that the framers did not intend to authorize determination of caste groups to be backward

classes. In his view, “.. .Article 340 refers not to caste-or-religion based communities

but to communities and groups sharing special conditions of hardship and difficulty due

to their environment,” like the inhabitants of the hill assertion suffers from some difficulty : it flies in the face of ing of the Constituent Assembly and of the same body Parliament (see chap. 6, §B, above); and it runs counter

area of U.P. As history this the unmistakable understandwhich sat as the Provisional to the clear understanding of

knowledgeable and disinterested contemporary observers (see chap. 6, n. 69, above.)

70. Expressions of this “modernist” view from diverse official, academic and journal-

istic source are found in chap. 6, §C, above. This theme, present though not dominant in Balaji, reaches its judicial apogee in Chitraletha and subsides with only occasional later

outcroppings, as in the dicta of Pradip Tandon. 71, A.LR. 1963 S.C. at 659. A recent reiteration is in Jayasree v. State of Kerala, A.1.R. 1976 S.C. 2381, 2386.

Backward Classes and the Judiciary—Selection

205

selected because they possess the traits that define backwardness? It seems to imply that personal possession of the traits that define backwardness is neither a sufficient nor a necessary condition for member-

ship in a class for purposes of Articles 15 (4) or 16 (4). For example, all five-year-olds are educationally backward, yet it is clear that five-year-old

children are not an educationally backward class within the meaning of Article 15(4)— if only because Article 15(4) is totally irrelevant to the

State’s power to make suitable provisions for overcoming this sort of backwardness. Similarly, among the applicants for a high government

post, none are likely to be educationally backward by any standard of

comparison with the general population. But the Constitution clearly envisages preferential treatment for some competitors for high posts, even though they will not be personally backward. And arguably, it envisages special provisions for some of the five-year-olds— those least

likely to overcome their presently shared educational backwardness because of their membership in a group with other traits. Apparently, it is the backwardness of the class that is constitutionally relevant, not the

backwardness of the individual beneficiary. Indeed, it has never been

suggested that individuals be selected on grounds of personal backwardness, a procedure that would select those least likely to make use of the opportunity. Selecting individuals on the basis of the backwardness of the group involves the opposite problem: inevitably the benefits flow to the most able and advantaged—the least backward, personally— within the backward group. If the constitutional provisions are specifically designed to overcome the transmitted inequalities of past social distinctions, especially those distinctions whose recognition is otherwise forbidden, then the classes would be those groups that had suffered arbitrary hindrances to advancement because of past application of such distinctions. Clearly, caste and communal groups might fit as such classes. If castes and

communities

are such classes, the question remains

whether other sorts of grouping are also possible. Are illiteracy, low

income, or menial occupation sufficient likenesses to constitute a class

for purposes of these provisions? May the State for purposes of its equalizing policies designate backward classes without reference to the historic social formations to which these provisions were addressed? This set of issues was nicely crystallized by two former Supreme Court Justices at a 1973 Jurists’ Seminar on Backward Classes.72 As retired

Chief Justice Subba Rao poses it,

72. Speeches at the Jurists Seminar on Backward Classes, held in Bangalore in

September 1973 are reprinted as Appendix 8 in vol. I, Part 2 of Government of Karnataka (1975).

206

Identifying the Beneficiaries

the real question is whether the Constitution provides only for the existing or subsequently formed castes or classes with definite characteristics and which are accepted by society as backward castes or classes or can the State form a new backward classes out of different castes and classes with specific

qualities of backwardness.”>

Justice Hegde (who had recently resigned his seat on the Supreme Court after being passed over for the chief justiceship) puts forward what we have called the “historic” view: [T]here can be no doubt that the expression “class” in Article 15(4) as well ‘as in Article 340 denotes a well-knit and organized section of the society based on status or rank and not any classification made by the State on

diverse considerations. . .. The ‘“‘classes” mentioned in Article 15(4) refer to the existing organized sections of the society, and not any new groupings

of individuals.”*

The “historic” view, as argued forcibly by several acute students of the subject, is that class can mean nothing other than caste or communal units.75 Thus, Mr. Havanur argues that Articles 15(4) and 16(4) do not

permit ‘‘new classifications: they only permit specification of the already existing and recognized social classes.”7° To this Chief Justice Subba Rao opposes another reading which, borrowing his term, we label the “elastic” view. [There is . . . another view, ie., backward class is more comprehensive

than the backward caste or community . . .The expression ‘‘class” is wider

. . -in expression than “‘caste”’ or “‘community;” it takes in, in addition to “caste,” other groups based on language, race, religion, occupation, location, poverty, sex, etc. Caste is also a class; . . . The expression “‘backward

class” is an elastic and changing concept. It takes in not only the classes

existing before the Constitution but also those formed after the Constitu-

tion... .77

73. Id., at 65.

74. Id., at 106.

75. Havanur 1965 : If, Radhakrishnan 1965 : 270. A more specific and less persuasive version of the “historic” view reads the provisions for “backward classes” as a commitment.to a definite stratum of the population made up of certain castes. Thus, Mr. C. L.

Sathi (n.d.), a lawyer, politician, and backward classes activist, decried the income test

as violating Art. 340 of the Indian Constitution since “Art. 340 refers to all those castes

which have been considered low like the barber, potter, blacksmith, carpenter, the household servants or shepherds, etc.” But as our discussion of the background and

drafting of the constitutional provisions indicates (chap. 6, §B), there was never any clear agreement about the identity of this stratum. Was it the 12% of Bombay or the 70% of Mysore? There was never any decision as to the cut-off point. Who were the

Backward Classes was not pre-determined, but left subject to later determination. 76. Havanur 1965: 107, 115. 77. Government of Karnataka (1975) Vol. I, Part II, pp. 65, 67.

Backward Classes and the Judiciary—Selection

207

There is some judicial support for the “elastic” view. In Chetralakha, the Supreme Court upheld a scheme that did not use castes or other historic social formations as the units whose backwardness was

measured.78 And the broad definition of class propounded in Sagar and

since reiterated by the courts would seem ample enough to accommodate departures from communal units. On the other hand, there is also support for a reading that is closer to the less permissive “historic”

view.79

The repetition of the requirement that a class display “certain likenesses or common traits . . . some common attributes” suggests that the recipients of preferences under Articles 15(4) and 16(4) have to share traits that make them a class other than the traits that constitute or define their backwardness. The Supreme Court in Janki Prasad draws close to the historic view when it rejects provisions for small cultivators and small pensioners on the ground that neither of these categories are classes in the sense of “‘a homogeneous social section . . .. with common traits and identifiable by . . . common attributes.””®° Instead, they are “artificial groups” created by the state for purposes of receiving benefits under a particular policy. The Court is looking for palpable social groups with perceptible traits like castes, territorial groups, or the followers of an occupation—and there is a hint that such grouping should not cut across kinship lines.81 The judicial attachment to a requirement that such classes be shown to be socially as well as educa-

tionally backward also argues for groups whose social backwardness can be measured.5®? In part the problem is an artificial one, for even if “classes” in Articles 15(4) and 16(4) were confined exclusively to caste or community groups, there is nothing in Articles 15 (1) or 16 (2) or 29 (2) to prevent the

State from using other classifications (like income, literacy, etc.) instead of,

or in conjunction with, caste.®9 That is, Articles 15 (4) and 16 (4) empower

the State to use these otherwise forbidden categories, but they do not 78. Chitralekha v. State of Mysore, A.1.R. 1964 S.C. 1823. This result is anticipated in the

strong commendation of the income criterion in Balaji v. State of Mysore, A.1.R. 1963 S.C.

649, 664. 79. Cf. Viswanath v. Government of Mysore, A.1.R. 1964 Mys. 132 at 139; Gargv. State of

Punjab, A.1.R. 1966 Punj. 476.

80. A.I.R. 1973 S.C. 930, 941.

81. Discussed at n. 53 above. 82. See chap. 8, §C, below.

83. E.g., Art.

15 does not reach a classification exempting the property of “ag-

riculturists” from attachment and sale, since the classification does not depend upon any of the forbidden criteria (Rura Ram v. Gurbachna, A.1.R. 1954 Punj. 254). Of course, if

one placed sufficient emphasis on “only” in Arts. 15(1) and 16(2), one might conclude

that so long as it was combined with other criteria, even caste would not require Art. 15(4) to legitimate it.

208

Identifying the Beneficiaries

compel the State to use them. Nor, when the State does use these categories, is it confined to using them without any admixture of other categories. The difference between the historic and the elastic views would be crucial in a doctrinal setting in which the provisions of Articles 15(4) and 16(4) were thought to exhaust governmental power to confer certain benefits (like reservations) or to give preferment to the disad-

vantaged. But at present the State is thought to enjoy broad powers to

confer advantages, including reservations, on various categories of the

disadvantaged quite apart from Articles 15(4) and 16(4).

Hence these rival readings have little constitutional import in the present doctrinal setting. But they are more than readings of the Constitution; they are, at least implicitly, prescriptions for policy. They speak both to the allowable means and to the goals to be pursued. The historic view is soundly anchored in the observation that the provisions for backward classes were intended to enable the government to address

the problem of’caste/communal disparities by adopting measures along caste/communal lines. Reduction of caste/communal disparities was a central purpose of the inclusion of Articles 15 (4) and 16 (4): the contention

that these provisions permit only measures which address other inequalities is unsupportable. But these provisions are not a detailed blueprint of how the government must pursue this objective; they are but

an authorization to employ one method that would otherwise be barred to

it. There is nothing to forbid the government from attempting to reduce caste/community differentials entirely by other means or to mix the

specifically authorized means with other methods. Surely, the government may attempt to articulate its attempts to reduce communal disparities with measures directed at the general problems of poverty, backwardness, and inequality.®5 Thus the Constitution would seem to provide wide leeway to the State. Whether to use caste units or not is a question of policy, not of constitutional prohibition or command. But if, in programs purporting to be empowered by these provisions, it aban-

dons communal! units in favor of classes defined in other ways, presum-

ably it could be required to show that the classes employed are relevant to the elimination of the kind of backwardness that these provisions are designed to overcome. So far, no court has required such a showing. 84. See chap.

12. §D, below, and the general doctrine of classification discussed in

chap. 11, §D, below.

85. To some of the participants the problem of communal disparities may have exhausted the purpose of these provisions, but it is clear that many others regarded them as part ofan attack on a wider range of inequalities. See discussion at chap. 6, §B, above.

Backward Classes and the Judiciary—Selection

209

C. THE UTILITY OF COMMUNAL UNITS The use of communal units has been the most intensely criticized aspect

of the policy of compensatory discrimination— both within the government and from intellectuals and scholars, who have overwhelmingly condemned it.® If the use of caste (or communal) groups as the units for determining Backward Classes is a matter of policy rather than of constitutional prohibition or command, we should attempt to examine the arguments about their use. Our examination must remain, at this point,

preliminary and abstract, for the arguments turn on factors we have not yet considered: the meaning of “‘social backwardness”; the feasibility of administering non-communal tests; and the wider consequences of using the caste units. Putting aside the argument of constitutional intent, which we have seen is inconclusive, let us concentrate on the

arguments for the utility of communal units in selecting beneficiaries and distributing benefits.87 The argument that caste units® provide a convenient and accurate way of identifying the backward can rest on either of two assertions. The first of these, which we might call the correlational proposition, is that members of castes tend to share a common level of opportunities and

resources. Because of this broad affinity of condition, membership in

an appropriately selected caste provides a convenient way of identifying persons with a low level of opportunity and resources.®? The second assertion, which we might call the causal propost86. See chap. 3, §D, and chap. 6, §C, above.

87. The arguments elaborated in the text concern the contribution of communal units to the redistribuuve and nation building goals of the compensatory discrimination policy. But communal units also appeal to those who, despairing of creation of a non-

communal sphere of public life in the near future, advocate a regime of conumunal quotas. Thus Prof. A. M. Dharmalingam commends the extensive reservations of pre-

Independence South India. Reservations are “in the nature of tariffs” protecting developing

groups. (The existing distribution of merit is the cumulative result of past “deliberate denial of opportunities to the mass of the people . . . not a thing to be proud of .. . ill gotten merit

is equivalent to stolen property.”) He proposes a regime of “proportional reservation and representation” in which “the rule must be the best from each section and not the best from the whole” (in Government of Karnataka 1975 : I, Part I, 78). And, of course, there is a strain of virulent and bigoted anti-Brahminism that in the name of “‘castelessness” would transform these policies from means of inclusion into a device for the complete exclusion of Brahmins from educational facilities and public services. For example, see Ponnarangam 1958:80 ff, 102. Other outcroppings of this substratum are

noted in chap. 13, §A, below. 88. The argument here is applicable to all communal units, but is stated in terms of caste to avoid additional cumbrousness. .

89. When the policy of preferences was first envisioned, it was taken for granted that it would be administered along communal lines. As Lelah Dushkin has pointed out, the notion of preferences along communal lines derives from the widespread (and unexcep-

210

Identifying the Beneficiaries

tion, is that backwardness is the result of caste status, or more accu-

rately, of a hierarchical caste system in which some groups were assigned low status and deprived of advantages.® That is, that the caste group occupies a given status or standing in the social hierarchy; this status or standing is an important determinant of the general level of opportunities and resources of its members;®" therefore membership in a group is a good predictor of backwardness. These propositions are independent: correlation may be maintained without causality and vice versa. The evidence to prove or disprove one or the other would differ. Typically, though, elements of both are merged in argument. To the extent that either of them is true, it can be argued that membership in a caste group is a good predictor of backwardness. In addition to their utility as a basis of selection, it may be argued that

caste units are a useful way of distributing preferences. First, it might be argued (suéprisingly, but with some factual basis) that communal membership is more readily ascertainable than, e.g., income level; so there is relatively little slippage in distribution— the benefits get to the target group, if not necessarily to the most needful within it.?? Dissimulation about communal identity does take place—both individually and on a collective scale—but on discovery is usually treated as some-

what scandalous; misrepresentation of one’s income, on the other hand, is easier to accomplish, harder to detect, and evokes little stigma.

Second, it may be argued that distribution of benefits to members ofan existing caste group may mobilize ties of kinship, loyalty, and mutual support to multiply the effect of the benefits more than does distribution to isolated individuals. Particularly, such communal groups have a greater potential capacity for political organization to facilitate the unimpeded flow of benefits through sluggish administrative channels. The correlational proposition obviously contains some truth. There tionable) assumption ofa correlation between groups with the lowest social statuses and those which suffered the greatest lack of oportunities and resources. The correlation

between the ritual status ofa caste and the political, economic, and educational levels of

the members of that caste was, she observed (1961 : 1666), the “primary assumption on which government policy is based and the principal determinant of the official definition process.” 90. See, e.g., the assertion of Mr. O. K. Yadav that “preferences need to be given by the caste criterion for the simple reason that social backwardness is not only a serious

handicap, rather it is the root cause of economic backwardness.” (Backward Classes Review 1(4-5) : 10)

91. Thus, e.g., Havanur (1965 : 64-65) argues that the relative advantages and

disabilities which have accrued to members of different caste units are transmitted by

caste loyalties, which lead to favoritism, even by the most well-intentioned. In this view open competition is tantamount to bias in favor of those castes whose members are responsible for selection; preferences for backward castes offset the formidable built-in

preference for members of those castes already in favorable positions. 92. See chap. 9, §G, below.

Backward Classes and the Judiciary—Selection

211

is a sufficient affinity of conditions among members of caste groups so that caste units serve as better than random predictors of levels of

education, literacy, income, or other indicia of backwardness. Just how

strong the correlation will be for any given list of castes cannot be predicted in advance. It depends on the definition of backwardness employed, and on the way in which the list of castes is compiled. Because of local and sub-group variation it may be expected that the broader the groups used, the less the force of the correlational argument. Members of the same caste in different parts of the same state or even within the same district may differ greatly in social standing, resources, and opportunities—as may component groups within a large cluster which goes by a common name.®? The correlation would be even weaker ifsuch broad composites as varna categories were used. The religious unit might generally be somewhat more suspect than caste, since it commonly involves a larger group with more economic and social levels, while caste is often associated with a traditional

occupation and a greater affinity of habits and social position among its

members. The less differentiated the units, both socially and territo-

Tially, the less likely that a list of communities would isolate groups with

distinctive social, economic, and educational characteristics. And the

less perfect the correlation the greater the danger that the more prosperous and advanced within the favored groups would be the recipients of the lion’s share of the benefits, thus subverting the purpose of helping the most disadvantaged. But the correlation is bound to be imperfect to some extent: some within the selected castes will have more resources; some outside them

will have less. How one responds to failures of correlation depends on whether one also accepts the causal proposition. If it is believed that membership in a low caste is a major determinant of backwardness, then the absence of correlation is less troubling, for if the members of a low caste were in the past disabled to roughly the same degree as a result 93. Cf. the situation in Mysore where the Nagan Gowda Committee used such large

units as Lingayats, Vokkaligas, and Muslims without distinguishing among component

groups of varying circumstances. See Havanur 1965 : 93 ff.

94. Although the units chosen as Backward Classes may be castes in the sense of

“communities” —jatis, or clusters of such groups with a common name—it is very

doubtful that the State could employ caste in the sense of “varna” (i.e., the four divisions of society propounded by classical Hindu socio-legal theory). It may be assumed that “‘varna” is included within caste, as it appears in the prohibition of discrimination in Arts. 15, 16, and 29. But it might well be argued that “varna” does not fall within the meaning of “class” for purposes of Arts. 15(4) and 16(4), since the members of these vast categories hardly display the degree of homogeneity which Sagar and Triloki Nath require of a “class.” In any event, the criterion is so unwieldy that no one has attempted to employ it : both the Backward Classes Commission and Mysore’s Nagan Gowda Committee rejected the suggestion that Backward Classes be equated with Sudras. See 1 BCC 44; Government of Mysore 1960 : 2.

212

Identifying the Beneficiaries

of their shared low status, even the relatively well-off individuals within

the caste may be thought of as worse off than they would otherwise have

been. Ergo, preferential treatment to restore the deprivations suffered as a result of low caste is justified at whatever level of advantages a member of that caste now possesses. The differential treatment of those

outsiders equally bereft of tangible resources becomes

a distinction

between those who are poor as a result of a heritage of disabilities and those who have failed to benefit from a favorable heritage.°5 Thus we see that these propositions tend to be associated with different views of the backwardness which requires and deserves remedy by preferential treatment. For those who argue the correlational proposition, backwardness is a low level of possession of present resources (such as education, income, cultural accomplishments, perhaps even confidence, poise, or status itself). Those who assert the causal

proposition conceive of backwardness as the accumulated effects of past disablements due to low status. In practice, the two assertions merge: there are few who would assert the causal proposition without reliance to some degree on correlation. The causal proposition, too, contains at least some truth. Those characteristics which can be thought of as most directly associated with

the low status of a caste—lack of cultural resources, lack of a home environment favorable to educational achievement, unfavourable self-

image, lack of access to opportunities, subjection to disabilities and discrimination— play a significant role in limiting the achievements of the members of that caste. But while the social position ofa caste may be

one of the factors affecting the extent of opportunities of its members, there is no reason to assume that it is the only one or that it is universally powerful. As in the case of the correlational proposition, the question is how much. 95. The proponents of preferences on a caste basis are not as callous as the argument makes them appear. Typically, they ask only that preferences for Backward Classes be administered separately from assistance for the poor among the advanced communities.

Thus,

Mr. G. Latchanna,

in his welcome address to the All-India Convention of

Legislators and Leaders interested in the welfare of the “Other Backward Classes” (September 1960), declared: “It is true that there are poorer sections even among the advanced classes of society, and we are not opposed to the continuance of special

fianancial assistance to them. We want only that the poorer sections of the “Other

Backward Classes” should not be deprived of social justice by mixing up their fate with that of the other poorer sections who are much more fortunate in their social and educational heritage.” The intensity with which many Backward Class spokesmen feel the salience of the differential heritage is poignantly expressed by a Backward Classes activist, Shri Nagappa: ‘Even if [most of the ‘advanced classes’] are backward economically, they are not poor mentally; the learning was their grandfathers’ property” (speech at Conference of All-India Backward Classes Federation in New Delhi, 12 March 1966). Acute awareness of the crucial role of structural advantages combines with desperately low

Backward Classes and the Judiciary—Selection

213

The arguments for caste units cannot be dismissed out of hand. The

correlational and causal propositions both lead us to the need for data. How great is the degree of homogeneity or spread of conditions within caste groups? To what degree is caste standing (and the conditions

intimately associated with it) a significant determinant of the kinds of backwardness to be remedied? Does a member of caste X with an income of Rs. 1,000 per annum encounter greater obstacles to mobility than a non-X with an equal or lesser income? The social scientists,

Indian and foreign, who have condemned the use of caste units have not

undertaken to disprove their utility in either the correlational or causal sense. They have instead based their objections on judgments of more far-reaching consequences alleged to attend the use of caste units. Briefly, the use of caste units is said to produce detrimental effects on the recipients (alienation, aggravation of their sense of dependence), on non-recipients, (unfairness, lower morale, etc.), and on the society and

polity

thought

generally.

The

distribution

of preferences

on

caste lines is

to perpetuate caste distinctions, accentuate caste conscious-

ness, aggravate group tensions, and encourage political abuse of the preference device.9© Although a number of these assertions could be self-image. Thus Prof. A. M. Dharmalingam, an unrepentant advocate of a regime of

communal quotas, addressed a Jurists’ Conference on Backward Classes: ‘For instance

a poor Brahmin or a poor Lingayat boy has enormous advantages derived from heredity and environment. A large number of his relatives and friends may be either rich or holding high position in government service or in the public life which combined to his higher intelligence will make him walk off with a seat or a post beating hollow the poor Kuruba or Yadava or Agasa....The poor backward class individual is not only economically poor, but also poor in influence, heritage, environment and resourcefulness and in consequence goes to the wall every time he is pitched against a poor Brahmin or a poor Lingayat.” (Government of Karnataka 1975 : I, Part IT, 86). 96. Thus, e.g., it is often asserted that lists of backward communities will inevitably

undergo a mushroom growth. Thus, Ghurye

1969:430

reports the ominous growth of

the Maharashtra list of Other Backward Classes from 125 castes in 1953 to 160 in early 1967 to 178 in October 1967. In the absence of accompanying population figures, it is

difficult to tell how substantial this increase really is. And in the case cited, one wonders

how much of the “increase” was due to states reorganization. The Punjab Evaluation Committee details the growth of the Punjab Backward Classes lists from 14 communities in 1953 to 25 by 1958 to 63 in 1959. But there had been no further growth when

the committee conducted its survey in 1965 (Punjab, Welfare Department 1966 : 142).

Not surprisingly, non-communal lists of recipients also tend to grow over time. Thus, in

the Punjab the 1956 list of 12 Backward Areas had grown to 25 by 1964 (id., at 146-47)

Possibly, expansion is more of a danger when any kind of “natural” groups are used to designate beneficiaries of preferences. It is then a question of whether their superior capacity to identify the deserving and their multiplier effect outweighs this additional danger. And the lists may shrink as well as grow. Thus in Andhra Pradesh there were 146 Communities on the 1956 list of OBCs; 112 on the 1966 list, and 92 recommended by the state’s Backward Classes Commission in 1970. But it would be misleading to assume any direct correlation of numbers of listed groups with population, since many entries are redundant. See n. 155 in chap. 8, §F.

214

Identifying the Beneficiaries

translated into testable propositions, there is a singular lack of data to support them (which, of course, does not prevent them from being widely believed). Even if it is conceded that all of these evils exist, surely not all of their presence can be attributed to the use of caste units in distributing preferences. For example, there is some evidence for the proposition that the use of caste units enhances caste consciousness, but it is clear that many other factors—especially the suffrage—play an even greater role. If the dependence of the recipients is aggravated by preferences, is it more aggravated by preferences on a caste basis than by preferences on, e.g., an income basis? We lack any reliable estimate of the net effect of the employment of caste units. Only with such an estimate can the alleged costs of this device be measured and weighed against the presumed benefits (and against the costs and benefits of the alternatives). We shall return to these broader questions. On the narrow question of the usefulness of the caste unit, the existing evidence does not permit us to condemn the use of these criteria per se. Each instance of their use

will have to be considered on its own merits. So long as detailed data on the relative backwardness of different sections of the population do not exist, the use of relatively crude classifications such as caste groups might be considered a reasonable method of isolating backward groups. The larger and more undifferentiated the group the more suspect its use. However, such suspicion might be allayed to the extent that other “neutral” standards are combined with the forbidden classification (e.g., an economic ceiling for enjoyment of benefits), particularly if the designated group is large and varied in circumstances. Once it is accepted that communal units are to be used in conjunction with other indicia of backwardness, the question arises whether it would not be simpler to use the latter exclusively. If backwardness is conceived of as the lack of tangible resources, it is presumably possible to find tests which measure more directly the conditions to be remedied. But the answer would seem to depend upon the range and distribution of resources and opportunities within communal groups and on the extent to which caste status is determinative of them. Ifit turns out that there is actually considerable uniformity of circumstance or that shared status is an important determinant of mobility, caste units might ap-

pear a reasonable way of selecting Backward Classes. Paradoxically, the very data which would assure equitable administration of preferences along caste lines might permit dispensing with caste units altogether. But the data needed to have equitable administration on non-communal lines would permit more effective use of caste units. The choice would then turn on the relative benefits and disadvantages of alternative methods— including the difficulties and costs of data col-

Backward Classes and the Judiciary—Selection

215

lection®’ and administration as well as effectiveness in identifying the backward and the wider effects of the various tests. D. FORBIDDEN CLASSIFICATIONS AND THE BURDEN OF PROOF Before proceeding to the other half of the problem—the measure of backwardness—let us return briefly to the technical legal setting. Caste® and religion,” along with the other classifications mentioned in Articles 15(1), 16(2), and 29(2)—race, sex, place of birth, residence, descent, and language—are forbidden criteria for State action. Articles 15(4) and 16(4), then, are exceptions to the prohibition on the use of

these classifications; they may be used to define Backward Classes even

though, when used for other purposes, their use may be struck down as 97. Data-gathering along caste lines may, when conducted under official auspices, produce some of the ill effects attributed to preferences along caste lines. Cf. the experience of the Backward Classes Commission and the frequent observation of the intensification of caste feeling resulting from collection of caste data in the Census (e.g., Ghurye 1957: 192). 98. Thus caste (and religion) cannot be employed for purposed of controlling crime (Sanghar Umar v. State, A.1.R. 1952 Saur. 124), suppressing disorder (State of Rajasthan v.

Pratap Singh, A.1.R. 1960 S.C. 1208,), delimiting constituencies (Bhopal Singh v. State

A.LR. 1958 Raj.41), or composing electorates (Nain Sukh Das v. State of U.P., A.I.R.

1953 S.C. 384). Indeed, quite apart from the- ban on caste criteria, the use of other

criteria for the purpose of promoting the identity and integrity of caste groups is itself invalid. The Supreme Court held unconstitutional laws providing for pre-emption on the basis of vicinage on the ground that the purpose of such laws was to promote and protect communal neighborhoods. “{Such] division of society ...into groups and exclusion of strangers cannot [any longer] be considered reasonable ...{or] in the interests of the general public” (Bhau Ram v. Baij Nath, A.1.R. 1962 S.C. 1476 at 1481) For a review of the post-Independence posture of law toward caste, see Galanter 1968. 99. Religion presents a somewhat more complicated instance, for while government is committed to eschew religious discrimination, it is simultaneously committed to protect

the identity and integrity of separate religious groups. Thus although religious differentia may not be used tor most purposes (see n. 98 above), it is permissible to use religion in determining the applicability ofa law of religious trusts (e.g., Motidas v. S.P. Saki

A.LR.

1959 S.C. 942) or in the application of family law. Though Art. 44 directs the

eventual elimination of separate personal law for members of different religions, the continuing validity of disparate rules of personal law and the power of the State to create new rules of personal law applicable to members of particular religious communities has been upheld (State of Bombay v. Narasu Appa, A.1.R. 1952 Bom. 84; Srinivasa Aiyar v.

Saraswati Ammal, A.1.R. 1952 Mad. 193). But cf. Sheokaran Singh v. Daulatram, A.1.R. 1955

Raj. 201, where a rule of Hindu law regarding recovery of interest was held invalid as discriminatory against non-Hindus. 100. The arrangement of forbidden criteria in the various provisions can conveniently be represented in tabular form: (see next page).

216

Identifying the Beneficianes

discriminatory.?°? Other than caste, religion, and residence,!™ these classifications are rarely used in designating the Backward Classes.1° While there is no constitutional ban on the use of these forbidden

criteria for this purpose, their use is not prescribed nor is it immune from judicial .review. When used, they are subject to review in accordance with the general standards of reasonable classification employed by the Court.

Ground of Discrimination Religion Race Caste

Sex Place of Birth Descent Residence Class

Language

15(1) (General) x x x

x x

Forbidden in Article: 16 (2) 23 (2) (Government) (Compulsory) Employment Public Service x x x

x x x xX

x x x

x

29(2) (Admission to) State or Aided Schools x x x

x

101. Thus, in State of U.P. v. Pradip Tandon, A.1.R. 1975 S.C. 563, 569, the Supreme

Court found reservation of admissions for rural students unconstitutional on the ground that it arnounted to discrimination by “place of birth.” This was in the context of a finding that inhabitants of rural areas did not constitute a socially and educationally

Backward Class, so the question did not arise whether a legitimate Backward Class

might be defined by, inter alia, place of birth. 102. The uses of caste and religion have been discussed in the preceding section.

Residence is a forbidden criterion for purposes of Art. 16 (see Kishori v. Board of Revenue, A.LR. 1957 Raj. 185). But it is allowable in educational matters (Joshi v. Madhya Bharat,

A.LR. 1955 S.C. 334). It seems never to have been doubted that residence may be used

in defining Backward Classes for all purposes (including reservations in government employment). Government lists of Backward Classes frequently employ territorial criteria—e.g., the members of caste X in district Y. Indeed, location may itself be a constitutive part of backwardness that the State may validly take into account. See Chap. 8, §H, below. However, reservations and distributions on territorial lines which are not related to backwardness must meet the general tests of reasonable classification. District-wise distributions of medical college seats have been struck down as having no reasonable relation to the object to be achieved—i.c., admission of best talent (P. Rajendran v. State of Madras, A.1.R. 1968 S.C. 1012; State of Kerala v. Jacob Mathew, 1.L.R.

1964[2] Ker. 53, 62, afPg A.ILR. Ker. 1964 39, 68; Abodha Kumar v. State of Orissa, A.LLR.

1969 Or, 80). However, distribution between regions on grounds of their different needs

or histories is acceptable (Sagar v. State of Andhra Pradesh, A.1.R. 1968 A.P. 165; Joseph Thomas v. State of Kerala, A.1.R. 1958 Ker. 33 [division of seats between Malabar and

Travencore-Cochin based on historical reasons]). Cf. Muralidhar v. State of Andhra Pradesh, A.1.R. 1959 A.P. 437 (distribution between metroplitan area and hinterland

acceptable).

103. Descent: in effect the commonly used criteria of caste and tribe contain an element

Backward Classes and the Judiciary—Selection

217

All governmental measures which distinguish between groups of citizens must meet the standards of ‘equality before the law” and “equal protection of the laws” laid down in Article 14.1% These tests, evolved in application of the general equal protection provision of Article 14, are well settled.1% It is also well settled that, owing to the intimate connection between Articles 14 and Articles 15 and 16, similar

standards of reasonableness are applicable to governmental measures challenged under the latter provisions.1° Thus where a measure is challenged as violative of Articles 15 or 16, of descent. See the discussion of converts to these groups in chap. 9, §F. Place of Birth: reservations for rural students have been invalidated as discrimination on grounds of place of birth (State of U.P. v. Pradip Tandon, A.1.R. 1975 S.C. 563, 569, discussed in n.

101, above). Inhabitants of specified areas can be deemed Backward Classes if the place

of birth criterion is avoided. Cf. Janki Prasad v. State of J. & K., A.1.R. 1973 S.C. 930, 943.

Race; Tribal classifications may be thought of as “on racial ground [s]"” (Mahendra Nath v.

State, A..R. 1970 A. & N. 32, 34). Apart from this the use of racial classification is confined to Anglo-Indans (discussed in n. 50 of chap. 9, below). Cf. the recommendation of the Backward Classes Commission (I BCC 28) that the small Eurasian commun-

ity of Trayencore be deemed a Backward Class. Sex: It is clear that the State may take

special provision for women independently of their membership in backward classes. Art. 15 is qualified by a proviso that the State may make “any special provisions for women” (Art. 15[3]), Unlike special provision for backward classes, that for women need not be favorable to them (Yusuf Abdul Aziz o. State of Bombay, A.1.R. 1954 S.C. 321).

Art. 15(3) qualifies only Art. 15 itself and does not engraft an exception on the prohibition of sex as a criterion in Art. 16. See Dattatraya v. State of Bombay, A.1.R. 1953 Bom. 311, and Venkataramana v. State of Madras, A.1.R. 1951 S.C. 229. Cf. State of Madras v. Champakam Dorairajan, A.1.R. 1951 S.C. 226 at 228, where the presence ofa backward

classes proviso in Art. 16 was thought to make the absence of one in Art. 15 significant.

Art: 16 concerns government employment specifically and overrides the more general

Art. 15 in this area. See chap. 11, §B, below. Reserved posts for women would depend on

their designation as Backward Classes. That such an eventuality is not too farfetched is licated by the recommendation of the Backward Classes Commission that all women

in India should be considered a Backward Class (1 BCC

31). While recommending

special concessions in education, the Commission did not mention government service (td., at 32). Cf. the suggestion of the court in the Dattatraya case that special provisions for women might not be discriminatory on grounds of sex “only,” since women as a group have been subject to other common influences and disabilities which make them a definable group. 104. But cf. the holdings of the Madhya Pradesh High Court that non-statutory rules

(e.g., university admission rules) are not “law” and therefore are not subject to the

reach of Art. 14 (although they would be covered by Arts. 15 and 29). Vinod Sagar Sood 0.

State of Madhya Pradesh, A.1.R. 1967 M.P. 182; Gokul Prasad v. Sohani, A.1.R. 1962 M.P. 126; Vishnu v. State of Madhya Pradesh, A.1.R. 1961 M.P. 247.

105. Budkan Choudhry ve State of Bihar, A.1.R. 1955 S.C. 191; Bidi Supply Co. v. Union of

India, A.1.R. 1956 S.C. 479.

106. State of Mysore v. Narasinga Rao. A.1.R. 1968 S.C. 349. This has been specifically

noted in litigation about preferences (General Manager v. Rangachari, A.1.R. 1962 S.C.

36). In Devadasan v. Union of India, A.1.R. 1964 S.C. 179 at 185, the Court notes that the

argument under Art. !4 is identical with that under Art. 16(1). For an early and clear

218

Identifying the Beneficiaries

the first test is that the classification on which [the government action] is

founded must be based on an intelligible differentia which distinguishes

persons or things grouped together from those left out of the group; and the

second is that the differentia in question must have a reasonable relation to the object sought to be achieved by the rule or statutory provision in question. . . .In every case there must be some nexus between the bases-of the classification and the object intended to be achieved by the statute.1°7

Ordinarily, there is a presumption in favor of the constitutionality of any enactment. The burden lies on the challenger to show a clear transgression of constitutional limits.1° However, it is not clear that such a presumption arises where the State classifies citizens according to the criteria forbidden by Articles

15 and

16.19 When

the State

employs caste and religion to classify backward groups, who has the burden of showing the relevance and adequacy of the classification? The first High Courts to address this question took divergent courses. In Kerala a majority ofa Full Bench clearly placed on the challenger the burden of proving that a scheme of reservations is not justified.!° The Andhra Pradesh Court, however, suggested that the onus of supporting

the classification lies with the State.121 In upholding that court’s decision in Sagar, the Supreme Court declared that “‘the conditions which justify departure [from the strictures of Art. 15(1) and 29 (2)] must be application of the Art. 14 standards to the Art. 15 area, see Partha v. State of Mysore, A.LR. 1961 Mys. 220 at 229. The incorporation of Art. 14 standards has been criticized

by Havanur (1965 : 115). 107. Pandurangarau v, Andhra Pradesh Public Service Commission, A.1.R. 1963 S.C. 268 at

271.

108. Moti Das v. Sani, A.1.R. 1959 S.C. 942; Mohd. Hanif Qureshi v. State of Bihar, A.1.R.

1958 S.C. 731. 109. Cf. the observations of Patanjali Shastri, J., in Kathi Raning Rawat v. State of

Saurasthra,

III

(1952)

S.C.R.

435

at 442,

that

while

“equal

protection

claims

under... Article [14] are examined with the presumption that the State action is’ reasonable and justified” where bias on any of the grounds mentioned in Arts. 15 or 16 is

disclosed, “it may well be that the statute will, without more, incur condemnation as

violating a specific constitutional prohibition unless it is saved by one . . .of the provisos to those articles.” A Full Bench of the Punjab High Court has held that a classification on any of these forbidden grounds is “bad without.anything more having to be proved,” since a classification forbidden by the Constitution cannot possibly be said to be reasonable (Pritam Kaur v. State of PEPSU,

A.1.R. 1963 Punj., 9 at 17). (Provision of

PEPSU Court of Wards Act allowing government to deprive a woman of management ofan estate where similar mismanagement by a male would not be grounds for removal is invalid.) 110. Hariharan Pillai v. State of Kerala, A.J.R. 1968 Ker. 42. Gopalan Nambiar, J., dissenting, would place the burden clearly on the State (id., at 53).

111. Explicitly in Sukhdev v. Government of Andhra Pradesh, 1966 Andh. W.R. 294 at 303, 304, and implicitly, by adopting exacting standards for data to support the classification in Sagarv. State of Andhra Pradesh, A.1.R. 1968 A.P. 165.

Backward Classes and the Judiciary—Selecton

219

strictly shown to exist.”"212 The Court would not assume that constitutional criteria were employed; mere assertions by the state that relevant criteria had been taken into consideration were not sufficient. Subse-

quently, another Supreme Court bench declared that “‘the onus of proof is on the State. to establish that the reservations are for socially and educationally backward classes of citizens.” 213 Whether or not the presumption of constitutionality technically

arises, jt is clear that such classifications suffer under an onus of

constitutional disrepute and will attract intense judicial scrutiny. The emergence of stringent judicial tests for sufficient data to support the classification indicates that wherever the burden of proof technically rests, it is the primary responsibility of the State to justify its classification!14— at least where the challenger has an interest in the composition of the whole category of Backward Classes.115 Where the matter ac issue is not the eligibility of the beneficiary groups but other aspects of the scheme of preferential treatment, no such burden is carried by the State: the normal presumption of constitutionality would seem to obtain. Thus ina case involving the operation of reservations for Scheduled Castes, where designation of beneficiaries was not an issue, the Supreme Court observed that “the 112. State of Andhra Pradesh v. Sagar, A.1.R. 1968, S.C. 1379 at 1384.

113. State of U.P. 0. Pradip Tandon, A.1.R. 1975 S.C. 563 at 569. But cf. Hridaya Narain

Singh v. Mhd. Sharif, A.1.R. 1968 Pat. 296, discussed in n. 101 of chap. 8. 114. Something more than mere assertions of unconstitutionality may be required to trigger the state’s obligation. See Periakaruppen v. State of T.N., A.I.R. 1971 S.C. 2303 at 2310-11. In some instances the courts are willing to relieve the state of this burden by taking judicial notice of the notorious backwardness ofa particular community (see Chait Ram v. Sikander A.1.R. 1968 Pat. 337) or a composite group like Harijans (Swain v.

Secretary A.1.R. 1974 or. 115). 115. In Chait Ram v, Sikander, A.1.R. 1968 Pat. 337, the Full Bench of the Patna Court dealt with a petitioner whose interest in challenging the Backward Classes designation

was somewhat different from that of the frustrated competitor for a government job or medical college seat. The state had made immune from sale in execution of decrees the raiyati holdings of members of the Backward Classes. Petitioner was a creditor of a

member of the Mallah community, a caste of fishermen and boatmen, and challenged their inclusion of the Backward Class list. The Court said that ordinarily the decision of

the state government that a particular class or caste is socially and educationally backward would prevail, subject to the right of petitioner to satisfy the court that the test of backwardness adopted was irrational or irrelevant. Government's classification

here was supported by gazetteers and common knowledge and petitioner offered no

refutation. Even if the State carries an onus with respect to selection of the entire Backward Classes category, the creditor here, who is concerned specifically with dis-

crediting the backwardness of the particular group to which his debtor belongs, need not

be the beneficiary of a rule designed to make possible the vindication of rights by those

with an interest in the composition of the entire Backward Classes category.

220

Identifying the Beneficiaries

burden of establishing that a particular reservation made by the State is offensive to Art. 16(1) is on the person” challenging it.116 The stringency of the standards of reasonable classification in designating Backward Classes derives from the exceptional nature of the authorization for preferential treatment. “Being in the nature of an exception, the conditions which justify departure must strictly be shown to exist.”117 Articles 15(4) and 16(4) permit the use of otherwise forbidden criteria for the achievement of specific objectives. The courts have been anxious to confine their use to those objectives, to prevent the expansion of group preferences into a general principle of operation,

and to reconcile these exceptions with their context of general fundamental rights. Thus the Supreme Court has continually emphasized the necessity of balancing the provisions for preferences against the general scheme of equality: It is implicit in. . . [Art. 16] that the doctrine of equality of opportunity

shall be reconciled with that of reservation in such a way that the latter while

serving the cause of the backward classes shall not unreasonably encroach upon the field of equality.118

Presumably, if other (i.e., non-forbidden) criteria were used to select Backward Classes, the State’s classification would still have to meet the

general equal protection standards (intelligible differentia, reasonable relation to the object, etc.). But if it is the technically “exceptional”

character of Articles 15(4) and 16(4) that is the basis of the enhanced stringency we have observed, then it is an open question whether it

would obtain where the scheme of “compensatory discrimination” did not technically exercise the powers conferred by these exceptions to deviate from the general provisions of Articles 15, 16, and 29. Would the

ordinary presumption of constitutionality and the burden of proof on

the challenger reassert themselves in the area of compensatory discrimi-

nation as reliance on forbidden communal criteria declined?

The courts have, as one scholar remarked of Balaji, “virtually read

the word ‘reasonable’ into” the provisions for compensatory discrimination.119 We shall in the sequel be examining various attempts of 116. State of Punjab v. Hira Lal, A.1.R. 1971 S.C, 1777 at 1780. Cf. State of Kerala v. N. M. Thomas, A.I.R. 1976 S.C. 490 at 524, where Justice Beg clearly places on petitioner the burden of establishing “a constitutionally unwarranted discrimination.”

117. State of Andhra Pradesh v. Sagar, A.1.R. 1968 S.C. 1379 at-1383. 118. Triloki Nath Tikku v. State of Jammu and Kashmir [1] A.1.R. 1967 S.C. 1283 at 1285; cf. State of Andhra Pradesh v. Sagar, A.1.R. 1968 S.C. 1379 at 1383; Balaji v. State of Mysore, A.LR. 1963 S.C. 649 at 662.

119. Rama Rao 1967 : 82. A clear example is provided by Partha v. State of Mysore, A.LR. 1961 Mys. 220, 229, where it is observed that the restraint on the fundamental rights of others “must be a reasonable restraint. ...It is ...well established that a

Backward Classes and the Judiciary—Selection

221

the courts to decide what is a reasonable use of this power by balancing competing factors. At this point it should be recalled that judgments of reasonableness may differ, reflecting deeper differences about the context and purposes of compensatory discrimination.12° What is reasonable depends upon the range of factors which are considerea relevant and the weights that are assigned them. restraint can be reasonable only when it is necessary to achieve a particular purpose and

should

not exceed what is necessary for the achievement of that purpose.”

120. ‘No abstract standard, or general pattern, of reasonableness can be laid down as

applicable in all cases. The nature of the right alleged to have been infringed, the underlying purpose of the restrictions imposed, the extent and urgency of the evil sought to be remedied thereby, the disproportion of the imposition, the prevailing conditions at the time, should all enter into the judicial verdict. In evaluating such elusive factors and forming their own conception of what is reasonable, in all the circumstances of a given case, it is inevitable that the social philosophy and the scale of values of the judges participating in the decision should play an important part, and the limit to their interference with legislative judgment in such cases can only be dictated by their sense of responsibility and self-respect and the sobering reflection that the Constitution is meant not only for people of their way of thinking but for all, and that the majority of the elected representatives of the people have, in authorizing the imposition

of the restrictions, considered them to be reasonable” (State of Madras v. V.G. Row,

A.LR. 1952 S.C. 196, 200).

8

The Backward Classes and the

Judiciary: the Measure of Backwardness

PREFERENCES FOR members

of a particular group are

not permissible if that group is not a Backward Class.1 In Venkataramana v. State of Madras,? the Madras Government had filled a group of posts

in accordance with its “Communal G.O.,” which divided all available

posts according to quotas among Harijans, Backward Hindus, Muslims, Christians, Non-Brahmin Hindus, and Brahmins. The Supreme

Court held that reservation was permitted solely for hackward classes and that it was “in the circumstances, impossible to say that classes of people other than Harijans and Backward Hindus may be called Backward Classes.”’3 The denial to petitioner, a Brahmin, of the opportunity to compete for posts reserved for other non-backward groups was caste discrimination within the prohibition of Article 16(1) and (2). In State of

Jammu and Kashmir v. Jagar Nath,‘ a cabinet order authorizing direct appointment of Muslims to certain posts, “to remove the communal disparity,” was held invalid; since there was no designation of the preferred group as backward. The same point was made by the Supreme

Court

in regard

to a subsequent Jammu

and

Kashmir

scheme.’ There must be a finding that the group in question is backward, something more than “a general assertion, unsupported by acceptable data.” 1. Cf. Dr. Ambedkar’s observation to the Constituent Assembly that “unless you use some such qualifying phrase as ‘backward’ the exception made in favour of reservation will ultimately eat up the rule altogether” (VII CAD 702). Preferential treatment for non-backward groups is taken up in chap. 12, §D. 2. A. 1.R. 1951S. C. 229. 3. Id., at 230 Cf. A. R. V. Achar v. State of Madras, Writ Petition No. 568, High Court at Madras, Aug. 25, 1952, aff'd on other grounds A. I. R. 1954 Mad. 563.

4. A. 1. R. 1958J. & K. 14, affg A. I. R. 1958J. &. K. 1. An alternative ground for

rejecting this scheme might have been that the device employed was not “reservation,” but the Court did not deal with this question.

5. Triloki Nath Tike v. State of Jammu and Kashmir{I), A. 1. R. 1967 S.C. 1283.

6. Id., at 1284.

The Measure of Backwardness

223

Thus, some finding by the State that the group (for whom the reservations are made) is a backward class is required. But the term “backward” is undescriptive: what criteria may the State use in determining whether a group is backward? Furthermore, “tbackwardness” is obviously a relative term: what may the State use as a standard and where may it establish the cut-off point? What kind of evidence must it have to support its application of these criteria? What kind of control are courts empowered

and competent

to exercise over the

standards employed by the State and over the way they are applied? A. EDUCATIONAL BACKWARDNESS

The judicial response to these questions emerged in good part from a series of cases involving Mysore’s schemes of reservation. Before the problem was encountered by the Supreme Court in the Balaji case, the Mysore High Court had already begun to elaborate the concept of educational backwardness. The old Princely State of Mysore had since 1921 provided extensive reservations for “backward communities,”

defined as all but Brahmins.” These arrangements were carried over

into the post- Independence State of Mysore. When it was attempted to extend them into the enlarged Mysore (formed by the States Reorganization of 1956) the Order was quashed by the High Court in 1958.® This judicial setback led the government finally to list the Backward Classes, and it compiled a list of 164 communities—all the Hindu communities in the state except Brahmins,

Banias, and

Kayasthas

and all the

non-Hindu groups except Parsis and Christians— altogether over 90% of the state’s population. In Ramakrishna Singh v. State of Mysore, the High Court struck down a scheme in which 45% of the seats in technical and professional colleges were reserved for the 164 communities.? In order to decide whether the reservations were constitutional, the High Court undertook to ascertain whether the beneficiaries of this scheme were “socially and education-

ally backward classes as envisaged in Article 15 (4). . . .”?° To qualify 7. For an enlightening account of these developments, see Dushkin 1974. The “all but Brahmins” system survived until the reorganization of the state in 1956 (Mysore Backward Classes Committee 1961: 2). 8. Writ petitions Nos. 369, 370, etc., of 1958, where the Coust quashed the communal

G. O. of 26 July 1958 in an uncontested consent order. (This information is from A. 1. R. 1960 Mys. 338.) Havanur (1965:88) explains the background of this unreported case: each of the five areas in the new state had retained its old list of Backward Classes. Members

of communities which were backward in old Mysore (i.c., all but Brahmins) but not in other areas of the state, succeeded in pressuring the Government to extend the old

Mysore order to the new areas in July 1958. 9. A. IR. 1960 Mys. 338. 10. Id., at 346.

224

Identifying the Beneficiaries

as such, they must have been selected by some _ intelligible principle” designed to further “the policy and object of the Constitution . . . to ameliorate the conditions of really backward classes... .” 11 Finding that communities with high percentages of literacy were included, the court struck down the state’s classification as arbitrary on the ground that “literacy is the only possible test for determining educationally

backward classes.””12 The Mysore Government responded by appointing a committee to suggest criteria of backwardness and the exact manner in which they should be applied, both for purposes of admissions to professional and technical colleges and for government posts. The Nagan Gowda Committee was formed on January 9, 1960, and submitted an Interim Report on February 19, focussing on a tentative classification for purposes of reservations in government service. It proposed and

apphied a dua! test of literacy and representation in government service:

communities above the state average on either of these measures were eliminated, leaving a list of 168 communities (an estimated 35.34% of

the state’s population), for whom the Committee recommended a re-

servation of 36%.3 Lingayats and the Bhunt section of Vokkaligas were excluded from the Committee’s list by the literacy test; Muslims

were excluded by the Government service test; the remaining Vokkaligas were on the list. On March 1, the Government issued an order adopting the Committee’s list for purposes of reservations of government posts, but reducing the reservations from.the recommended 36% to 25%. As the season for admissions to medical and technical colleges neared and the Committee’s final report remained unfinished, the

Government on June 9 adopted the same list for reservations in admis-

sions, ordering a reservation of 22%. When that year’s admissions were challenged in the High Court, the interim list survived challenge. The

Court found literacy an “intelligible” and, under the circumstances,

a reasonable measure of educational backwardness and was tolerant of the Committee’s extrapolation from old census data. “No one has suggested that the list includes any community which is not really socially and educationally backward.” 45 Indeed, there was an unsuc11. Id,, at 347. 12. Id., at 348. The Court indicated that it did not consider literacy in English a suitable test. 13. The number of communities is somewhat misleading. The population figure of 35.34% is for 13 major communities with 128 “sub-communities.” In addition, there were 40 other groups for whom no individual population figures were available but whose total population was probably less than 3% of the state total. 14. Partha v. State of Mysore, A. 1. R. 1961 Mys. 220. 15. Id., at 232.

The Measure of Backwardness

225

pesaful challenge by those who felt they were unfairly excluded from the ist.16 In preparing its final report, the Committee shifted its test of educational backwardness from literacy to enrolment in the last three years of high school: communities were deemed educationally backward if they had fewer high school students per thousand population than the state average (6.9 per thousand).17 The interim report had not developed any separate test of social backwardness. The Committee considered and rejected income, occupation, and vama as possible tests and decided that it would employ the test of status accorded to the communities by society in general,*® though it is not clear how far they actually applied such a test. (In addition, for purposes of government posts,

they

added

the

test of representation

in government

ser-

vices.) These tests were applied to large clusters like Lingayats’? and Muslims; among Vokkaligas they were applied separately to Bhunts and others. Among the backward the Committee distinguished those who were

“more

backward”

purely by its educational test— those

whose high school enrolments were below half of the state average were

“more backward”. In its final report, submitted on May 15, 1961, the

Committee found 214 communities (57% of the population) backward

for purposes of Article 15(4): 79 backward (33%) and 135 more back-

ward (24%). For puposes of government 185 groups backward (45% of the total (21%) and 135'more backward (24%). Of Vokkaligas (minus Bhunts) were on both

posts, the Committee found population): 50 backward the large borderline groups, lists; Muslims were on the

16. Particularly the Lingayats. On these claims, see §1, below.

17, The entire procedure was based on the assumption that accurate population

figures could be extrapolated from the old census reports. There had been no caste

enumerations on the 1951! or 1941 censuses, so the population figures were projections

from 30-year-old data. This, compounded

by the Committee’s unawareness of the

possible bias introduced into its school enrolment test by the differing age profile of the

various communities

(see n. 23 below)

casts some doubt on the precision of the

Committee's figures carried out to several decimal places. 18. Mysore Backward Classes Committee 196: 19. The Committee concluded that

social backwardness in Article 15(4) referred to “social hierarchy based on the caste

system”. Their status test sought to determine “the position of that community in the social hierarchy” “‘ ‘Social backwardness’ of a community should be judged mainly by the status accorded to that community in society in general (apart from individuals). A contributory factor is the general economic condition of that community.” 19, The application of the school enrolment test to the Lingayats was surrounded by heavy dispute. The Director of Public Instruction’s initial figures on Lingayat enrolments showed them with 6.9 students per thousand, comfortably below the state average. Some members of the Committee moved the Director to recheck Lingayat figures, and his second report raised their average to 7.1. The procedure and the figures are disputed by Shri Patil in Mysore Backward Classes Committee 1961 : 34-35.

226 Identifying the Beneficiaries Article 15(4) list but not on the government service list; Lingayats were on neither. The state government was unwilling to allow Lingayats to be omitted, so it raised the educational cut-off point from 6.9 per thousand to 7.1 for the purpose of including them (in the process they had to include the intervening Ganigas), bringing the total Backward Classes to 74% of the population (not counting the Scheduled Castes and Tribes).2°, The state government also refused to carve out the Bhunts from the Vokkaligas on the ground that Vokkaligas had to be treated as a whole. The Government retained the Committee’s recommended figure of 50% for reservations in educational institutions. It was in this scheme that the Supreme Court encountered the question of educational backwardness in Balaji v. State of Mysore.24 The Court found it “doubtful if the test of the average of the student population in the last three High School classes is appropriate in determining the educational backwardness.”2? But it did not press the merits of the student population as a measure,” but directed its concern to the state’s cut-off point. The Mysore scheme included as backward all communities whose enrolment in secondary schools throughout the state was less than the state average of 6.9 high school students

per thousand of population. Assuming enrolment to be a rational test of backwardness,

the Court asked “whether it was legitimate to treat

castes and communities which are just below the state average as educationally backward classes?” It answered that it is not: only communities “well below” the state average may be deemed backward.

(This standard is discussed below in §E.) In spite of the Balaji court’s misgivings about the high school population test, composite tests along these lines have become the most com20. Havanur 1965: 96. 21. A. I. R. 1963 S. C. 649. This was the first of these cases to go directly to the Supreme Court under the Article 32 writ jurisdiction. Counsel opposing the reservations, in spite of their earlier successes in the Mysore High Court, apparently felt that the composition of the bench hearing writ petitions was unfavourable to them and thought their chances were better in New Delhi. The earlier successes and the intense public debate surrounding the work of the Nagan Gowda Committee and the State Government's subsequent actions may have increased interest and made available the resources required for this more expensive undertaking. 22. Id., at 660, The Court also opines that “‘the literacy test supplied by the Census

Reports may not be adequate. . . .” (id.) 23. It would seem that this test contains at least one serious probability of distortion due to differences in the age profile of different communities. A community with fewer youngsters of school age might appear more backward than one with many young people, even ifa higher proportion of the former's young people were in high school than of the latter's. 24. Id., at 660.

The Measure of Backwardness

227

mon way of determining educational backwardness. Thus, Kerala’s 1965 Kumara Pillai Commission utilized a more sophisticated composite test of educational backwardness: communities were educationally backward if they had a low student population in the Xth standard or if the percentage of persons who had not completed primary education was low. Where

Xth standard data was not available, communities

were judged backward if they displayed an abnormally high fall-out (i.e. drop-out) rate. The cut-off points were pegged at slightly below the state averages.25 The

1970 Kerala Commission, however, felt that a

test focussing on the high school level was not appropriate, since many positions required advanced training. The Commission recommended that communities should be considered educationally backward if the percentage of persons in that group passing the Secondary School Leaving Certificate (S.S.L.C.) examinations, or the number of students in first degree programs in arts and science, law, medicine, ayurveda, agriculture, veterinary science, and engineering was lower

than the group’s percentage of the state population.26 The Jammu and Kashmir Backward Classes Committee in its 1969 report determined educational backwardness by comparing the number of students in the 9th and 10th classes per thousand population in each group, with the statewide student population per thousand, but did not specify how the cut-off was arrived at.?7 In 1970, the Andhra Commission, citing the Jammu and Kashmir Committee’s use of standards IX and X, adopted a test utilizing student population in standards X and XI: communities whose student population in these standards was “well below” the state average were educationally backward.28 The Commission did not specify what it meant by “well below” the state average. In the cases of several communities which showed a “slightly higher level of education,” the Commission used personal knowledge of the living conditions of the communities to conclude that they were in fact educationally backward.29 The inclusion of these groups was one of the grounds upon which the Andhra Pradesh High Court struck down the state’s list. But the Supreme Court reversed on the ground that a few such instances was not sufficient to strike down the entire list.3° Balaji’s doubts about student populations as a test, and notions about ‘“‘well below the state average,” had been “misap25. Kerala (Kumara Pillai Commission) 1966: 43-44. 26. Kerala (Backward Classes Reservation Commission) 1971: I, 87. 27. Government of Jammu and Kashmir 1969: 28.

28. Andhra Pradesh Backward Classes Commission 1970: 58-59. 29. Id., at 59.

30. State of Andhra Pradeshv. Balaram, A. 1.R. 1972 S.C. 1375, at 1396. The decision of the High Court in this case was unreported.

228

Identifying the Beneficiaries

plied;” for Balaji did not “lay down any hard and fast rule,” but only indicated “broad principles to be kept in view.”’31 In Karnataka

(as Mysore

was renamed

in

1973),

the Havanur

Commission in 1975 used the S.S.L.C. level in its test of educational backwardness. It compared the number of students in a community passing the examination in April 1972 with the average number of students passing per thousand state population. Groups with less than the state average but more than 50% of the state average were classed as Backward Communities. Separate lists of Backward Castes and Backward Tribes were composed of communities with student averages below 50% of the state average.>? Unfortunately, all of these measures of representation in the student

population suffer from the speculative population figures on which they are calculated. But alternative tests of educational backwardness are still inchoate. One alternative measure was rejected in Dilip Kumar v. Government of U.P.33 The shortage of Higher Secondary schools in an area and low marks in the Pre-Medical Test obtained by residents of the area, were

insufficient to classify all those residents as educationally backward, since the area undoubtedly included classes that were not educationally backward. The Supreme Court in a later encounter with this same U.P. scheme allowed that “lack of educational institutions and educational aids” might be one factor in the educational backwardness of an area, but the Court put more stress on attitudes toward education than on the absence of facilities: “traditional apathy for education on account of social and environmental conditions or occupational handicaps ... [illustrates] educational backwardness.”35 Thus people in the hill and Uttarkhand area of U.P. are educationally Backward

Classes “‘because lack of educational facilities keep them stagnant and they have neither meaning and values nor awareness for education.”>¢ Earlier this ingrained apathy principle had been elaborated by a

Constitutional Bench in Janki Prasad, as a test which seems to measure

both the social and the educational backwardness referred to in Article

15(4).37 The Court remarks the spread of facilities for education into 31. dd., at 1397.

32. Government of Karnataka 1975: I, 315-316.

33. Dilip Kumar v. Governmentof U. P., A. 1. R. 1973 All. 592.

34. Id., at 595. In State of U. P. v. Pradip Tandon, A. 1. R. 1975 S. C. 563, the Supreme Court concurs that the rural areas of U. P. can’t be classified as educationally backward, on the curious ground that candidates from rural areas obtained 85 of 758 seats (12%)

on merit. The Court thought this spoke well of the high standards of education in rural areas, although it estimated that 80% of the population of the state resided in those areas (id., at 568-69). 35. A. 1. R. 1975 S. C. 563, at 567.

36. Id., at 567.

37. Janki Prasad Parimoo v, State of J. & K.. A. 1.R. 1973S. C. 930. at 938.

The Measure of Backwardness

229

rural areas and the growing sector of the village population which has

embraced education as a means of social advancement. However, there

remain “sectors of the population which show extreme apathy towards education due to age-old customs and habits of living, fostered by poverty, ignorance, superstition and prolonged social suppression.”’3* While other sectors in the rural areas deserve encouragement by the State, these apathetic sections “require to be goaded into the social stream by positive efforts of the State.”39 Whether these attitudinal factors can be developed into a workable substitute for, or supplement to, representation in student bodies remains to be seen. The absence of census data suggests that surveys might be an appropriate way of gathering information relevant to educational backwardness. Once such survey apparatus were established, one would be free to devise more precise and differentiated measures of educational backwardness. B. SOCIAL BACKWARDNESS In

addition

to

educational

backwardness,

the

Balaji

Court

scrutinized the Mysore scheme for evidence of the “social backward-

ness” of the beneficiaries. While holding that a classification solely on the basis of caste would be unconstitutional, the Supreme Court suggested that “it may not be irrelevant to consider the caste of the . . . group of citizens.”4° The use of castes as units or classes of citizens designated as Backward Classes has already been discussed in chapter 7 above. But caste may be used in another way: to determine the backwardness of groups and their eligibility for inclusion among the recipients of preferential treatment. These uses involve two different senses of the word “caste.” The use of caste as the unit or community of citi-

zens emphasizes the caste as an entity, an endogamous group or cluster

of such groups with a common name, etc. The use of caste to determine

backwardness emphasizes rank, status, prestige (or their opposites),

and disabilities, the absence of resources and opportunities associated with them. Unfortunately, the Supreme Court does not make this dual usage explicit.

It is quite clear that the Balaji Court finds it permissible to use castes

as the units of citizens to be designated backward. It is in the second

sense of caste that the Court limits its use—i.e., in the sense of rank,

status, or prestige of the group,as a test of its social backwardness. The Court’s first objection to the exclusive use of caste as a test of social 38. Id., at 938. 39. Id., at 938.

40. A. I. R. 1963 S.C. at 659.

‘

230 Identifying the Beneficiaries backwardness is that the test would inevitably break down in relation to many sections of Indian society. How is one going to determine whether Muslims, Christians, or Jains, or even Lingayats are socially backward or not?...[Though] castes in relation to Hindus may be a relevant factor to consider in determining the social backwardness of groups or classes of citizens, it cannot be made the sole or dominant test in that behalf.41

Since it is well known that these non-Hindu groups comprise (or are composed of) er\dogamous communities, the only reason this would not work would be the difficulty of applying to them the notions of caste

rank and standing in traditional Hindu terms. Again, says the Court,

poverty as well as caste is relevant:

that social backwardness which results from poverty is likely to be aggravated by considerations of caste to which the poor citizens may belong

but that only shows the relevance of both caste and poverty in determining the backwardness of citizens.‘?

It is clear that the caste that is objected to here is not the caste as a unit,

but

caste

in the sense of standing

or status—i.e.,

a level of

opportunities and access to resources, respect, etc., which characterizes that group. The Court notes that not only poverty, but also such factors as occupation and place of habitation, go toward determining the backwardness “of a community of persons.”43 It criticizes the Mysore Committee, which “was inclined to treat the caste as almost the sole basis in determining the question about the social backwardness of any community.” “The predominant, if not the sole test that weighed in their minds was the test of caste.”45 And the “classification of the socially backward classes of citizens made by the State proceeds on the consideration only of their castes without regard to the other factors which are undoubtedly relevant.“ Thus, “the social backwardness of the communities . . .has been determined in a manner which is not permissible under Article 15 (4).”47 Inspection of the Mysore scheme suggests that the role of the “‘caste test” in determining social backwardness was neither as dominant nor as clear as the Supreme Court indicates. To be sure, the Nagan Gowda

Committee

announced

that “the status accorded

to [the] commu-

nity in society” would be its test of social backwardness.“* But it 41. Id., at 059. 45. Id., at 660.

42. Id., at 659. 46. Id., at 660.

43. Id., at 659. 47. Id., at 660

48. Mysore Kackward Classes Committee 1961: 19.

44. Id., at 660.

The Measure of Backwardness

231

never specified any findings about status and it never applied any such test. Its list of backward communities is the list produced by application of its education (high school enrolment) test. No group was dropped from the list by the caste status test, nor was any group added. In the case of the three large borderline groups (Lingayats, Vokkaligas, and

Muslims) the Committee merely announced its opinion of their social backwardness. In each case that opinion corresponded exactly with the results of the education test: Lingayats and Bhunts forward, Vokkaligas and Muslims backward. The “caste test” made no visible difference in the listing of these groups. The only real effect of the caste status test was in the case of the many minor communities for whom population figures were not available and to whom, consequently, the school enrolment

test could not be applied. Presumably, it was reputation of their low

status that was used to identify 40 of these groups as backward. These

groups were probably no more than 5% of the total population of Backward Classes. The ghostliness of the caste status test is clearly demonstrated by the way in which the Committee divided its list into Backward and More Backward—it deemed as More Backward those “‘socially backward communities whose standard of education is less than 50 per cent of the state average.”“? The Supreme Court notes at one point how hollow the purported status test is, when it observes that in relation to Muslims the finding of social backwardness was “stated merely as a conclusion and no data or reasons are cited in support of it.”50 But this is literally the case for all communities— there were no findings about their social backwardness.51 It is difficult to escape the conclusion that the state government’s subsequent additions to the Committee’s list were more result-oriented than guided or impelled by any test of caste status. The state altered the educational test so that it accommodated Lingayats*? and announced that they were also socially backward. Even its stated rationale—that

49. Id., at 22-23.

50. A. I. R. 1963 S. C. at 661. The Committee decided Muslims were backward by majority vote. See Havanur 1965: 93-94. 51. The most elaborate entry in the Committee’s chapter on social backwardness

reads in its entirety : “ (ii) Vokkaligas— All sections of Vokkaligas excluding Bhunts are socially backward. Bhunts are socially forward. The standard of education among Bhunts is above the State average, being nine per thousand according to the figures

available to the Committee” (Mysore Backward Classes Committee 1961: 20). It is

particularly revealing that the only piece of evidence the Committee cites is a repetition

of its educational test. 52. The state government

professed concern that it was not possible to obtain

absolute mathematical precision in the matter, so it decided the figures should be rounded to the nearest integer. It thus raised the state average from 6.9 to 7.0 and the Lingayats’ average was lowered from 7.1 to 7.0 (A. I. R. 1963 S. C. at 660-61). (A slightly different account is given in Havanur 1965: 96.)

232

Identifying the Beneficiaries

“it observed that a large percentage of Lingayats live in rural areas and most of them are engaged in agriculture and menial labour and suffer from all the consequences of illiteracy and poverty’’53— does not reveal exclusive reliance on a test of caste status. But if the “caste test” (in the sense of measurement of group status)

played a restricted role in the making of the Mysore list, caste politics was of the essence. The use of caste units, the strong flavor of caste politics, and the announcement that the caste status test was -being applied —all made it easy to overestimate the extent to which it was in fact applied. Perhaps the Court’s alarm at the obtrusion of caste politics and its discomfort with the use of caste units (to which it put forward no objection) found a tempting target in the “caste test” of social backwardness—a target which loomed larger than life because the Court had not arrived at a clear distinction between castes as units and caste status as a factor in backwardness. The state’s use of castes as units is equated with their selection by caste status. Having labelled the scheme as one of exclusive application ofa caste test, the Court passed without mention several problematic aspects of Mysore’s use of this

test. May caste status be determined as casually as the Committee and the state government appear to have done it, without any supporting

evidence, merely on opinion of the committee members? May this status test (and other tests) be applied to such large and diverse composite clusters as Lingayats, Vokkaligas, and Muslims in which homogeneity of status, much less any close correlation with social conditions, is unlikely?s4 If caste standing cannot be the sole or dominant criterion of social backwardness, Balaji provides a catalog of other factors which the State might employ: poverty, occupation, habitation.55 It is not clear 53. A. I. R.

1963 S. C. 649 at 654. Cf. the argument of Shri Patil, the leading

proponent of the Lingayat cause on the Nagan Gowda Committee that Lingayats were

not only educationally backward

(see n. 19 above) but also socially backward. His

argument emphasizes traditional and current occupation and rural residence and

concludes that these are the sources of the group’s social inferiority (Mysore Backward

Classes Committee 1961 : 33-38). 54. Havanur 1965: 93-94.

55. A. I. R. 1963 S. C. 649 at 659. An unusual opportunity to observe a judge putting

his precepts into practice was provided by the circumstance that in 1968, former Chief Justice Gajendragadkar (the opinion-writer in Balaji) constituted a one-man commission to make recommendations about the revision of the list of Backward Classes in

Jammu and Kashmir. In terms that contain recognizable echoes of Balaji, the Commis-

sion recommended the following criteria for selection of the Backward Classes: (1) economic backwardness of a class; (2) occupation(s) of a class; (3) habitation; (4)

average student population per thousand in that class; (5) caste, in relation to Hindus.

Thé Commission does not define what it means by a class, but presumably communal

units are included. (Reported in Backward Classes Review 1 (3): 23.) It is notable that

The Measure of Backwardness

233

whether any of those other measures may be used solely or dominanuly, although Balaj? 's commendation of an income test implied that they

might.56 The extent to which these have been sufficient to support a

designation as backward is discussed in §§ G and H below. Neither Balaji nor any subsequent case has addressed the problem of the incommensurateness of the various kinds of backwardness to which

the state may legitimately attend in designating beneficiaries. There is

no simple way in which low income, low status, undesirable habitation,

etc., can be.added together. But if the State may take all of them into account, are there any limits on the formulae it may use in cumulating them? Perhaps part of the appeal of economic tests— quite apart from the desire to avoid caste—is their susceptibility to quantification. It is established that it is not necessary for all of these factors to be used simultaneously. When Mysore later used only income and occu-

pation as tests of backwardness, the High Court advised it that it was

necessary

to add

tests of caste and residence.5? In Chitralekha, the

Supreme Court repudiated this view, allowing that caste might be omitted entirely if backwardness could be ascertained by other relevant criteria.58 Subsequent Supreme Court decisions have concurred that caste (in the sense of standing or rank) may play some part in the determination of social backwardness, though it may not be the sole criterion.5? But they have had little to add about the way in which caste

status is to be employed.

In Balaji, the most complete judicial discussion of this point, the

Court finds that “caste” is “not irrelevant” to social backwardness,

because “considerations of purity,” “feelings of superiority and infe-

riority,” “narrow caste loyalties,” and the rigidity and inflexibility of

the caste system lead to caste’s playirig an unfortunately large part in determining the status of a citizen. The “social backwardness” which results from poverty is likely to be aggravated by “considerations of caste.”"60 The relation between low status and social backwardness remains somewhat equivocal. Is low status itself to be taken as constituting the

social backwardness to be remedied by preferential treatment? Or is

in spite of Balan’s embrace of the income test, the former Chief Justice does not recommend its exclusive use here. These recommendations provided the foundation for the work of the Backward Classes Committee that reported in November 1969 (Govern-

ment of Jammu & Kashmir 1969).

56. A. I. R. 1963 S. C. 649 at 664.

57. Viswanath v. Government of Mysore, A.1.R. 1964 Mys. 132.

58. A. I. R. 1964S. C. 1823. 59. State of Andhra Pradesh v. Sagar, A. 1. R. 1968 S. C. 1379; P. Rajendran v. State of Madras, A. 1. R. 1968 S.C. 1012. 60. A.I.R. 1963 S. C. 649 at 659. 61. The Balaji Court’s argument for the relevance of poverty and occupation seems to

be that they too confer low status (A. I. R. 1963 S. C. 649 at 659).

234

Identifying the Beneficiaries

low status merely an indicator of a low level of tangible resources and opportunities which are the object of government policy? Both views point to factual questions on which the courts have been presented with little systematic data. If caste status is an indicator of the absence of resources and opportunities, the question arises how strong is the association? To what extent are those characteristics which can be thought of as intimately associated with caste status— low self-image, disabilities, etc.—a significant determinant of the low level of resources or opportunities or ofan impaired capacity to utilize them? Iflow status is in itself tantamount to backwardness, how largea part does caste play in the determination of the status of individuals, families, and groups? How does it interact with occupation, wealth, education, residence, and

cultural traits as a determinant of status?

The answers will turn, in part, on what is meant by status, a question

that is not without conceptual and practical difficulties of its own.

Status can be measured in innumerable ways, and the relative status of a particular caste may differ with the measure employed.® Is status to be measured by the self-image ofa group? or by the regard in which it is

held by those in the locality? the region? Or is status to be measured by

observation of patterns of interaction with other groups? Or is it to be

measured by the possession of attributes which give standing in some hierarchic scale—local, regional, or pan-Indian? Presumably, a measure might be any of these types if there were some plausible connection with the backwardness to be remedied and if the status were ascertainable with some accuracy and without excessive cost.

Presumably, though, there are some measures of status that are not

sufficiently reliable or relevant. It has occasionally been suggested that the socially backward should be equated with Sudras®— the fourth or

lowest vara (class, estate) in the traditional fourfold classification of

Hindu

socio-legal theory.

Varna has long been recognized as both

62. On the multiplicity and ambiguity of status rankings, see Silverberg 1967.

63. Many would accept this proposition as a general description. Thus the Tamil

Nadu Backward Classes Commission (1971: 1, 37) noted: “The castes included in the

Backward Classes are regarded as ‘Savarnas,’ i.c., inside the Chatur-Varna, but largely falling within the Sudra fold.” But taken prescriptively, it is more troublesome. This criterion of backwardness was advocated by Shri S. D. Singh Chaurasia, a member of

the Backward Classes Commission (III BCC 22 ff), but it was rejected by the full commission on the ground that there were a small number of Sudra communites which “thave made some advance in recent decades” ({ BCC 44). The equation of backward-

ness with Sudra status seems to carry more plausibility to northern Indians than to southerners. The late Dr. Ram Manohar Lohia was a proponent of the Sudra theory (Lohia 1964: passim). A similar proposal was rejected by the Nagan Gowda Committee, which found it difficult to justify, “especially when the Committee is aware that some sections of the so-called Sudras are socially well-advanced today” (Mysore Backward Classes Commission 1960: 2).

The Measure of Backwardness

235

inexact in its correlation with social conditions and exceedingly difficult to determine in many cases. Determinations of caste status are rendered difficult by the tendency to describe them in the traditional idiom of purity and pollution. Thus, for example, the Mysore Government characterized farming as a low occupation. But such traditional measures may play a less important

role than it appears to those wno view rural India from the state capital. For example, studies of evaluations of occupations in rural India suggest that purity/impurity values may have relatively little influence in the evaluation of occupations: compared with other societies, farming ranks unusually high in rural opinion, the priesthood unusually low.© Although the ritual or ceremonial standing of a caste © may be one of the factors affecting its status in situ and thereby the extent of opportunity open to its members, there is no reason to assume any exact correspondence.*” There are few examples of courts and governments attempting to assess the social backwardness of a community—particularly that associated with caste status. In Jacob Mathew v. State of Kerala, the single judge found fault with the state’s assessment of the social backwardness of the populous Ezhuva caste. The Government claimed that it classified the Ezhuvas as backward because they had been denied access to the temples of high-caste Hindus, were regarded as “untouchables” and polluting in use of tanks, their traditional occupation was the degraded one of toddy-tapping, there was no intermarriage between them and higher-caste Hindus, and they had been latecomers in the field of education. The Court, upon consulting the state’s District Gazetteers published a year before, concluded that except for the lack of intermarriage, all of these disabilities were more historical than current. Lack of intermarriage, the court asserted, was an expression of social distance not sufficient to establish social backwardness. Reversing, the Division Bench was skeptical that the legal abolition of untouchabil64. On the indeterminacy of varna classifications, see chap. 9, §D. 65. See Marriott 1959: 104. 66. Bopegamage and Veeraraghavan 1967; Harjinder Singh 1967. Cf. Rama Rao

(1967: 83), who questions Balayi’s assertion that traditonal notions of caste status are

relevant to determining backwardness: “Obviously, in the context of the present Indian society, ritual purity ofa caste has nothing to do with social status ofa community which depends upon several factors like the extent of wealth possessed by its members, the political pull it can exert, etc. A poor, orthodox, tufted Brahmin priest is more often an object of ridicule than of veneration, and at any rate, sociologists have observed, and as everyday observation of rural India will bear out, that the non-landowning Brahmins in the villages are subservient to the members of the dominant castes.” 67. Determination of status by governmental authorities may involve a problem of reactivity, as status may be affected by governmental patronage in the form of preferences. (See chap. 10, below.) 68. A. I. R. 1964 Kerala 39.

69. Id., at 60

236

Identifying the Beneficiaries

ity and opening of temples had altered the character of the Ezhuva community.7°

[T]ime alone transmutes the ideals of law into the realities of everyday life.

No one can say that the introduction of progressive measures is the end, not

the beginning, of a process of amelioration. Habits of thought die hard and

slow, and occupations like toddy-tapping carry their special stigma from one generation to another and through the decades of conduct and

behavior.7!

The government and are available. But what the kind of standards establish. We shall take

the courts have to make do with the data that data becomes available will depend in, part on of precision and pertinence that the courts this up in §F below.

After Balaji, discussion of social backwardness gained in sophistica-

tion and rigor, moving away from simple estimates of caste standing to analysis of conditions that impede group advancement. In a thoughtful discussion of the relation of status and backwardness, the 1965 Kumara

Pillai Commission in Kerala equated social backwardness with subjection to disabilities—i.e., those invidous social distinctions which hinder

members of a group by denying them access to facilities for improvement. Social backwardness, then, can be measured by popular conceptions

of the

status

of a community

because

those

conceptions

are

had, like its predecessor, been declared unconstitutional. Upon

the

translated into disabilities which impinge on the tangible opportunities of the group.72 In 1967, Kerala set up another commission, in response to judicial prodding,73 this one to determine who were the Backward Classes for purposes of reservations in state services. While the Kerala Commission was sitting, Andhra set up a Backward Classes Commission to redesignate its Backward Classes after its Cabinet Sub-Committee’s list advent of the Dravida Munnetra Kazagham (DMK) government in

Tamil Nadu in 1969, a mandate to consider the question of who should

be the Backward Classes was pointedly withheld. That state set up a Backward Classes Commission limited to reviewing the progress of the then listed Backward Classes. All of these commissions reported in 1970 70. State of Kerala v. Jacob Mathew 1. L. R. 1964(2) Ker. 53.

71. Hd., at 58-59. Subsequently, the Kumara Pillai Commission carefully weighed

recent advances against the heritage of disability and concluded that the Ezhuvas’ disabilities had not completely disappeared, that caste prejudice still persisted and impinged on the lower income group among the Ezhuvas (Kerala [Kumara Pillai Commission] 1966: 46 ff.). 72. Kerala (Kumara Pillai Commission) 1966: 29-34.

73. See §F, below.

The Measure of Backwardness

237

and all had something to say about the measurement of social backwardness.74 The Tamil Nadu Commission moved the shortest distance away from the simple equation of social backwardness with caste standing, which it measured in terms of public opinion: “Social backwardness is a matter.of local or regional opinion based on taboos and other concepts of a hierarchical society.”75 The Commission was emphatic that poverty is not the sole cause of social backwardness; it is “only one factor affecting social position.”’76 The Kerata Commission thought that Backward Classes under Article 16(4) should be delineated by four tests: education, economic position (because prosperity produced a “home atmosphere congenial to education”), participation in governmental service (and consequent participation

in governmental

power

and

the

benefits

that it af-

fords) ”77 and “social backwardness due to historical reasons.”7® The latter includes such factors as stigmatized occupations, vestiges of caste discrimination, a heritage of purdah; and aversion to education in English. The Commission proposed that from communities that met these

tests, members of families with incomes below Rs. 8,000 should be

considered backward for purposes of Article 16(4). The Andhra Pradesh Commission was more historically minded and more focussed on occupation. Social backwardness was defined in terms of the heritage of restricted opportunities resulting from a group’s traditional association with an occupation which was “inferior or unclean or undignified or unremunerative or . . . [without] influence or power.”’79 These occupational characteristics were applied along with “general poverty vu: une community” and “caste in relation to Hindus.” By the latter, the Commission referred to restrictions on intercourse with the advanced classes that had “hindered social progress in the past by denial of access to educational institutions, public offices and offices -74. The Kerala Commission was appointed on Oct. 17, 1967, and reported on Nov.

30, 1970. The Andhra Pradesh Commission was appointed on April 12, 1968, and

reported on June 20, 1970. The Tami] Nadu Commission was appointed on Nov. 13, 1969, and reported on Nov. 26, 1970. The three commissions exchanged views. The

Andhra Pradesh report mentions a visit to Madras and Kerala (1970: 51), and there was a meeting between the Tamil Nadu and Kerala Commissions (Tamil Nadu

Backward Classes Commission 1971: 1, 231). 75. Tamil Nadu Backward Classes Commission 1971: 1, 36. 76. Tamil Nadu Backward Classes Commission 1971 : I, 37. 77. The Commission carefully distinguished its use of representation in services in measuring backwardness and in assessing under-representation in the services (Kerala [Backward Classes Reservation Commission] 1971 : I, 76). 78. Kerala (Backward Classes Reservation Commission) 1971:

79. Andhra Pradesh Backward Classes Commission 1970: 50 ff.

I, 73 ff.

238

Identifying the Beneficiaries

of public resort

.. . and have left vesugzes which

.. . [constitute] a

serious impediment to social life.”®° This segregation was imposed upon the followers of lowly occupations who in turn were members of “certain castes and communities.”’81 The Andhra Pradesh classification was struck down by the High Court of that state on the ground that the Commission had classified Backward Classes mainly on the basis of caste and had failed to collect the data necessary to ascertain the social and educational backward-

ness of the group.8? The Supreme Court, reversing, found acceptable the use of caste units and thought the Commission had collected “‘sufficient material to enable [it] to be satisfied that the persons included in

the list are really socially and educationally backward.”® Earlier, a Constitution Bench had expressed some doubt about the criterion of traditional occupation when detached from evidence of present circumstance. In Janki Prasad v. State of Jammu and Kashmir the state deemed as Backward Classes 62 “‘traditional occupational classes” including various kinds of artisans, peddlers, and laborers. Traditional occupation was defined as “the main occupation of his ... grandfather’s.”’®4 Finding that it would be wrong to give benefits where the father has been mobile, the Court thought the validity of such a classification required a showing that the family livelihood is still

earned from this occupation.®s

But if present economic circumstances cannot be ignored in determining social backwardness, social backwardness is not to be equated with poverty. Balaji had confidently asserted that social backwardness resulted from poverty. But later courts have not seen the connection as so direct. Even if poverty and backwardness are so related that the well-off members of a community might not suffer from the social backwardness that characterized the community in general, “poverty is not the decisive and determining factor of social backwardness.” In Janki

Prasad the Court indicates that it will not take low income as in itself

constitutive of social backwardness.®7 Although the measure of social

backwardness is never articulated, it seems to lie in the direction of 80. Id. at5l. 81. Id., at 51. 82. See State of Andhra Pradesh y. Balaram, A. 1. R. 1972 S.C. 1375 at 1394.

83. Id., at 1396.

84. A. 1. R. 1973S. C. 930 at 940.

85. Id., at 940.

86. Jayasree v. State of Kerala, A. 1. R. 1976 S. C. 2381 at 2384.

87. Janki Prasad v. State of J. & K., A. 1. R. 1973 S. C. 930 at 937: “In this country,

except for a small percentage of the population, the people are generally poor—some

being more poor, others less poor. Therefore, when a social investigator tries to identify

socially and educationally backward classes, he may doit with confidence that they are bound to be poor. His chief concern is, therefore, to determine whether the class or group is socially and educationally backward.”

The Measure of Backwardness

239

cultural traits—shared attitudes and dispositions that render it unlikely that members of a group will, without special help, take ‘advantage of opportunities to improve their position.®* Thus the Court notes that some sectors of the village population have taken to education, and their embrace of education is “almost the measure of social advance they have made recently.” The socially backward, it seems, are those “sectors of the population which show extreme apathy toward education due to age-old customs and habits of living, fostered by poverty, ignorance, superstition and prolonged social suppression.””® Thus the trend is to think’ of social backwardness in terms of a heritage of deprivation. This is usually visualized in terms of the accumulated effects of low position in a social hierarchy. But one Supreme Court Bench, upholding Uttar Pradesh’s designation of the hill and Uttarkhand areas as backward, included in its catalog of “the traits of backwardness” not only the lack of economic development but the absence of social hierarchy.® Generally, however, social backwardness is visualized in terms of the accumulated effects of low position in a social hierarchy. The most ambitious attempt to characterize social backwardness was the work of the Karnataka Backward Classes Commission, under the

chairmanship of L.G. Havanur, a committed and scholarly student of the subject. In its 1975 report, the Commission set forth a catalog of factors which contribute to social backwardness: (i) [Residence in] rural, isolated and segregated areas.

(ii) Their economic condition is so poor that the majority . . . are incapable of owning land, house or other property. (iii) . . . they are engaged in very unremunerative . . . occupations [that]

are considered unclean or inferior, and therefore low.

(iv) People of the advanced castes and communities entertain prejudice against them. (v) Low status or inferiority associated with their castes makes it difficult 88. Id., at 941. Social backwardness in this view is clearly a trait ofa group—not an “artificial” group like pensioners or small cultivators, but some palpable historic social formation. See the discussion of Janki Prasad in chap. 7, §B, above. 89. Id., at 938.

90. State of U.P. v. Pradip Tandon, A. 1. R. 1975 S. C. 563, at 567, containsa somewhat

puzzling catalog of the traits of social backwardness. “There is no social hierarchy. There are no means of controlling the environment through technology. There is no organization of the society to create inducements for uplift of the people and improvement of economy. Building of towns and industries and] growth of cash economy .. .

are absent .. . . [t]hey do not make effective use of resources. When large areas .. .

maintain a sparse, disorderly and illiterate population whose property is small and

negligible . . . social backwardness is observed.”

240

Identifying the Beneficiaries

. .. to have access to places of cultural training, or to have religious and secular education.

(vi) Many . . . have been segregated from the advanced communities due to social taboos against inter-dining and inter-marriage, and against similar opportunities of association.

(vii) [They are] prevented from enjoying status in society due to the age-old

social customs.

(viii) [consequent] . . .apathy for education®*

Although there are formidable conceptual and measurement prob-

lems in operationalizing such a catalog, it is clear that the notion of

identifying the Backward Classes has moved far beyond summary assertions about their status. The Commission was emphatic that “we have ignored the principle of caste—the principle of status—the principle of high and low—in deciding social backwardness.”"9? Instead, the Commission devised a complex test based on poverty and isolation, social enforcement and reinforcement of that poverty and

isolation, and the development of cultural traits that are adaptive to poverty and exclusion. The basic principle for ascertaining social backwardness was “‘eco-

nomic backwardness based on occupational and residential backward-

ness.”93 Applying its test of backwardness, the Commission included among the Backward Classes several castes of temple functionaries as well as a group regarded as Kshatriyas, and it excluded some groups with lower ritual status on ground of their ‘economic and educational advancement.” The Havanur report, the most comprehensive and thoughtful of the Backward Classes reports, is marked by a conceptual mastery of the

subject that has so far evaded: the courts. The Commission moved beyond the confusions imparted by the inability to separate caste as a unit and caste as a status. It set a challenging agenda for conceptualization and measurement of backwardness that may help to address the real and perplexing questions of whether, how, and at what cost

disparities among communities can be dispelled.

C. CUMULATIVE BACKWARDNESS The relation between educational and social backwardness is entwined with the relationship between Articles 15(4) and 16(4). Article

16(4) refers to ‘‘any backward class,” but Article 15(4) mentions classes 91. Government of Karnataka 1975: I, 313-14.

92. Id., at 307. Cf. id., at 287.

93. Id., at 267.

94. Id., at 307.

The Measure of Backwardness

241

that are “socially and educationally backward.” If the difference in wording?5 can support any inference, it is that Article 15(4) covers a

narrower spectrum of groups while Article 16(4) includes also those

who are backward in other respects than “socially and educationally.”96 Governmental bodies— including the Backward Classes Commission and the Home Ministry—have never seen fit to make any distinction whatsoever.9” No emphasis was put on these qualifying words until the high court in Ramakrishna Singh, in striking down the scheme of reservations there, objected that the scheme provided no standard for the determination of social backwardness.% “It would not be enough ” said the court, “to say that these communities are educationally backward. It will also

have to be seen whether they are socially backward.” The court thus

read the phrase “socially and educationally backward classes” con-

junctively: to be entitled to preferences, a group must be both socially backward and educationally backward. It would seem equally plausible

to read the phrase disjunctively,1 allowing government to give preference to groups which are either socially backward or educationally

backward.1°1 But the Balajt Court, anxious to impose a more restrictive notion of the legitimate beneficiaries in opposition to the expansive

Mysore view, found this conjunctive reading convenient. “The backwardness under Article 15(4) must be social and educational. It is not either social or educational, but it is both social and educational.”192

95. See the explanation of Prime Minister Nehru (n. 55 to chap. 6, above) that the

intention was to align the wording of Art. 15(4) with that of Art. 340. 96. Art. 16(4) could, of course, be narrower instead of broader and include only those

who are backward in all respects, not only socially and educationally.

97. Cf. the Nagan Gowda Committee’s conclusion that the backward classes refer-

red to in Art. 16 (4) are the socially and educationally backward classes of Art. 15(4)

(Mysore Backward Classes Commission 1961: 17). 98. A. I. R. 1960 Mys. 338. 99. Id., at 348. 100. ‘lhe parliamentary debate over Art. 15(4) provides no indication about the preference of the drafters. Near the end of the debate, Prime Minister Nehru, rejecting an amendment that would have added “economically” explained that these words were chosen “because they occur in Article 340 and we wanted to bring them bodily from

there. Otherwise, I would have had not the slightest objection to add ‘economically’.

But ifI added ‘economically’ I would at the same time not make it a kind of cumulative thing but would say that a person who is lacking in any of these things should be helped”

(Parliamentary Debates, Vol. X1I-X111 [Part IT], col. 9830 [1 June 1951]).

101. This result may be approximated by keeping the burden of disproving backwardness on the party who complains about a scheme. Cf. Hridaya Narain Singh v. Mhd. Sharif,

A. I. R. 1968 Pat. 296 (where educational backwardness of a community is beyond question, lack of a showing of social backwardness is not fatal, where there was no data to show that they were not socially backward). 102. A. I. R. 1963 S. C. 649 at 658. The Court provides no argument for this reading, but an interesting argument in its favor is put forward by the Kumara Pillai Commis-

242 Identifying the Beneficiaries The conjuctive reading propels the courts into serious difficulties,

both textual and substantive. First, there is the textual paradox. The great weight which the conjunctive reading places on the qualifying

adjectives in Articles 15(4) suggests that their absence in Article 16(4) is significant and that the latter provision covers a much wider range of groups. Yet generally the courts have refused to attribute any import-

ance to the distinction in phrasing. Thus they have held the Balaji tests of backwardness applicable to the government employment area, governed by Article 16. In Dasa Rayudu v. Andhra Pradesh Public Service

Commission the High Court says flatly that the “absence of some words in Article 16 (4) hardly makes any difference.” !°9 In Triloki Nath Tiku v.

State of Jammu and Kashmir [I] the Supreme Court clearly indicates that Article 15(4) and Article 16(4) backwardness are commensurate. The

principles of Balaji, which turned on Article 15(4), are “equally applicable” to the area of government employment, regulated by Article 16(4).1° It seems somewhat paradoxical for the Court put so much

emphasis on “socially and educationally” in Article 15(4) and then to turn around and say that their absence in Article 16(4) makes no

difference.1% Second, the conjunctive reading would severely restrict the flexibility of State policy. It seems to imply that it would be unconstitutional for the government to provide special educational aid to a community which was educationally backward ifit were not also socially backward; nor would it be able to give aid, even of a non-educational type (e.g., loans or housing), to a group that was not educationally backward but might suffer unusual economic or other disadvantages. Finally, the conjunctive reading would seem to constrict—or perhaps invalidate altogether— the use of purely economic tests of backwardness. If, for example, a test of family income (of the sort that Balaji itself commends)

sion. Contrasting the State’s duty to provide educational facilities for all with the special commitment to provide educational aid to those classes whose educational backwardness is the result of the denial of opportunities entailed by their social backwardness, the

Commission concludes that special provisions under Art. 15(4) are intended only for

those who suffer this conjunction of disadvantages (Government of Kerala 1965: 15-16). 103. A. I. R. 1967 A. P. 353 at 361. 104. A. I. R. 1967S. C. 1283 at 1285. 105. But cf. the dissenting judgment of Gopalan Nambiar, J., in Hariharan Pillai v. ‘State of Kerala, A. 1. R. 1968 Ker. 42 at 53, which draws a sharp distinction between the coverage of Arts. 15(4) and 16(4). He asserts that while backwardness for purposes of Art. 15(4) must be “cumulative” (i.e., both social and educational), for purposes of Art.

16(4) any sort of backwardness will suffice. Even the majority in that case suggests that Art. 16(4) “has perhaps a wider import than Article 15(4),” but is sure that “the

“backward class’ mentioned therein refers to the weaker sections of the citizens ” (id., at

48).

.

The Measure of Backwardness

243

were used to identify the backward, there would be no way of knowing that the benficiaries were drawn from groups or families that were socially and educationally backward. Notwithstanding these conceptual and practical difficulties, the conjunctive reading has become firmly established.? It is not clear that it adds much of a hurdle where a court is inclined to credit a group’s backwardness, for as the Supreme Court suggests in Janki Prasad, failure to take advantage of educational opportunities may be the best indicator of social backwardness.'°” Conversely, the Court noted that “Gf a class as a whole is educationally advanced, it is generally also socially advanced because of the reformative effect of education of that

class.” 108 Perhaps the enduring appeal of this formula lies in the scope .that it gives courts for assessing schemes: “educationally” enables a court to demand evidence of backwardness that can’t be reduced to the low status of a group; “socially” enables the court to ask for something more than merely some educational data about a group. One cost of this

firmer judicial control is to tie compensatory discrimination schemes firmly to such groupings as caste, locality, or occupation whose social and educational backwardness can be assessed separately—unlike income strata. The formula also supplies a rationale for court approval of income cut-offs which eliminate from benefit coverage the affluent sections ofa

backward group. In Jayasree v. State of Kerala, the Court found it unob-

jectionable to eliminate the rich because they were not socially backward.!® This raises the possibility that a court itself might impose such a cut-off or require the state to impose one in order to ensure that only the socially backward were included. D. UNDER-REPRESENTATION The courts have read the qualifying words “socially and

cationally”

in Article

15(4)

as an

implicit

explanation

edu-

to Article

16(4). However, the designation of recipients of reservations in government employment under Article 16(4) is distinctive. For Article 16(4) permits reservations in favor of any “backward class of citizens which in the opinion of the State is not adequately represented in the services under the State.” 106. Thus in Dilip Kumar v. Goo’t. of U.P., A. 1. R. 1973 All. 592, 595, the Court finds the government designation insufficient as to the educational backwardness of the areas and totally lacking ‘application of mind” as to the social backwardness. 107. Janki Prasad v. State of Jammu and Kashmir, A.1.R. 1973 S.C. 930 at 938. 108. /d., at 937-38. 109. A. I. R. 1976 S. C. 238] at 2384.

244

Identifying the Beneficiaries

A determination of inadequacy of representation is a necessary but not a sufficient condition for reservation. Before reservation is permissi-

ble, the inadequately represented class first must be determined to be backward. In Triloki Nath Tiku v. State of Jammu and Kashmir [1] the

Supreme Court noted that the power to make reservations under Artic-

le 16 (4) may be exercised only in favor of backward classes:11© “[W]hether a particular class of citizens are backward is an objective factor to be determined by the State.”121 The State’s determination of

backwardness ‘“‘is not the final word on the question: it is a justiciable issue.”112 As to under-representation, however, the opinion of the State “may ordinarily be accepted as final, except when it is established that there is an abuse of power:”13 Thus the determination of the inadequacy of representation is left to the discretion of the State— presumably subject to some minimum standard of plausibility. Just where the minimum lies is not yet clear. A percentage of representation in government posts lower than a group’s population in the state is the commonly proffered measure of inadequate representation.144 A more

exacting standard has sometimes been applied: thus the Andhra

Pradesh Commission imported (from Balaji’s discussion of educational backwardness) the notion that representation in the services must be “well below the state average”’ to determine inadequacy of representa-

tion.15 But the general assumption is that disparity with population

figures is sufficient. The disparity need not be massive to support a judgment of under-representation. Thus in the Thomas case, although it did not turn technically on a finding of under-representation under Article 16(4), the Court concludes that the fact that Scheduled Castes

and Scheduled Tribes made up only 8.7% of employees in the Registration Department (while they were approximately 10% of the population of the state) “clearly shows [they] ... were under-represented.’”"116 Provisions to overcome such under-representation are not automatically invalid if they produce “‘over-representation” of the beneficiaries in a particular grade.1?7 Is the backwardness required to support reservation under Article

16(4) the same as that required for measures under Article 15(4)? It 110. A. I. R. 1967S. C. 1283. 111. fd., at 1285. 112. Id., at 1285. 113. Id., at 1285.

114. Typically, such representation is measured by the total of'appointments secured,

but the Kerala Backward Classes Reservations Commisson (1971: 1, 105) ascertained

adequacy of representation by counting only posts gained in open competition, excluding those gained by reservation to prevent any “relapse into inadequacy” by cutting off reservations.

115, Andhra Pradesh Backward Classes Commission 1969: 107. 116. A. I. R. 1976S. C. 490 at 554. 117. Harpartap Singh 0. Union of India, 1970 S. L. R. 40 (Punj. & Haryana).

~

The Measure of Backwardness

245

might be argued that under-representation is itself a species of back-

wardness and that Article 16(4) which, unlike 15(4), is not confined only to the “socially and educationally backward” authorizes reservations for groups which are backward in this specific respect.118

Inadequate representation has been employed by the Backward Clas-

ses Commission and by other bodies as one criterion in designating Backward Classes.119 However, it is clearly used as an indicator and component of educational and social backwardness, rather than as in

itself a kind of backwardness sufficient to justify reservations.12° But

representation in government service is not invariably regarded as a

suitable measure of backwardness. The Havanur Commission in Kar-

nataka deliberately omitted it on the ground that “inadequacy of representation is the result of social and educational backwardness and not the cause.”’121 Even if the absence of “socially and educationally” in Article 16(4) might conceivably indicate that a backward class for purposes of that Article need not meet the same tests of social and educational backwardness required by Article 15(4), it does not justify dispensing with some requirements of backwardness beyond mere under-representa-

tion.122 To permit " under-representation as a sufficient justification for reservations would in effect read the word “backward” out of Article

16(4) and would open the door for a general regime of communal

quotas.123 It would also, as the Supreme Court points out in Triloki Nath, confer a windfall on groups which had taken to occupations other than government service. The Court there concludes that there are two

distinct conditions for the application of Article 16(4) : a class of citizens _ 118. In the now discredited case of Keseoa Iyengar v. State of Mysore, A. 1. R. 1956 Mys.

20, the state had designated as backward “all communities other than Brahmins who are not adequately represented in the services.” This was upheld as within Art. 16 (4), the court declining to scrutinize the adequacy of this classification. 119. I BCC 45-47. 120. Thus the Kerala Backward Classes Reservation Commission regards representation in government service as “‘a sign of ability to stand on their own, a sign of non-backwardness of that class of citizens” (1971: 75). The Commission carefully distinguished its use of representation in services as an indicator of backwardness from

its use in assessing “adequate representation” in the services (1971 : 76). 121. Government of Karnataka 1975: I, 315.

122. Thus in State of Jammu and Kashmir v. Jagar Nath, A. 1. R. 1959J. & K. 14, the Government Order, presumably based on a finding that Muslims were “under-

represented,” was not sufficient to satisfy the court that they thereby comprised a

backward class. 123. If the State: may apportion reservations to specific recipients (see chap. 13, §C), these might be quotas much like those condemned in the Venkataramana case, with the difference that there each group was assured a certain number of seats, but now, since

reservations are minimums but not ceilings (see chap. 13, §B) it would be possible for

members of the favored groups to compete for and obtain additional seats.

246

Identifying the Beneficiaries

who are backward by the Balaji tests and who lack adequate representation in the services.124 The justification for accepting under-representation as itself a conclusive criterion of backwardness is the notion that members of any communal group are entitled by right to a portion of posts commensurate with their ratio to the total population. The Constitution repudiates this notion by making it a Fundamental Right of every citizen not to have another preferred on the ground of his caste or religion.125 The only exception to this Fundamental Right is in favor of members of backward classes. However broad the discretion enjoyed by the State todesignate backward classes, it does not include the power to deem as backward any group which may be “under-represented,” for this would turn the exception into a general principle. E. ABSOLUTE VS. RELATIVE BACKWARDNESS The main contribution of the Balaji opinion to the problem of designating the Backward Classes lies, not in the provision of clear and applicable standards for testing backwardness, but in its severe limitations on the range of groups that it will allow to go by that name. In Ramakrishna Singh v. State of Mysore, the state’s scheme was invalidated on the ground that some of the communities designated as “backward” were not “really backward” but indeed were “relatively forward.”126

In the Balaji case the Supreme Court, visibly perturbed at the large portion of the population included among the beneficiaries, clearly indicates that those who may be deemed backward under Article 15(4) area restricted group. (T]he concept of backwardness in Article 15(4) is not intended to be relative in the sense that any classes who are backward in relation to the most advanced classes of the society should be included. . . .127

Indeed, the Backward Classes intended to be covered by that Article are those who “‘are in the matter of their backwardness comparable to the Scheduled Castes and Scheduled Tribes.”’228 And the Constitution contemplates that such backward groups will be “treated as being similar to the Scheduled Castes and Scheduled Tribes.”’129

In its handling of educational backwardness, the Supreme Court

124. A. I. R. 1967S. C. at 1285. 125. Earlier proposals for reserved seats and posts for religious minorities were eliminated from the final version of the Constitution (IX CAD 701). 126. A. I. R. 1960 Mys. at 349 ff. 127. A. I. R. 1963S. C. at 658. 128. Id., at 658. 129. Id., at 658.

The Measure of Backwardness

247

gives an indication of which groups might qualify as these “really backward” classes. The Mysore scheme included as educationally backward all communities whose enrolment in secondary schools through-

out the state was less than the state average of 6.9 high school students per thousand of population. All communities whose enrolment was below 6.9 per thousand were deemed backward; those whose enrolment was less than 50% of 6.9 per thousand were deemed “more backward.” Assuming this to be a rational test of educational backwardness, the Court asks Whether it was legitimate to treat castes or communities which are just below the state average as educationlly backward classes? If the state average is 6.9 per thousand, a community which . . . is just below the .. . [average] cannot be regarded as backward. It is only communities which are well below the state average that can properly be regarded as educationally backward classes of citizens.3°

What is “well below?” The Court goes on to concede that classes of citizens whose average of student population is below 50% of the state averages (i.c., less than 3.45 per thousand) are obviously Backward Classes of citizens. Muslims, whose student population is 5.0 per thousand, “are not so below the state average that the community could

be treated as educationally backward in the State... .” 151 If the test of student enrolment is used, “the legitimate view . . . would be that the classes of citizens whose average is well or substantially below the State average can be treated as educationally backward.”32 While it avoids prescribing any hard and fast rule and says that drawing the line is up to

the State, the Supreme Court clearly indicates that it will accept as “backward” only those whose resources and accomplishments are

considerably below the local average.

The Balaji Court does not indicate how its two tests of “absolute” backwardness— “‘well below” state averages and “comparability to Scheduled Castes and Scheduled Tribes” —are to be combined. Nor

has there been any further development of either standard by subsequent courts. While the “well-below” test has not recurred, the test of backwardness comparable to that of the Scheduled Castes and Tribes has exercised an attraction on some courts.133 130. Jd., at 660 (emphasis added) 131. Id.. at 661. _ 132. Id., at 661. 133. This theme of resolution of the Backward Classes into the Scheduled Castes

surfaces in Chitralekha v. State of Mysore, A. 1. R. 1964S. C. 1823, where Subba Rao, J., after animadversions against the use of caste units says, “If . . . [an] entire sub-caste, by

is backward, it may be included in the Scheduled Castes by following the and large, appropriate procedures laid down by the Constitution” (id., at 1834).

248

Identifying the Beneficiaries

The heaviest use of this standard has been in Andhra. In Sukhdev v. Government of Andhra Pradesh the court in striking down the Backward Classes list noted that the Backward Classes for whom special provisions are contemplated are those comparable in their backwardness to the Scheduled Castes and Scheduled Tribes.45 In Sagar v. State of Andhra Pradesh, where the principle ground for striking down the revised list was the State’s exclusive reliance on caste in constructing the list, the High

Court explicitly applied this criterion. The court noted that the criterion

of backwardness set forth in Balaji, that the Other Backward Classes be

similar to the Scheduled Castes and Scheduled Tribes, was not applied in the preparation of the list in question. “There is no material from which we

could determine that the classes shown in the list are similar to the Scheduled Castes and Scheduled Tribes.”137 The Supreme Court, upholding the judgment, repeated and presumably endorsed this admonition.13® Nowhere do these courts indicate exactly what kind of comparability they have in mind—whether of status, of disabilities suffered, of economic or educational condition, or of representation in

government services: Particularly since the assertion of this standard is coupled with strictures against selection based solely on caste, it is difficult to see how such comparability could be shown. In response, the Andhra Pradesh Backward Classes Commission formulated a rather ingenious reading of the “comparable to Scheduled Castes and Tribes” standard. They found the communities comprising the OBCs to be those which are hindered at present by the accumulated results of a heritage of past disablement which, like that of the Scheduled Castes, “thas hindered social progress in the past by denial of access to educational institutions [etc.] ... because of traditional

occupation

... and

...

left vestiges which

... constitute ... a

serious impediment to social life.” 139 In the next round of Andhra litigation, the Supreme Court relaxed its comparability requirement without giving it further specificity. Backward Classes must be comparable to Scheduled Castes and Tribes, but if they are “really . . . backward” they need not be “exactly similar in

all respects.” "4° In_Janki Prasad a Constitution Bench asserted that this “comparability” was a “fundamental requirement .” The Court sug-

134. CE. the dissenting judgment of Gopalan Nambiar, J., in Hariharan Pillai v. Stateof

Kerala, A. 1. R. 1968 Ker. 42, which found troubling the fact that Balaji’s suggestions had not been adverted to in making the reservations (id., at 55).

135. 1966(1) Andh. W.R. 294. The Court cites the provisions of Art. 340(1) as

evidence for the “comparability” test.

136. A. I. R. 1968 A.P. 165. 137. Id., at 187. 138. A. I. R. 1968 S. C. 1379 at 1383. 139. Andhra Pradesh Backward Classes Commission 1970: 51. 140. State of A.P. 0. Balaram, A. 1. R. 1972S. C. 1375 at 1395.

The Measure of Backwardness

249

gests that the groups may be equated in terms of failure to develop

“enthusiasm for education.” That is, it is those “sectors of the popula-

tion which show extreme apathy toward education due to age-old customs and habits of living, fostered by poverty, ignorance, supersti-

tion and prolonged social suppression” that need to be “goaded into the

social stream by positive efforts of the State.” The Scheduled Castes and Tribes exemplify such social and educational backwardness, and their backwardness in these respects provides the model for the selecting of Other Backward Classes. ‘41 Subsequently, the Supreme Court reiterated the comparability formula in the course of upholding Kerala’s income cut-off provisions, but there was no further development of specific bases of comparison.142 So far it is difficult to see that this doctrine has made any difference in outcome, but it clearly provides the Court with doctrinal potential for resisting expansive listing of Backward Classes. It also provides a basis for eventual merger of the Backward Classes and Scheduled Caste categories and undermines attempts to achieve a radical separation

between them. 143

F. THE SUFFICIENCY OF DATA

In Kesava Iyengar v. State of Mysore,** the court declined to examine

the adequacy of the classification, finding that “{d]oubtless the State is

the sole authority to classify the communities as ‘backward classes.’ ”145 The state “list” was actually a Government Order passed in 1921—34 years before—and the fact that the Government had not seen fit to revise it was accepted as sufficient evidence of its current accuracy. With the Ramakrishna Singh case, however, the Mysore Court became much stricter in scrutinizing the data used to make the determination of backwardness.’ It found the state’s population-and 141. Janki Prasad 0. State of J.& K. , A. 1. R. 1973 S. C. 930 at 938. The Court is not

Suggesting that there is a definite line between the backward groups and others. Its discussion of comparability follows immediately after this observation: “The words ‘advanced’ and ‘backward’ are only relative terms— there being several layers or strata

of classes hovering between ‘advanced’ and ‘backward,’ and the difficult task is which

class can be recognized out of these several layers as being socially and educationally backward.”

142. Jayasree v. State of Kerala, A. 1. R. 1976 S. C. 2381 at 2383.

143. E.g., Justice Krishna Iyer’s attempt in State of Kerala v. N. M. Thomas to separate Scheduled Castes and Tribes as a “‘superclassification” for which treatment quite distinct from Other Backward Classes is warranted. See discussion in chap. 11, §D, below. 144. A. I. R. 1956 Mys. 20. 145. Id., at 28. 146. A. I. R. 1960 Mys. 338.

250

Identifying the Beneficiaries

literacy figures, taken from the 1941 census, defective because “considerable changes have taken place . . . [so that] the census report of 1941

cannot be any basis for the determination of Backward Classes in

1958” 447 However, in the Partha case, the Mysore Court relented to

the extent of finding reasonable the projections of figures on caste population and literacy from older censuses—from the 1941 census where _ figures were available, but in some cases the 1931, 1911, and even 1901

figures. The court found this not to be too unreasonable.1* Balaji envisioned much higher standards in regard to the data upon

which determinations of backwardness were to be made. The Court

observes that

evolving proper criteria for determining which classes are socially backward

is obviously a very difficult task; it will need an elaborate investigation and collection of data and examining the said data in a rational and scientific way. That is the function of the State which purports to act under Article 15(4).1499

Several high courts embraced and applied this high standard. In Sukhdev v. Government of Andhra Pradesh,*®° a single judge struck down a scheme of reservations in medical college admissions on grounds, inter alia, of the absence of evidence that the government had acted on concrete and reliable data or had held any inquiry to ascertain the social

and educational backwardness of the listed groups. Dismayed by the

fact that the list in question was 25 years old and that there were not

even state averages against which to measure the listed groups, the court found this procedure a violation of Article 15.151 Subsequently, in Sagar v. State of Andhra Pradesh, the same judge (now Chief Justice), sitting on a Division Bench, had an opportunity to refine and extend these standards.152 After the earlier list was struck down, the state

switched to an income test while a Cabinet Sub-Committee was established to determine state policy regarding preferences for Backward

Classes. The Cabinet Sub-Committee, after soliciting information on criteria from various states, directed that the Director of Social Welfare

should check the lists of Backward Classes to select

those castes or communities which, apart from the point of view of caste are considered backward on account of the low standard of living, education,

poverty, place of habitation, inferiority of occupation followed, etc. 453 147. 148, 149. 150. 152.

Id., at 347. A. I. R. 1961 Mys. 220 at 232. A. I. R. 1963 S. C. 649 at 659. 1966(1) Andh. W. R. 294. A. 1. R. 1968 A.P. 165.

151. dd., at 304, 305. 153. Id., at 185.

The Measure of Backwardness

251

The Director accordingly drew upa list of 113 communities, which was

submitted to the Cabinet Sub-Committee, which in turn recommended

them to the cabinet, which deleted certain items from the list, overruled the priorities suggested by the Director of Social Welfare, and promulgated the list. The High Court expressed dissatisfaction with this procedure. It found no indication that materials had been placed before the cabinet

from which it could be concluded that the Balaji criteria had been used

in preparing the list.

[We are not able to ascertain whether any material, and ifso what material, was placed before the Cabinet Sub-Committee, upon which the list of Backward Classes was drawn. . . . [I]t was stated below that the Law Secretary and the Director of Social Welfare sat down together and drew up a list, the former specifying the legal requirements and the latter as an expert advising on the social and educational backwardness of class or classes.154

The Director of Social Welfare, who disclaimed personal knowledge that the 113 communities were socially and educationally backward, deposed that he arrived at his assessment by his direct knowledge and discussion with officials, reinforced by consultation of standard

works.455

To accept the contentions . . . that we should rely on the expert knowledge of the Director of Social Welfare would be tantamount to saying that we have to be satisfied with the personal knowledge of an officer of the government who himself is frank enough to admit that the classification is a difficult one and that he cannot say that he has personal knowledge of the life conditions and needs of these classes and that his knowledge was acquired as a result of his considerable touring of many districts . . . and information derived from books, like Thurston and Siraj-ul-Hasan, which information

. . is available to everyone.456

The Court pointed out that these authors wrote nearly 60 years before, and that they had not ascertained the educational backwardness of these castes, nor would their books disclose what were the criteria

adopted in determining social and educational backwardness. Indeed, 154. Id.,at 186.

155. A measure of the accuracy of these methods is provided by the work of the Commission, which recommended a new OBC list for Andhra Pradesh in 1970. It found that of the 32 groups listed by the Director as “Aboriginal Tribes,” 15 were synonyms, titles, mistaken renderings of, or subgroups within, groups already on the Scheduled

Tribes or Scheduled Castes lists; 3 were unlocatable; 2 were groups that were well-off.

This left 12 groups out of 32 to be included in the new list of OBCs (Andhra Pradesh Backward Classes Commission 1970: 60 ff.). 156. A. I. R. 1968 A.P. 165 at 187.

252

Identifying the Beneficiaries

according to some of these books, the listed castes are relatively forward.157 Neither the Director of Social Welfare nor the Cabinet Sub-Committee nor

the Cabinet had before them the population of each of these classes, their

economic conditions, percentage of literacy or education, inferiority of their occupations, their habitation, or their social and economic status vis-a-vis

the Scheduled Castes and Scheduled Tribes . . . . Phe fact that the informa-

tion cannot be easily obtained from the Census reports can afford no valid

ground of justification for sustaining the list of backward classes under

Article 15(4). This information could be obtained, if the state required it,

and it is not for us to suggest the manner in which it can be done, a duty which it should perform however inconvenient or difficult it may be. . . .158 Before the Sagar case reached it, the Supreme Court had encountered

the data problem in two Article 32 petitions (in which it sat as a court of first instance). In Triloki Nath Teku v. State of Jammu and Kashmir (1),*9 it

found among other defects of the scheme that there was no material on

the relevant facts: the population and state of backwardness, social,

economic, or cultural, of the groups listed.16 “What is placed before

us is a general assertion, unsupported by any acceptable data, that all Muslims . . . and the majority of the Hindus . . . are backward.”161 Census figures from 1941 had been placed before the Court. But the

Court cautioned that “the 1941 figures may not afford any workable

guide, as a quarter ofa century has passed by since then and there must have been revolutionary changes during the period.”162 Rather than invalidate the scheme in default of this information, the Court adopted the unusual course of directing the High Court to gather the necessary material and report back to the Supreme Court. In P. Rajendran v. State of Madras, the Court was quite receptive to Madras’s claim that the main criterion upon which its list of castes was chosen was “the social and educational backwardness of the caste based on occupations.”63 There is nothing in the judgment from which it might be gathered that the Court was exacting in its demands for data to

substantiate this; indeed, it seems not to have probed to see whether it was actual or traditional occupation that the state was employing.

When it encountered Sagar, the Supreme Court put considerable stress on the absence of evidence of the state’s use of criteria other than 157. Id., at 186.

158. Id., at 187.

159. A. I. R. 1967S. C. 1283.

160. Earlier, the Jammu and Kashmir High Court struck down medical college

reservations on the ground that the State had provided no data on either educational or

social backwardness (Lalita Shuri Tiku v. State of J. & K., A. 1. R. 1966 J. &. K. 101). 161. Id., at 1284. 162. Id., at 1284. 1603. A. I. R. 1968S, C. 1012.

The Measure of Backwardness

253

caste and appeared to endorse the High Court’s standards on data.1

“Being in the nature of an exception [i.e., reservations under Article

15(4)] the conditions which justify departures must be strictly shown to exist.”165 The mere assertion by the state that its officers had taken into consideration relevant criteria was not sufficient. Whether such criteria had been employed was up to the courts, who would not assume that constitutional criteria had been employed. The High Court’s decision striking down the state list was affirmed; the state had not

made a Suitable investigation nor collected the material necessary to

ascertain the social and educational backwardness of the listed groups. Two weeks after the Supreme Court decision in Sagar, the Andhra Pradesh Government appointed a nine-member Backward Classes Commission, chaired by the retired Chief Justice of the High Court, including six legislators and an I.A.S. officer as secretary. Unlike its discredited predecessor, this Commission worked openly, sending out questionnaires, touring the districts, recording evidence. Its attempt to apply measures of school enrolment and other quantitative standards was severely “handicapped for want of up-to-date statistical information in regard to population of the several communities in this State.”’166 The Commission was forced to rely on projections from old census data: “[t]he population figures for each caste for 1968 have been

arrived at by inflating the 1921/1931 census figures by the percentage of the increase of the total population.”167 Data on student enrolments were based on projections from returns by about half of the schools. Occupation and income data were derived from the testimony of informants, either at hearings or on the Committee’s questionnaires. Although, unlike the Kerala Commissions, this one does not set out. its

statistical findings, but merely reports its conclusions, it represents an advance in precision and system over the earlier Andhra efforts. After two years of work, the Commission recommended a list of 92 communities, estimated to constitute over 38% of the State’s population, to be the Backward Classes. It recommended an array of programs to benefit them, including reservations of 30% of places in government service and in admissions to professional education, these reservations to be reviewed after ten years. The state government accepted the Commission’s recommended list of Backward Classes and reserved 52% of places for them. When the validity of the new rules was challenged, the High Court once again struck down the Government Order on the ground that no proper investigation and collection of data 164. A. 1. R. 1968 S.C. 1379. 165. Id,, at 1384. 166. Andhra Pradesh Backward Classes Commission 1970: 12. 167. Id., at 13.

254

Identifying the Beneficiaries

had taken place. On appeal, the Supreme Court thought that this time the High Court had gone astray in reading Balaji to forbid the use of castes as units in designating Backward Classes. 168 The petitioners who challenged the state’s Order had not specifically attacked the inclusion of any of the groups on the list, and the Supreme Court was indisposed to accept an attack on the overall process of the

Commission. The Court was sanguine about the inevitable crude-

ness of any such investigation. Thus the use of 1921 and 1931 census data where it was the best available was not fatal.17° Nor was the use of “the personal impression gathered by members of the commission

. . .

to augment the various other materials gathered as a result of detailed investigation” in any way improper. Indeed, the Commission was commended for its efforts to find out about actual living conditions.171 The Commission’s efforts were not to be judged by inappropriate standards:

In spite of best efforts that any Commission may make in collecting materials and data, its conclusions cannot be always scientifically accurate in such matters ....[T]he proper approach . . should be to see whether the relevant data and materials referred to in the report of the Commission

justify its conclusions.

.. . [T]here was sufficient material to enable the

Commission to be satisfied that the persons included in the list are really

socially and educationally backward.172

Thus in the setting. of the conscientious performance of the Commission, the Court relented on the sterner pronouncements about data inspired by its earlier encounters with more political commissions. The course of developments in Kerala illustrate a different judicial approach to the data problem. In Kerala, immediately after Balaji, a single judge struck down the scheme of reservations in professional colleges.173 Among the Court’s reasons for rejecting findings of backwardness was that no investigation had been made or data collected.

“[N]o conclusion can be arrived at in 1963 regarding the educational

backwardness .. . of any group on the basis of figures collected in 1941.”174 The data must refer to circumstances existing “within a reasonable time of the passing of the order declaring a particular group as a backward class.”?75 The Division Bench, reversing the striking down of reservations for several groups, found the paucity of up-to-date data a source of considerable worry.176 For the short run they were 168. State of Andhra Pradesh v. Balaram, A. 1. R. 1972 S. C. 1375. The High Court decision was not reported. 169. Id., at 1396. 170. Id., at 1399. 171. Id., at 1399. 172. Id., at 1396. 173. Jacob Mathew v. State of Kerala, A. 1. R. 1964 Ker. 39. 174. Id., at 60, 175. Id., at 61. 176. State of Kerala v. Jacob Mathew, 1. L. R. 1964 (2) Ker. 53.

The Measure of Backwardness

255

willing to rely on a kind of personalized judicial notice.17” “It is impossible to say that our conclusion had not been influenced to some extent at any rate by our experience of life and work in this state.”?78 But no enduring conclusion could be based upon outdated data or “on judicial experience which such data may disprove or modify.”279 It was essential that the state embark immediately upon a fact-finding inquiry. When the case was appealed to the Supreme Court, the state agreed to do just this, and the appeal was withdrawn after argu-

ment. 180

The resulting Commission,"1 chaired by a retired High Court Judge, assembled and analyzed in a thorough and workmanlike way a variety of data: mainly on educational accomplishments and the existence of disabilities, but with some reference to occupation, income, and

landholding. Although the Commission relied upon projections from old census data for its population figures, it not only collected data from vernment

and

educational

institutions,

but

commissioned

the

jureau of Economic and Statistics to conduct a sample survey of certain aspects of socio-economic conditions. In addition, it collected an enormous number of submissions by individuals and groups. Recommending the continuation of reservations in educational institutions on a diminished scale for ten years, it proposed retention of communal units with the addition ofan income ceiling. On the basis of these recommendations, Kerala revised its lists for educational reservations, but continued to use the old ones for reserva-

tions in government service. When the service reservations came before a full bench in Hariharan Pillai v. State of Kerala,1*? the dissenting judge, responding to numerous defects of the scheme, found the data meagre and out of date. He found it “impossible to say that the assessment of backwardness had not been either solely or predominantly on considerations of caste or community, or that [it] has proceeded [along the

lines indicated in Balaji].”1®3 However, the majority, while equally

troubled by the paucity of current data, opted for the strategy of stimulating an inquiry that had worked se well in the admissions area. The Court was particularly concerned lest there were non-backward sections within the backward communities and that investigation be made to insure against this and to insure that the groups dénominated as backward were indeed “by and large backward, socially and econo-

177. Cf. judicial notice of'a group’s notorious backwardness in Chait Ram v. Sikander, A. I. R. 1968 Pat. 337. 178. I. L. R. 1964(2) Ker. 53 at 61. 179. Id., at 61. 180, Interview with Shri P. Govinda Menon, M.P., New Delhi, Dec. 15, 1965.

181. Government of Kerala 1965. 182. A. I. R. 1968 Ker. 42.

183. Id., at 56.

256

Identifying the Beneficiaries

mically.”1& The Court was unwilling, on the basis of the materials

before it, to hold that this was not the case, but found it “necessary to

strike a serious note of warning.”185 The data (a 1935 Committee, the 1941 Census, a 1957 Committee) were out of date. “‘It is essential that

relevant data be collected periodically.” 28 Action under Articles 15(4)

and 16(4) must be ‘“‘modulated from time to time.” “This can only be done if surveys are made at regular intervals and detailed information collected.” 187 It was desirable that the state undertake a detailed

survey as early as possible. “There will be no justification in continuing to apply the principles embodied in Rules 14.and 17 . . . after 31-3-1968 without a fresh appraisal of the questions involved.”’188

Less than nine months later, the Kerala Government established the Backward Classes Reservation Commission. Its six members included

legislators, two advocates, and (a revealing innovation) the Director of the Bureau of Economics and Statistics as its member-secretary. Unlike

all the other commissions to this date, the Nettur Commission aban-

doned reliance on extrapolations from old census figures and ascertained group population by a sample survey.1® It also collected some data on household income as well as educational and employment data from government and schools. By an extremely circuitous route, the Commission recommends that 38% of posts be apportioned among nine Backward Classes. (These are nine of twelve groupings into which all the communities in the state are sorted; the Backward Classes are

those members of each grouping with annual family income of less than Rs. 8,000. These nine, having obtained fewer posts on merit than their portion in the population, are deemed inadequately represented.)19° The Commission’s recommendations, “accepted” by the government in 1973 but later rejected by a different ministry, had not been

put into effect as of 1979.91 The Karnataka Backward Classes Commission was established in 1972 in response to persistent criticism of the “income-cumoccupation” test that prevailed there from 1963, under the chairmanship of its most redoubtable critic, L.G. Havanur.19? Setting out forthrightly to work on the basis of community units, the Commission compiled an immense fund of data on the various communities of the 184. Id., at 51.

186. Id., at 51.

185. Id., at 51.

187. Id., at 51.

188. Id., at 51. The Hariharan judgment was delivered on Jan. 31, 1967, 14 months

before the date mentioned. The Commission was established on Oct. 17, 1967.

189. Kerala (Backward Classes Reservation Commission) 1971: I, 87; II, 462.

190. Id., at 1, 60 ff.

191. See “The Politics of Reservation” in The Hindu, 29 Jul. 1977. Namboodiripad

1976: 336. 192. See the chronology in Government of Karnataka 1975: I, 7.

The Measure of Backwardness

257

state. Unfortunately, it reverted to the use of projections to calculate the populations of these communities.193 But in spite of lack of cooperation, the Commission carried out a massive survey. Two hundred villages, at least one from each taluk, and 204 blocks or wards from

urban areas were selected. Of these, 193 villages and 185 urban blocks were actually surveyed. Every individual in the selected villages and urban blocks/wards was covered under the said survey. . . . About 63,650 families with a population of about 3.55,000 have been covered. 378 investigators collected the information for more than six months under the supervision of 20 District Social

Welfare Officers and 19 District Statistical Officers. In addition, the Deputy

Secretary of our Commission, who is also a Joint Director of Bureau [sic] of Economics and Statistics, supervised the survey work. Training was given to the Investigators in the respective headquarters and two meetings of all

the District Social Welfare Officers were held in the office of the Commission for giving them necessary instructions. Out of the 404 villages and town/city

blocks the Commission received the survey material from 193 villages and 185 urban blocks. In addition to the office staff, the Commission appointed

100 Tabulators. The complexity and magnitude of the work was such that

the entire staff had to work beyond office hours for more than one-and-half

years, 195

Those surveyed were 1.18% of the entire population of the state.19¢ Household interviews elicited information on personal characteristics, education, employment, household particulars, land holdings, ownership of houses, living conditions, health conditions, livestock, agricultural implements, vehicles, radios and news-papers, mem-

bership in co-operatives and political organisations, cottage industries, income and expenditure and traditional occupation.” Information was also collected at the village, ward, and taluk levels about panchayats, cooperatives, money lending, educational institutions, forced labor, and a host of other matters.19*

Notwithstanding the thoroughness and rigor of its work, the Com-

mission’s recommendations were the focus of heated controversy. As in Kerala, it was the exclusions as well as the inclusions that delayed 193. Government of Kamataka 1975: II, 1-2. Cf. the note on estimated population

by the Andhra Pradesh Backward Classes Commission 1970:156.

194. The Commission notes that it could not get the Directorate of Social Welfare, the

Bureau of Economics and Statistics, or the Census Department to conduct the survey (I,

308). But the Commission did secure additional funding to enlarge its data collection activities. 195. Id., at 1, 308. 197. Id., at 1, 309.

196. {d., at I, 310. 198. Id., at I, 309-10.

258

Identifying the Beneficiaries

acceptance. When the Government did implement the Commission’s recommendations, it added to the 185 communities recommended by the Commission some 20 groups that the committee had omitted, including the relatively prosperous and politically potent Lingayat community.1%? The measure and moderation of the Commission was also abandoned in the portion of places reserved: where the Commission recommended 32%, the state government reserved 50% (thus return-

ing to the pre-Balajt level).

A Backward Classes commission in Gujarat, sitting contemporaneously with the Havanur Commission in Karnataka, marked a new high pdint in the utilization of social science expertise.2© The five-member commission included two eminent social scientists, and other highly qualified social scientists were utilized in staff and advisory positions. In addition to industriously assembling data by hearings and inspections, the Commission set out to conduct a comprehensive two-stage sample survey of households throughout the state. The first phase—a statewide socio-economic survey of 6,449 households— was conducted,

but the second phase—a purposive sample to augment the frequency of

observations of groups not adequately included in the first phase— was

unfortunately abandoned owing to time pressure. The Karnataka and Gujarat reports mark an advance in the scope of

data collected and the rigor with which it is analyzed. They suggest

possibilities for development of systematic surveys as a tool to identify ckward Classes and to monitor programs for the amelioration of their conditions. Government need not employ such Commissions when it designates Backward Classes. Of course, the evidentiary standards applied to commission data are equally applicable when the Government makes the determinations without benefit of such an inquiry. “[T]he State

should produce material before the Court to show that there was a proper inquiry, relevant criteria were adopted and the decision on [who

are backward] is reasonable. Mere expression of opinion is not sufficient.”201 Thus a State scheme was struck down where “there was no application of mind” in determining the social backwardness ofa group

and the criteria for educational backwardness were faulty.202

199. Government of Kamataka: G.O. No. DPAR-1 SBC, 77, dated March 4, 1977. 200. Government of Gujarat 1976.

201. Dilip Kumar v. Government of U.P., A. 1. R. 1973 All. 592 at 595. These requirements may be applied with leniency. Thus a state may act on incomplete data—it need

not wait until all the data is in before acting (Viswanath v. Government of Mysore, A.1LR. 1964 Mys. 132 at 136). Nor need all the supporting data be included in the order providing

for reservation. Justification to satisfy the court can be provided later and separately

(Jacob Mathew v. State of Kerala, A. 1. R. 1964 Ker. 39 at 51). 202. A. I. R. 1973 All. 592 at 595.

The Measure of Backwardness

259

Alll of the recent commissions recommend that reservations be made for a fixed period and then reviewed. In Periakarrupan the Supreme Court made it clear that peripdic re-assessment was obligatory: . Government should not proceed on the basis that once a class is considered a backward class it should continue to be backward class for all time. . .. The Government should always keep under review the question of reservation . .only the classes which are really socially and educationally

backward should be allowed to have the benefit of reservation.2°

Candidates on the existing list of Backward Classes had secured about half the seats in the general pool. This showed that “the time has come for a de novo comprehensive examination of the question.” 2 Thus the courts

have moved

from a position of unquestioning acceptance of

government action to one of close, if intermittent and uneven, probing to insure that the government has acted on data sufficient to support its claim to have used acceptable criteria. Some courts have applied such standards unrelentingly to strike down an order, others have invoked them to prod the state into upgrading its performance. In so doing, the courts have helped institutionalize periodic review and pushed the states beyond primitive data collection and simple assertion into systematic generation and analysis of data. The generation and analysis of data suggests new possibilities for designing programs for the Backward Classes and monitoring their implementation. And it foreshadows a new generation of problems for judicial control. The data in surveys varies in quality and in the inferences that may be drawn with confi-

dence from it. Problems of sampling error, coding reliability, interven-

ing variables, and such threaten the inferences that may be drawn. As the designation of Backward Classes involves the collection and interpretation of survey data, courts will have to devise some minimum standards for the handling of empirical data. The first target of such standards might well be the practice of extrapolating population figures from old census data. The preIndependence enumerations of population by caste were open to some suspicion about their accuracy— perhaps most in the case of groups of low status.205 And there is reason to think that different communities

have experienced different rates of fertility, mortality, and migration in

the intervening years.2

As time goes by, the speculativeness of such

203. A. I. R. 1971 S.C. 2303 at 2314. 204. Id., at 2311. 205. The Andhra Pradesh Backward Classes Commission (1970: 14) observes that in

pre-preferential treatment days there may have been systematic understatement of the population of communities ‘about whom there was some stigma.” Thus projections would understate their population and overstate, ¢.g., their current educational accomplishments.

206. On differential fertility, see e.g., Mandelbaum 1974: 42-59; Driver 1963: 89; cf.

K. Davis 1951: 179.

260 Identifying the Beneficiaries estimates is compounded and the interpretation of good current data confounded. Continued judicial indulgence of this practice seems out of line with the judicial thrust toward more adequate data in designating the backward. But a substitute is not easy to come by. Are governments to be required to carry out surveys? If there are to be Backward Classes defined in part by caste units, the absence of population figures for these units renders difficult the task of assessing their condition. Central and state governments have been reluctant to generate data broken down along caste lines.2°7 The Census did not collect caste data (except regarding Scheduled Castes) in the post-Independence censuses. The caste enumeration and the census recording of social precedence was perceived as a device of colonial domination, designed to undermine as well as to disprove Indian nationhood. There was revulsion at the unseemly scramble to

use census listings to upgrade caste status and to inflate numbers for political advantage. And abandonment of caste enumeration seemed a step toward the disestablishment of caste. In 1955 the Backward Classes Commission remarked on the shift in emphasis in Census operations from ethnographic to economic. Finding itself adrift without accurate data, the Commission recommended collection of caste data

on the 1961] census.2° But Chairman Kalelkar in his covering letter denounced the recording of caste as a violation of voluntarism and beset by immense practical difficulties.20° That a caste enumeration would

aggravate caste consciousness and undermine modernization has re-

mained the received view. The practical difficulties raised by the caste-blind policies of the Census have bedevilled Backward Classes commissions. The Andhra Pradesh and Karnataka Commissions have urged resumption of caste enumeration.210 For some, governmental unwillingness to collect caste data is a deliberate move to “preserve the status quo of disabilities, distinctions and injustices.”’211 Thus Havanur urges that census data be required to implement the commitment to uplift the weaker sections. He dismisses the disestablishment argument with the observation that government has not found it necessary to cease recording data on the

basis of any of the other suspect criteria— religion, language, sex, etc. The argument that omission of caste enumeration lessens caste con-

sciousness he finds “amusing and absurd as when a merchant omits

labelling grain bags to prevent bandicoots from eating the grains.”?!? 207. Cf. Government of Karnataka

1975: I, 309.

208. 1 BCC II. 209. I BCC xvii ff. 210. Government of Kamataka 1975 : 1, 332 ff, Andhra Pradesh Backward Classes

Commission 1970: 15. 211. Government of Karnataka 1975: I, 336.

212. Id., at 336.

The Measure of Backwardness The

261

controversy about collection of caste data echoes the larger

conflict over the use of communal or of neutral units in designating

Backward Classes (discussed in chapter 7). Avoidance of communal units obviates the need for caste data; to measure the backwardness of

communities requires data on caste lines. Since the courts have accepted the use of caste units, their insistence on data of higher quality implies some responsibility to assemble appropriate data to justify their designations of backwardness and the desirability of data to assess the results of programs. Such data might be produced by regular decennial census, by special census operations, or by surveys conducted by other

agencies. Not only have courts insisted on fresh and detailed investigation as a prerequisite to preferential treatment,?%3 but also the collection of “necessary materials” under judicial auspices has been directed by the Supreme Court.2" The possibilities of judicial action to require assembly of an appropriate data base have not been pursued, although

apparently such possibilities are present.

G. ECONOMIC TESTS Economic tests of backwardness have been highly praised by critics of community-based preferential treatment. Income and occupation tests of various kinds are used in many schemes for Other Backward Castes and

to a lesser extent in schemes for Scheduled Castes and

Tribes. It is important to distinguish three different ways in which such

economic tests can be employed ini designating the Backward. First, income and occupation may be used as criteria for selecting

backward communities. Second, income and occupation tests may be employed in conjunction with community membership to determine individual eligibility for preferential treatment. Third, income and occupation may be used independently as criteria for selecting individuals who compose a Backward Class. 1. Income and Occupation as Criteria for Selecting Backtuard Communities The thrust of the Balaji case was that backward communities should be selected by criteria other than merely caste standing — especially po-

213. Hariharan Pillai v. State of Kerala, A. 1. R. 1968 Ker. 42 at 51. 214. Triloki Nath Tika v. State of Jammu and Kashmir [1], A. 1. R. 1967 S.C. 1283 at 1286.

The Court directed the High Court, either directly or through the district court, to

gather material on population, community strengths in the services, the extent of backwardness of the various communities, etc.

262 Identifying the Beneficiaries verty, occupation, and residence.745 In P. Rajendran v. State of Madras the Supreme Court upheld the list where the state claimed that occupation was the principal ground upon which the communities had been selected.216 The Court apparently did not ascertain whether it was actual or traditional occupation that the state had employed.?!7 Apparently, it was the latter 2%—a standard which seems not very different in its result than the “‘caste test” that the Supreme Court found objectionable in Balaji.249 The courts have never formulated any standards for the application of income or occupational tests to communities. It is not clear whether such limitations as “well below the State average” and “comparable to Scheduled Castes and Tribes,” which were formulated in respect to educational backwardness, would

apply to economic factors as well. The absence of castewise census data would make it exceedingly difficult to apply such tests. Nevertheless, assessments

of the income

and occupational

distribution of com-

215. Shri Havanur, a staunch proponent of the use of communal units, objected to the

Nagan Gowda Committee’s failure to apply these tests. He suggested that a caste

should be considered backward if its per capita income is less than half the state

average, or if more than 50% of its members are engaged in occupations regarded as inferior (1965:110).

216. A. £. R. 1968S. C. 1012. 217. Ifit was actual occupation, this involves an intriguing problem of data. How did

the State ascertain this? Castewise occupational figures had‘ not been collected in censuses since 1941 or earlier.

218. Hardgrave (1969: 144) recounts the vacillation of the Madras Government in

regard to the listing of Nadars as a Backward Class: “The government wanted ‘to make it clear that the concession ... contemplated for the Shanan community is admissible only to persons whose traditional occupation is toddy-tapping.’ This only compounded

the difficulties, as the traditional occupation of all Nadars was toddy-tapping. In 1962 the government clarified its position, saying that the concessions ‘should be given only to the children of those persons who had been actually toddy-tappers until the introduction of total prohibition in their native districts.’ In a memorandum to the chief

minister, the president of the D. M. N. S. [Dakshina Mara Nadar Sangam] opposed

‘actual’ occupation as the criteria [sic] for eligibility to ‘backward classes’ benefits. This is patent injustice,’ he wrote. ‘In the case of no other community is actual occupation of the individual parent made the criterion!’ ” In 1963, Nadars were put on a par with other Backward Classes in this regard. 219. It appears that traditional occupation was the single factor most heavily relied on by the Backward Classes Commission in the determining the backwardness of various communities in a way that was labelled as a caste test. Traditional occupation is not always readily ascertainable, particularly if the groups are large. Cf. the rejection of this standard by the Nagan Gowda Committee on the ground that it was too difficult to define (Government of Mysore 1960: 2-3). If a test of disabilities or low status is employed, traditional occupation may be redundant. Thus the Kumara Pillai Commis-

sion rejected occupation as a separate test on the ground that the incidence of the stigma attaching to certain occupations was on the low income groups of those communities which had traditionally followed those occupations, groups that were already included by the “caste” test (Kerala [Kumara Pillai Commission] 1966: 37 ff.).

The Measure of Backwardness

263

munities play a prominent role in the assessment of backwardness by some of the state Backward Classes commissions. Traditional occupation and current employment in unclean, undignified, and unremunerative occupations bulk large, and commissions have attempted to take income into account in spite of formidable difficulties of ascertaining it. Of course, income and occupation here are used to characterize the group. The individual beneficiaries of programs may not share these characteristics, and indeed in the case of high-echelon benefits they often (but not always) will not.

2. Income and Occupation Tests in Combination with Communal Criteria Economic tests may be combined with communal criteria in another way: the economic test may act as a ceiling, excluding from preferences the most prosperous section of the backward community. Thus, e.g., Kerala adopted a “means-cum-community” test after the Kumara Pillai Commission concluded that the heritage of disabilities did not seriously impinge on the well-off families within the listed communities. Medical-college seats were reserved for members of listed communities

whose annual family income was less than Rs. 6,000.22 This scheme

was challenged from two directions: by an applicant who met the income test but was not a member of one of the listed communities, and

by members of the listed communities with higher incomes. The High

Court rejected the claim of the low-income non-member on the ground that it was open to the state to give caste some relevance along with income— in effect, to apply the income test only to members of certain

castes,221

,

Surprisingly, the second challenge was initially successful. A single judge of the Kerala High Court struck down the income limitation on the ground that the absence of poverty cannot be the sole determinant of the absence of social and educational backwardness.222 (If it is a fallacious inference that since poverty is not a sufficient ground for

inclusion its absence cannot be a sufficient ground for exclusion, then it

is one that stems from a perception that this is income discrimination

among

the backward.) The Court also finds the Rs. 6,000 limit ar-

bitrary: it is improperly low, for it would exclude a family with two wage earners in Class IV government employment, and it is less than would enable a family to support a child in medical college.229 On appeal, a 220. The Kumara Pillai Commission had recommended an income ceiling of Rs. 4,200,

but the state government raised it to Rs. 6,000 when putting the recommendations into effect. 221. Laila Chacko v. State, A.1.R. 1967 Ker. 124. 222. Shameem v. Medical College, A.1. R. 1975 Ker. 131 at 138.

223. Id., at 138.

264

Identifying the Beneficiaries

Division Bench had no difficulty in finding that the more prosperous members of backward communities might be excluded on the ground that they were lacking the requisite social backwardness to entitle them to preferential treatment.?% Since poverty is relevant to the determi-

nation of social and educational backwardness, its absence may be

taken into account. To the arbitrariness point, the court responds that the inevitability of borderline cases does not preclude dividing the

group to exclude those who are not backward, shifting benefits from the

more advanced to less advanced sections of the listed community. The question is whether the Commission has material from which it could

conclude that the better-off among the listed castes were not socially

backward. Having satisfied itself that the Commission kept in mind the guiding legal principles, the court declined to second-guess the Commission’s weighing of the evidence.?25 The following year the matter was taken directly to the Supreme Court in Jayasree v. State of Kerala.?26 In the meantime the state government had raised the cut-off point from Rs. 6,000 to Rs. 10,000 annual

income. Both caste and poverty are relevant in determining the social backwardness of a group. Here the Commission’s finding—that the

wealthy sections of the listed groups were not stigmatized or disabled by

their group membership and were thus not socially backward—was a permissible application of relevant tests. In the course of its judgment, the Court

makes

some

tantalizingly obscure observations about the

interaction of community and poverty. It suggests that not only is the classification of Backward Classes solely on the basis of caste impermis-

sible, but that “classification . . . based solely on poverty . . . will not

be logical.”227 The argument, never made explicit, seems to be that the “social backwardness which results from poverty is likely to be magnified by caste considerations,”228 and thus “special provision is contemplated for classes of citizens and not for individual citizens.”229 Thus the Court seems to move away from the notion propounded by

Balaji that income might be used alone— “neither caste alone nor poverty alone will be the determining tests [of backwardness]”230—at

the same time that it commends Chitralekha’s upholding of Backward Classes based on income combined with occupation.?31 If income and community are both relevant, they may be combined in a different way to produce what might be called a staggered income test. In accordance with a policy of a higher ceiling for members of 224. 225. 226. 227. 229.

State of Kerala v. Krishna Kumari, A.1.R. 1976 Ker. 54 at 60. Id., at 61. A. I. R. 1976 S.C. 2381. Id., at 2385. 228. Id., at 2385. Id., at 2386. 230. Id., at 2386. 231. Id., at 2386.

The Measure of Backwardness

265

communities “socially looked down upon,”23? the Punjab Govern-

ment reserved seats in medical colleges for members

of listed com-

munities with annual family incomes of less than Rs. 1,800 or any other person with annual family income of less than Rs. 1,000. This staggered income test has been upheld.793 Among the advantages ascribed to income cut-offs is that, by exclud-

ing the well-off within the backward communities, they insure that preferences go to the really disadvantaged and deserving within these communities. Just how many are actually excluded depends, of course,

on where the line is drawn and on how tightly the test is administered. Usually the limits have been fixed so high that only a small porton of these communities would be excluded. An extreme example would be Kerala’s limit of Rs. 6,000 (later 10,000) annual income, which prob-

ably eliminates no more than 1% in the populous Ezhuva community

and

even

less in other communities.234 Of course, this 1%

may ac-

count for a far higher proportion of those who receive benefits like medical school admissions—so there may be a real redistributive effect.235 Where the community is poor and the means test high, the number of those excluded may hardly justify the cost of administration of the income test.236 But in so far as such tests assuage the fears of other communities that well-off sections within the backward are the undeserving beneficiaries of preferences, such tests perform an important symbolic function— possibly their most important one.237 Income cut-offs, usually at far lower levels, are a common feature of

many schemes for fee concessions and scholarships and other benefits of a less exalted sort than medical admissions. For example, the largest scheme of all, scholarships for Scheduled Castes, is limited to those SC

232. RCSCST 1958-1959: 1, 12. 233. Gurinder Pal Singh v. State of Punjab, A. 1. R. 1974 P. & H. 125. See Ramesh Chander

Garg v. State of Punjab, A. 1. R. 1966 Punj. 476.

234. The Kumara Pillai Commistee estimated that its original ceiling of Rs.4,200

would exclude about 4% of the Ezhuvas, Muslims, and Latin Catholics (App. VIII). Presumably, the Rs. 6,000 standard would restore the great majority of these.

235. As noted by the Division Bench in State of Kerala v. Krishna Kumani, A. 1. R. 1976

Ker. 54 at 60, 61.

236. Thus the Study Group on the Weaker Sections of the Village Community (1961:

I, 12) dismissed as an idle exercise the application ofa Rs. 1,000 test to Scheduled Castes and Tribes. 237. Preferences in higher echelon positions (medical school admissions, legislative Seats, etc.) are in the nature of things a symbolic payment to groups for whom the government cannot do much in-more tangible currency. It is those within these groups who are already better off who are most likely to serve as symbols of group participation in the achievements and control of the society. A policy of eliminating them by income

ceilings might, if the line were drawn low enough, exclude most of the likely candidates for this symbolic function.

266

Identifying the Beneficiaries

students whose parents’ or guardians’ income is less than Rs. 750 per month.238 The income cut-off in these cases has never been challenged, although it rests on somewhat different premises than the cut-off upheld in the Kerala cases just discussed. There the Commission concluded,

and the courts concurred, that income serves as a measure that social

backwardness associated with membership in the listed group had been

dispelled. It is not self-evident that this conclusion would be sustainable

at the much lower levels typically used for fee concessions, or that it would apply in the case of Scheduled Castes. It may seem curious that competitors for high-echelon benefits are aided at much higher levels of income than those who aspire to avail themselves of elementary school. In part, the explanation is that income cut-offs are really used for two distinct purposes. At low levels, they are used to define priorities: they provide a crude measure of ability to pay—or at least they isolate those with the most pressing and unrefusable need. But where they are used for higher-echelon benefits their purpose is quite different: they are used as an indicator of the presence of family advantages that offset the disablement associated with membership in a particular community.” 3. Income and Occupation as Independent Criteria of Backwardness.

Proponents of economic tests have sought not only more accurate

identification of backward communities and the exclusion of the well-off from benefits, but the total abandonment of the communal element in selection of Backward Classes. There was considerable sentiment for

this in the late 1950s. The Ministry of Home Affairs recommended to all the states an occupational test with an income ceiling. Erstwhile 238. RCSCST 1973-74: 84. 239. Thus the Tamil Nadu Backward Classes Commission (1971: I, 104) recom-

mends a higher income level for eligibility to apply for admissions and reserved posts than for scholarships on the the ground that “parents with less than the income . . . may not be able to give the environmental advantages. . . .” 240. The Ministry’s proposal was that a list of backward occupations be drawn up. This list was to comprise: “(a) any non-agricultural occupation in any state of India in which 50% or more of the persons belong to the Scheduled Castes or the Scheduled Tribes, (b) any non-agricultural occupation in which general literacy percentage of the persons depending thereon is less than 50% of the general literacy in the state” (RCSCST

1958-1959: 1, 12). This was to be used with a family income ceiling of Rs.

1,000 per year (a figure based on the notion that the average family consisted of five members and per capita income in the countryside did not exceed Rs. 200). Presumably, the occupational test was to apply to actual rather than traditional occupation. This proposal was subsequently declared unworkable. See Letter of the Minister of Home Affairs to the Chief Secretaries of all State governments/Union Territaries, 14 August 1961, reprinted at RCSCST 1960-1961: 11, 366.

The Measure of Backwardness

267

Bombay did adopt an income test for some scholarship aid.%! The central government continued to urge economic tests on the states.?4? This sentiment was given a boost by Balaji in an aside on the tamily income test used in Maharashtra. The Court observed that if any state adopts such a test and attempts to enlarge educational opportunities by increasing the number of institutions, it can claim to proceed on a rational, broad and scientific approach which is consistent with, and true to, the noble ideal of a secular welfare democratic

State set up by the Constitution of this country.43

The Court notes that such an approach might, if necessary, be supplemented by “special provision by way of reservation to aid the Backward Classes and Scheduled Castes and Tribes.”2 Although some who welcomed the Balaji endorsement seemed to believe that Maharashtra had abandoned communal tests entirely, the income test had in fact been adopted only for the giving of financial aid in education. It was not employed for distributing seats in medical colleges or posts in government services, which continued to be distributed to the Backward Classes as communally defined. Although they enjoy Supreme Court commendation, such income tests

are

by

no

means

mandatory.%5

Nevertheless,

such

“pure”

economic tests have proved attractive to a number of states and agencies.?46 In 1963 the principal central scheme for Other Backward Classes, post-matriculation scholarships, shifted from using the state lists of backward communities to a test of family income. Scholarships were to be given first to students with family incomes of less than Rs. 1,500 per year; second, if any funds remained, to those with family incomes of Rs. 1,500-2,000 per year (in the case of technical courses,

Rs. 1,500-2,400).247

241. Ironically, Bombay's adoption of the income test may have been inspired less by

an animus against communal tests than by a desire to accommodate the new converts to

Buddhism (interview with Professor R. D. Bhandare, 242. See the letter cited in n. 240. 243. A. I. R. 1963 S. C. 649 at 664.

Bombay, 1966).

244. Id., at 664.

245. As indicated by all the subsequent litigation in which the absence of economic

tests has not troubled the court—e.g., P. Rajendran v. State of Madras, A. 1. R. 1968S. C.

1012. This non-mandatory character of economic tests is even more clear in the case of Scheduled Castes and Scheduled Tribes. In Devadasan v. Union of India, A. 1. R. 1964S.

C. 179 at 188, the Court rejected the argument that the Government was obliged to implement Art. 16(4) by selecting classes by income rather than along communal lines. 246. Particularly in disbursement of benefits in primary and secondary education. See, e.g., RCSCST 1960-61: 1, 148-9 (free schooling to those with annual income of less than Rs. 1,200 in Maharashtra and Gujarat and, up to Form III, in Madras). 247. No. F 5-3/63 Sch. 4, Government of India, Ministry of Education. These

scholarships were to be distributed without a “merit test” and in “the order of

268

Identifying the Beneficiaries

In Andhra, after the High Court struck down the list of backward

communities, benefits were given to Backward Classes defined as those

with incomes below Rs. 1,500.48 West Bengal provided educational concessions for those whose parents’ annual income was below Rs.

3,000.49 Mysore, in reaction to judicial invalidation of its earlier schemes, switched over to an income-cum-occupation test; the Back-

ward Classes were those families who had annual incomes of less than

Rs. 1,200 and pursued the following occupations: (1) Actual cultivator; (2) Artisan; (3) (4)

Petty businessman; Inferior service (ie. Class IV in Government

service and

corresponding class or service in private employment) including casual labour, and

(5) Any other occupation involving manual! labour.25°

In the course of determining the boundaries of these notions, the Mysore Court has indicated that it is “the habitual occupations of families” rather than “casual or temporary occupations’ that are intended.?51 But no court there or elsewhere has addressed the problem of which occupations the State might find to be indicators of backwardness. It was suggested by the Mysore Court that the object of this poverty—i.e., students from families with lower incomes will receive preference over those from families with larger incomes.”

248. Andhra Pradesh Backward Classes Commission 1970: 7. 249. RCSCST

1964-65: 313.

250. Govt. of Mysore 1965:32 ff. The 1963 education order is reprinted in its entirety

in Shetty v. State of Mysore,A. 1. R. 1969 Mys. 48. Havanur (1965: 98-99) eloquently

portrays some of the consequences of the then new Mysore test: “The notification states that a student is backward if his parent’s or guardian’s income is less than Rs. 1,200 per annum and that he is either an agriculturist, petty business-man, artisan or a menial servant. If the student owns lands and buildings fetching income, that does not matter. If the parents have a dozen of children and the income is Rs. 101 per month, all his children are forward. Whereas the father of the only child whose income is Rs. 100 per month the child is backward. A primary school teacher whose pay is Rs. 70 per month with half a dozen children is forward. A peon whose pay is Rs. 90 with a child is

backward. A landlord who owns hundreds of acres of land is backward if the Tahsildar certifies that the income was Rs. 1,200 per year. A widow who gets Rs. 15 by way of rent

from her small house is forward. The backwardness is of an individual, and on the date of the certificate. If the business-man gets profits to any extent after the date of certificate,

the child

is still backward.

In the matter of reservation of posts, the

candidate may be a gazettesl officer whose pay may be Rs. 500 per month, but he is backward when he applies for a higher post or other posts, provided the parent or guardian of the gazetted officer secures a certificate of the type described above from a Tahsildar. The gazetted officer is backward, but the children of the backward gazetted

officer are forward because their father is a gazetted officer.” 251. Id., at 50-51.

The Measure of Backwardness

269

classification is to “help those who are at the bottom rung of the ladder of social and economic development.”252 But it is not suggested that a parent’s being a cultivator or an artisan has in itself stultifying effects on children. Rather, it seems to be assumed that those who engage in these callings occupy positions of low status and command few resources, so that opportunities for personal development are less likely to be present

in their homes. Therefore, it may be assumed that their children will

enjoy a less favorable environment for educational accomplishments. The assumption that the low status of these occupations can be taken as an indicator of the unfavorable developmental environment is the same as that which supplies the rationale for the communal units. Whether occupation or community correlates more highly with affinity of conditions remains to be tested. The Supreme Court’s only discussion of occupational backwardness provides little guidance. In Janki Prasad v. State of Jammu and Kashmir,?53

the Court says it is “quite open to the State to declare that persons

belonging to low income families following a traditional occupation should be regarded as persons belonging to a backward class.”254 But it is not enough that the occupation be traditionally associated with the ily: “his family must be still following the traditional occupation.”255 The Court then proceeds to confine “traditional occupations” to those which “require special skills developed by tradition,”256 thus excluding such menial callings as agricultural laborers, bearers, load carriers, and cart drivers. But since it is open to the State

to define Backward Classes by combining income with actual as well as

traditional occupation, and since these callings would surely be associated with backwardness, it is difficult to see the purpose of interposing this skill requirement. In addition to excluding the well-off among the backward communities, a further and more important virtue is claimed for a purely economic test: that it allows benefits to reach the deserving in all communities, including those members of “advanced” communities who live in depressed circumstances. Thus in Hariharan Pillai v. State of Kerala the Court cautioned the Government that not only was it necessary to eliminate the non-backward sections of the backward communities, it was also necessary to include the backward sections within other communities. The Backward Classes, said the Court, must be

drawn from “‘all the weaker sections of citizens.”257 The application of economic tests involves a number of problems 252. Sridhara v. Director of Technical Education, 1966(2) Mys. L. J. 744 at 777. 253. A. I. R. 1973 S.C. 930. 254. Id., at 940. 256. Id., at 940.

255. Id., at 940. 257. A. 1. R. 1968 Ker. 42 at 51.

270

Identifying the Beneficiaries

which raise some doubts about how efficiently they accomplish their

avowed objectives. Such terms as “manual labor” and “‘family income” give rise to a host of conceptual problems. Joint families, income in kind and services rather than cash, and generally the difficulties of determining agriculturists’ net income— all of these add formidable problems of accurate assessment.?58 And all these difficulties are compounded by a widespread inclination to understate income.?5? If a means test applied with precision might be thought to have unwelcome disincentive

effects, the difficulties have not been in this direction. In the absence of income and tax records, government has relied on income certificates

issued by government officers, who have neither the resources nor the inclination to make accurate assessments.26 In many areas income certificates have been generally available to almost anyone except those

who, like schoolteachers or inferior governments servants, had incomes

that were matters of general knowledge. Of course, the opportunities for obtaining such certificates are not spread evenly: they are greater for those who have contacts with government officials (and their clerks). Although the prosperous have on the whole succeeded in obtaining such certificates, in some instances the truly destitute have experienced considerable difficulty. Economic tests, like communal tests, entail the “‘lion’s share” prob-

lem. Those who get the benefits are those among the poor who are relatively better off and have the resources to take advantage of the opportunities offered them.?61 Whatever criteria are used in selecting Backward Classes, there is the problem of getting benefits to those at the bottom. When caste units are employed, it is the elite within the caste who receive a disproportionate share of the benefits. Income tests in fact broaden the field against which the most backward and deprived members of these communities have to compete: it substitutes for their more prosperous caste fellows a much larger array of poor (or allegedly poor)

families among the more prosperous and powerful groups. It was apparently with such considerations in mind that the Mysore High Court insisted that the state should have supplemented its income and occupational tests with a caste test, ““There is hardly any doubt that the scheme . .

. has not helped the really Backward Classes amongst the

Hindus.”262 The court substantiates this with some details from the 258. One report mentioned as instances of “rural ingenuity” that “income was invariably interpreted as net savings after all the normal family living needs were met,” while employed persons omitted dearness allowances, etc. in computing their income. 259. Cf. the Nagan Gowda Committee's rejection ofan income test on the ground that

“it is difficult to enforce in practice” (Mysore Backward Classes Committee 1960: 3).

260. Problems of issuance of such certificates are discussed in chap. 9,§ G below.

261. These resources might include contacts, high self-esteem, a tradition favorable

to education, or other by-products of membership in a high or prosperous caste. 262. A. I. R. 1964 Mys. 132 at 139.

The Measure of Backwardness

271

1963 selection for Mysore engineering colleges, which tend to suggest

that the more advanced a group was educationally, the more it benefit-

ted from the application of the income-cum-occupation test (table 21). While the test seemed to help the less well-off in the advanced com-

munitites and the middle strata, the communities at the bottom were

almost entirely excluded. Of course, the differences which emerge here may have been produced, not by the operation of the income-cumoccupation test, but by selective processes at earlier stages which limited the numbers of Korubas and Bedars who entered this competition.2® But evidently, distribution of preferences along income/ TABLE

21

RELATIVE Success OF SELECTED Castes ON MERIT AND ON SELECTION BY INcome-Cum-OccupaTion Test IN Mysore ENGINEERING CoLecEs, 1963

S$

Se

as oa

gb ~¢ g 22

£€s

25

#

&£s S°

&B Ee

BRAHMINS

4.28

LINGAYATS

15.57

Group

x

42

$5

t

; i ee %

§ 5

a

225 cB 8 gl

‘ss=

BE

‘g FF

22

s3

g g2

Se

FR | ak

2

§

2 §&

g a3

285

(35.9

8.40:1

22

i

13.9

0.89:1

35

a

3s=

Ss os

3

&§ fe

2

&

15.5

3.6:1

Forward

24.6

16:1

(Borderline)®

12.98

32

40

O31

| 35

46

191

6.88

6

0.75

OWN

2

14

0.2:

4.74

3

0.38

0.08:1

1

0.7

0.15:1

794

a8

s

z

KORUBAS

TOTALS

4833

*8

VOKKALIGAS BEDARS

=

Backward More Backward More Backward

142

Source: Calcylated from data provided in Viswanath v. Government of Mysore, A.1.R. 1964 Mys. 132 at 137, 139.

2 Mysore Backward Classes Committee 1961, Statement 9.

:

> Classified as forward by the Committee, backward by state government. 263. Table 21 includes all the figures that the court provided, so we ‘cannot exclude the possibility that the pattern for these five castes was atypical. Nor do we really know

how many Koruba and Bedar applicants failed because they were (1) excluded from the reserved seats by the income-cum-occupation test, or (2) beaten out for reserved seats

by members of more advanced groups who met the income-cum-gccupation test.

272

Identifying the Beneficiaries

occupational lines did little to offset the effects of earlier inequalities— at least at the bottom of the scale.

The Mysore scheme here, like most economic tests, combined in-

come/occupation with merit. Income/occupation determined eligibility to enter the reserved-seat competition; among the eligibles, the seats were awarded on the basis of merit. Presumably, the tendency of economic tests to exlcude the lowliest could be mitigated by the use ofa “pure” income test—i.e., using income to determine both eligibility

and the order of distribution. This is done, e.g., in the central postmatriculation scholarhsip scheme, where the student with the lowest income gets his scholarship first, and the remainder are distributed in “order of poverty.” This kind of arrangement seems workable where the economic test does not involve incommensurables (like income and

occupation) and where it is assumed that all of the group are equally eligible and deserving. Typically, however, that assumption is not present, and it is felt that merit inter se should play a role in the selection.

In considering the suitability of economic tests, we should distinguish what we might call higher-echelon preferences (which require for their utilization a degree of prior success and some matching resources) from what we might call early-stage preferences, which can be utilized without prior success or matching resources. Reservations in medical college admissions and higher government posts are examples of higherechelon preferences; fee concessions in primary school and free lunches

for schoolchildren are early-stage preferences. Obviously, the distinction is not absolute, but a matter of degree. While economic tests may be suitable for distributing fee concessions or free lunches to schoolchildren, it is not clear that they are as well adapted for distributing higherecholon preferences. Since some degree of prior success is required even to enter the competition for medical colleges or high posts, it is not, with rare exception, the most destitute who will utilize these preferences. Thus an economic ceiling for such higher-echelon preferences would have to be set fairly high in order to encompass a significant number of potential competitors.’For example, the family income limit (for medical college admissions) of Rs. 10,000 in Kerala or Rs. 3,000 in Bihar

excludes only a small fraction of the population. It has been estimated that over 80% of village families have incomes of less than Rs. 1,000 per year.26 264. The Study Group on the Welfare of the Weaker Sections munity (1961 :5—6) cites two 1961 studies, one in Orissa and one in which concurred that about 80% of rural households had annual Rs. 1,000. In the Punjab, where the ceiling was Rs. 1,000, a single

of the Village ComSouth India, both of incomes of less than judge questioned its

appropriateness on the ground that 90% of the Indian population would fall within the Backward Classes thus defined (High Court of Punjab, Civil Writ No. 1917 of 1965,

referred to Division Bench on 13 August 1965; order and opinion by R. S. Narula, J.).

The Measure of Backwardness

273

The Backward Classes Reservations Commission in Kerala exposed the underlying problem of the income ceiling. Having recommended that Backward Classes be families with aggregate annual incomes of less than Rs. 8,000, it then turns its attention to higher government posts: [When

promotion to a

selection grade becomes due to an employee in a

feeder category for the Gazetted post, he would almost have reached the

maximum of his scale of pay, which when combined with the other ingre-

dients of his aggregate annual family income, would be above the Rs. 8,000 per annum. So by the time he is ripe for promotion, he goes out of the Backward Class, because the economic limit for a Backward Class family is Rs. 8,000 per annum. If this is the case of officers in the feeder category for each of the lowest Gazetted categories of posts, no mention need be made about cases of those in the feeder category of higher gazetted posts. To illustrate: In the case of posts of Section Officers which are in the feeder category for posts of Under Secretaries in the Government Secretariat, Public Service Commission, etc., the economic limit may have to be raised

at least to Rs. 9,000 per annum per family. In the case of posts of Under

Secretaries, which are in Secretaries, the economic per annum per family. In in the feeder category for

the feeder category to selection posts of Deputy limit may have to be raised at least to Rs. 13,000 the case of posts of Deputy Secretaries, which are promotion to selection posts of Joint Secretaries,

the economic limit may have to be raised at least to Rs. 15,000 per annum per family. There are certain other feeder categories, the posts which carry scales of pay, the maximum of which is Rs. 1,300 per mensem with the usual

D.A. in certain Departments. In the case of this feeder category, the economic limit may have to be raised to Rs. 17,000 per annum per family. Therefore, when the State finds that the Backward Classes are inadequately

represented in the Gazetted category in the services under the State and

wants to make provision of reservation of posts in their favour, it can effectively do so, only if the economic limit of Rs. 8,000 is suitably raised to

different limits on the lines indicated above.26

The Committee envisions that the Balaji pronouncements against layers of Backward Classes may cause difficulties.26¢ But of course the layering here is an ironic reversal of the attempt to preserve some part of the benefits of the least well-off. Here, what is proposed is a layering to enable benefits to be extended to those who are doing well by any all-India standard. In such a scheme a high school graduate applying for a lower post would not be backward if he had a family income of Rs. 8,500, whereas a highly placed government officer earning twice 265. Kerala (Backward Classes Reservation Commission) 1971: 128-129.

266. Id., 129-134. CE. the discussion of layers in chap. 13, § C, below.

274

Identifying the Beneficiaries

that would be backward when he sought promotion to a yet higher position. We reach what is surely the reductio ad absurdum of the income

test—in which income is not used to identify deserving beneficiaries,

but the deservingness of beneficiaries is used to fashion the income test. In somewhat milder form, all income tests entail this paradox. If the cut-off is set high, so many are included that few benefits will go to the members of the most disadvantaged groups. If they are set low, it is difficult to employ preferential treatment to include historically disadvantaged groups in the upper reaches of national life.26” Thus different income tests are needed for different kinds of programs. Were an

income test designed to enable the more mobile members of backward communities to obtain medical college admissions used to distribute early stage benefits, it would amount to a provision for virtually the entire population. A state committed to such a definition of Backward Classes could not in effect provide them with any benefits it was not ina position to supply to the population at large. Or, it could only give some late-stage higher-echelon benefits in which most potential recipients would have been eliminated by their own lack of resources and earlier failures. But in the transition from early stage to higher echelon, the function of income tests is transformed. In the case of early-stage benefits, the income test is a measure of need or at least priority in responding to need.

At later stages, income becomes an indicator of cultural disadvantage.

But the later the stage at which a particular preference devolves on the individual—and most higher-echelon preferences devolve fairly far on in life and depend on prior success— the more questionable the rele-

vance of such tests as family income. If the father’s income has some

relevance to the deservingness of an applicant to medical college, why should his father’s present income affect the chances of a lawyer applying for a munsif’s post or the chances of a gazetted officer applying for a promotion?26® If the father’s low income is thought to reflect a constric267. The Supreme Court, noting the insufficiency of poverty as a test of backward-

ness, observed that “‘if poverty is the exclusive test, a very large proportion of the population of India would have to be regarded as socially and educationally backward. ... An untenable situation might arise. In this country except for a small

percentage of the population the people are generally poor—some being more poor, others less poor. Therefore when a social investigator tries to identify socially and educationally Backward Classes, he may do it with confidence that they are bound to be poor. His chief concern is, therefore, to determine whether the class or group is socially

and educationally backward” (Janki Prasad v. State of Jammu & Kashmir, A.1.R. 1973S. C. 930 at 937). 268. If it were the applicant’s own income that was measured, either the scale would have to be set high or one would confine the competition to lawyers who were either failures or dissimulators.

The Measure of Backwardness

275

tion of opportunities earlier in life, it is income at that earlier time that would be relevant. All preferences that directly redistribute access to higher-echelon positions (legislative seats, medical college admissions, higher Government posts) are going to go to individuals who already have more resources and opportunities than the average of the population. Presumably, the justification for conferring preferential treatment of this

kind on an individual is that he comes from a

class that enjoys fewer

resources and opportunities than the average, and preference for him is expected to benefit that class. It might do so in various ways, depending

on the way in which the classes are chosen. When backward communities are employed, it is the more prosperous and mobile within them

who receive these higher-echelon benefits. Presumably, they are linked to their less-fortunate caste fellows by ties of kinship, loyalty, and mutual support, so there is some reason to hope for a “‘trickle-down” that will spread and multiply the benefits. But where income tests are

used, there is little reason to think that those at the bottom will be

indirectly benefitted. Besides the spread of tangible benefit, the recipient might be expected to play a representative function.He might be a representative of the interests of the disadvantaged group—as in legislative seats and to a certain extent even in higher government posts— providing them with a spokesman or access or leadership. Or he might be a representative in a symbolic sense, giving the group a sense of participation in the achievements of the society and a model of possible

mobility

(as in higher professional

training). There is no

self-evident reason to think that the recipients under an economic test will serve this kind of representative function, tangible or symbolic, for those in the lowest groups.2 Pure economic tests would in several respects be incompatible with some of the doctrines developed in Balaji and accepted by subsequent courts—which need not, of course, be taken as a defect of economic

tests. First, as we have seen, many of the most prominent benefits are at levels of accomplishment that would require any economic ceiling to be placed fairly high.?”° Such a test could not possibly conform to Balaji’s notion of “absolute backwardness”—that the Backward Classes be 269. Unfortunately, in spite of numerous assertions about the broad effects of

preferences, Indian social scientists have not attempted to measure their impact on the

recipients, nor to compare the impact of alternative policies. So the statements above must remain speculative.

270. For example, in Jayaswalv. Principal, A. 1. R. 1968 Pat. 504, the court concluded

that it was not unreasonable to consider an annual family income of Rs. 3,000 (far above

the national average) as “poor” for purposes of attending a medical college. Cf. the response of the judge to the arbitrarily low cut-off of Rs. 6000 per annum in Shameem v. Medical College, A. 1. R. 1975 Ker. 13] at 138.

276

Identifying the Beneficiaries

confined to those “‘well-below the state average” or comparable in backwardness to the Scheduled Castes and Tribes.271 These standards could only be used with economic tests of the first type—i.e., economic tests applied to determine the backwardness of communities. Second, if it were recognized that the economic ceiling for purposes

of, e.g., medical admissions, would have to be higher than that for free

lunches, economic tests could be used only by sacrificing the notion ofa single uniform standard for all of the Backward Classes.272 Third, economic tests make it difficult if not impossible to apply Balaji’s notion that the Backward Classes are both socially and educationally backward. How is one to know that the low-income and/or low-occupation families are both socially and educationally backward?

In Viswanath v. Government of Mysore, the High Court found it difficult to conclude that the high-scoring successful students chosen in the “Backward Classes Pool” were educationally backward.273 Presumably, the high scores of individuals were not to the point— the individuals need not be educationally backward, but supposedly they belong to an educationally backward class. But there is no assurance that they are members of such a class unless it is assumed that the whole low-income and low-occupation class is socially and educationally backward. But if that is the case, then the “‘lion’s share” problem will be reproduced on a vastly larger scale, for it will be these low-income families which are least backward (in the sense of being least deprived of other advantages) who will get the benefits. At least in part, this will include those families with higher status, better contacts, and the other advantages that are associated with higher caste. So economic tests, when used in conjunction with merit, may be expected to minimize the convergence with other forms of social backwardness.?74 We have seen that the courts have erected increasingly stiff requirements of accurate data to support Backward Classes’ classifications. So far the courts have only invoked these standards where a state has employed the normally forbidden communal criteria. But in so far as economic (or other non-communal) tests are used to identify the reci-

pients of preferences which are purported to be authorized by Articles

15(4) and 16(4), it would appear that courts could require from states a 271. See §E above. 272. See chap. 13, §C, below.

273. A. 1. R. 1964 Mys. 132.

274. Furthermore, the economic test may well run afoul of the notion that reserva-

tions should be guaranteed minimums rather than provisions for places “over and above” the places gained on merit. (See chap. 13, §B.) Presumably, some of the backward (economically defined) will win places on merit. If they have to be subtracted before distributing of the reserved places, it will be necessary for the state to obtain

income certificates from all candidates, not only from those candidates who claim to be

The Measure of Backwardness

277

convincing showing of the relevance of such tests to remedy the sorts of backwardness that these provisions were designed to overcome. H. GEOGRAPHICAL CRITERIA As we noted earlier, Balaji said that the State might take into account habitation in identifying the Backward Classes.?’5 But a residence test, like economic tests, raises many problems with existing doctrine.

Reservations in admissions to medical colleges in favor of residents of

backward areas started to come before High Courts in northern India in the late 1960s. The judicial response was mixed. The Jammu and Kash-

mir Court allowed reservation of places for residents of Ladakh,?76

and one bench of the Allahabad Court allowed reservations for candi-

dates from rural areas, hill areas, and Uttarkhand Division because the

residents of these areas undeniably formed a “socially and educationally backward class” in relation to medical education.?77 However, another Allahabad bench found these reservations troub-

ling on the ground that the government had failed altogether to indicate the basis for regarding these classes as socially backward, and its criteria for finding them educationally backward were insufficient. The Court appeared willing to take judicial notice that Uttarkhand was so

backward that the designation of its residents as a socially and educa-

tionally backward class was justified, but it struck down the rural and

hill

areas reservations.?7® Similarly, in Gurinder Pal Singh v. State of

Punjab,” reservations for candidates from “Backward Areas” were struck down on the ground that such preference on residential grounds

did not distinguish between “[a] millionaire and a pauper living in such

areas,”2° and thus did not direct benefits to “the really deserving residents of such areas.” In the meantime the Supreme Court had encountered backward areas in the modified Jammu and Kashmir scheme which designated as Backward Classes two “areal” groups: residents of areas within five miles of the cease-fire lines, and residents of 696 villages in ten “bad pockets” (about 8% of the total population of the state). The border

areas were characterized?®! by inhibitions on mobility and investment

275. 276. 277. 278. 279. 280. 281.

A. I. R. 1963 S. C. 649 at 659. Sardool Singh v. Principal, Medical College, A.1.R. 1970 J. & K. 45. Subhash Chandra v. State, A. 1. R. 1973 All. 295, 297. Dilip Kumar v. Government of U.P., A. 1.R. 1973 All. 592. Gurinder Pal Singh v. State of Punjab, A. 1. R. 1974 P. & H. 125. Id., at 127, Janki Prasad Parimoo v. State of J. & K., A. 1. R. 1973S. C. 930.

278

Identifying the Beneficiaries

as well as severe restrictions on free movement; the bad pockets were

distinguished by difficult terrain, absence of vehicular communication, and general inaccessibility, as well as scant irrigation, lack of electric power, and so forth. The Supreme Court held it allowable to regard the

residents of the bad pockets and cease-fire areas as socially and educationally backward, but required that the government cure defects in the rules to insure that outsiders would not secure advantages that should

be confined to “genuine residents.”28? When the U.P. scheme was litigated again in Pradip Tandon,?® a Full Bench of the Allahabad High Court struck down the reservations on the ground that the government had indulged in unwarranted generalization, not justified by any investigation. Even if the paucity of candidates from these areas reflected their educational backwardness, there was no

data on their social backwardness, and to regard 80% of the population as socially backward would be “‘an appalling opinion, staggering to one’s imagination.”2® Basically, the reservation amounted to an impermissible classification by place of birth.285 In November 1974, the Supreme Court decided appeals from the confliciting Allahabad decisions.28 The Court was satisfied that both the populations of the hill areas (2.5 millions) and of the Utiarkhand Division (0.75 millions) were indeed socially and economically Back-

ward Classes. Because these areas displayed traits of economic undevelopment, isolation, and absence of education institutions, their populations might be regarded as socially and educationally backward classes. But the attempt to deem the 75 million inhabitants of Uttar Pradesh’s rural areas (80% of the state’s population) a Backward Class

was undone by the lack of heterogeneity of the class: “They are not of

the same kind. Their occupation is different. Their standards are different.””287 Weare left with the notion that the population ofa locality or a region may form a Backward Class. But as the area gets larger, it will become increasingly difficult to show that its inhabitants display sufficient homogeneity to qualify asa class. As in the use of communal units, some over-inclusion is unavoidable—a few who are not personally backward may be included. But geographical classes may be even more suscepti282. Id., at 942-43. 283. Pradip Tandon v. State of Uttar Pradesh, A. 1. R. 1975 All. 1.

284. Id., at 7.

285. Id., at 8.

286. State of Uttar Pradesh v. Pradip Tandon, A.1.R. 1975 S.C, 563. (This was an appeal

from Pradip Tandon, n. 283 above, and from Subhash Chandra, n. 277, above.)

287. Id., at 569. In addition, the Court stresses that the rural population is a majority.

Cf. chap. 12, §E, below.

The Measure of Backwardness

279

ble to over-inclusion than communities. While the inhabitants of geographical areas do share similar conditions of bad roads, isolation, lack of schools, etc., they may differ sharply in their resources and in their capacity for coping with these shared problems. Extending benefits that require prior training and advantages

(like medical admissions)

to

geographical classes would, one might expect, lead to the enjoyment of these benefits by the most advantaged within the beneficiary group, those who are personally least socially and educationally backward. "Geography can be used in conjunction with other factors as well as independently; and it can be used to disqualify as well as to qualify for benefits. Lists of eligible groups are typically based on area and contain area

restrictions.28

pose

area

Location

may

be used

not only

to select be-

neficiaries but to screen otherwise eligible beneficiaries— analogous to the income cut-off provisions discussed above. Government may imrestrictions

on

the

enjoyment

of certain

benefits.

The

Supreme Court held that it is permissible for the State to give favorable tax treatment to income of Scheduled Tribes earned within specified tribal areas and to withhold that beneficial treatment of income earned outside those areas.?49 I. NON-INCLUSION CLAIMS Of the 46 reported litigations involving preferences for Backward Classes, 29 were concerned with the designation of those classes by the State.29° Almost all involved challenges to the preferential policies on the ground that the list of Backward Classes was too inclusive. As we have seen, a significant portion of these claims have been successful. Of the 24 cases in which such a challenge was raised, the government’s action was struck down (or modified) in 17, and in several others the government was warned to take corrective action to avoid future invali-

dation of its list.294

On the other hand, claims that the lists of Backward Classes were

insufficiently inclusive have not found a receptive audience in the

courts; they have been reluctant to consider claims by members of an

allegedly backward class who assert that they should have been included. In Partha v. State of Mysore, a scheme for reservations was attacked 288. See discussion of area restrictions at chap. 5, §B, above.

289. I. T. Officerv. Rymbai, A. 1. R. 1976 S.C. 670 [NIC]. 290. These figures are based on a universe of 110 reported cases concerning com-

pensatory discrimination from 1951 to 1977. The derivation of that universe is described

below in chap. 14, n. 45.

291. To appreciate the significance of this rate of success, see chap. 14, §C, below.

280

Identifying the Beneficiaries

on the ground that the Lingayat community should have been included in the list of Backward

Classes, which included others who were no

more backward than they.29? The court rejoined that so long as the

communities which are included are all backward ones, “‘the fact that

the list may not be exhaustive of all the Backward Classes in the State” does not require that it be struck down, since the claim for inclusion is, in effect, a claim under a Directive Principle (i.e., Article 46) and therefore non-justiciable.293 The contention in Partha was, from the court’s point of view, that the line had been drawn at the wrong place; but it was not a case of demonstrated comparability with the groups under the line. Similarly, in the cases challenging the validity of the income cut-off, the issue was the validity of drawing a line separating members of listed communities whose annual incomes were more than Rs. 10,000 and

those with smaller incomes. Could the state include the latter but omit the former? In _Jayasree v. State of Kerala, the Supreme Court emphasized that it was a function of the State to “evolvfe] . . . proper criteria” for “determining who are socially and educationally backward classes,”294 while it is the jurisdiction of the courts “to decide whether the tests applied are valid.”295 The state’s determination that social and economic backwardness did not extend to the more affluent sections of the listed communities was found to be an application of valid tests. With the development of requirements of objective standards supported by sufficient data, it is doubtful that the courts would permit the exclusion ofa group which met the standards which the state purported to use in choosing Backward Classes. The claim that group X meets all

the criteria by which groups Y and Z were designated as backward would seem to be litigable on equal protection grounds. The more “objective” the criteria the courts demand of the state, the less reason for any reluctance to admit such claims for inclusion. Claims that the standards should be modified to include group X as well as groups Y and Z might still be dismissed as non-justiciable claims

under Article 46, as Partha suggests. But that case was decided at the

very beginning of judicial supervision of state selection of Backward Classes. That jurisprudence, requiring that the state employ valid tests of backwardness, provides many ways to challenge the validity of the

state’s tests. And the emergence of the theme that compensatory prefer-

292. A. I. R. 1961 Mys. 220. 293. Id., at 238. 294. The rationale of the State for the classification was that Lingayats had been excluded because their percentage of literacy was above the state average. See Govenment of Mysore 1960. 295. A.I.R. 1976 S.C. 2381 at 2385.

The Measure of Backwardness

281

ence is itself part of the Fundamental Right to equality guaranteed by Articles 14, 15, and 16 suggests fertile doctrinal ground for erecting claims

for inclusion.2% So far, however,

such claims on behalf of

excluded groups have not been forthcoming. There is, however, another judicial path to questions of inclusion, one that has been more heavily trafficked. Claims for inclusion may be put forward not only on behalf of groups who are not designated, but also by individuals whose inclusion within the groups that have been designated has been challenged. Such claims may take several forms:

group X of which I am a member is really part of (or equivalent to)

group Y, which is being provided benefit A; I am really a Y and I have not ceased to be a Y because I have become an X. Although I was formerly an X, I have now become a Y. It is typically more manageable for litigants to assert or contest personal membership in a listed group than to challenge the structure of the state’s list. When raised in the form of disputes about individual membership, courts cannot avoid the task of determining the boundaries of the designated groups. Thus some problems of inclusion are litigated in the guise of questions of individual membership. It is to these questions that we now turn. 296. See chap. 11, §D, below.

9

Membership in Groups Entitled to Preferential Treatment IN THE previous chapters we have been concerned with the problems of choosing the groups that may receive preferences. In this chapter we shall examine the corollary problems of determining whether an individual is a member of one of these preferred groups. We

have seen that Scheduled Castes, Scheduled Tribes, and Backward

Classes have for the most part been designated in terms of communal units: caste, tribe, and religion have been the primary differentiae. Some diversification of criteria has occurred: in some instances communal units have been supplanted by, or combined with, economic or occupational tests. The prospect is that in spite of such diversification, communal criteria will continue to be at least one element in the designation of the recipients of preferences—especially of the Scheduled Castes and Tribes. With membership in specified communal groups a qualification for preferment of various kinds, it is not surprising that disputes have arisen about such membership. In numerous cases the courts have had to pass on the question of whether a person was in fact a member of such a preferred group. This is not to say that such determinations are normally either disputable or disputed. Typically, for purposes of qualifying for most schemes of preferences, membership in the preferred group is evidenced either by simple declaration or by obtaining a certificate to that effect from some government official.1 These declara1. Where there is sufficient material to show that a person belongs to a preferred group, failure to submit the required certificate is not grounds for withholding the benefit. But to regard a person as Scheduled Caste, Scheduled Tribe, or Backward

Class “without any material whatsoever” is arbitrary and capricious and will be struck

down ( Mahendra Nath Pathak v. Stats of Assam, A. 1. R. 1970 A. & N. 32, 34). (Petitioner had made a bare assertion: “I am belonging to the Scheduled Tribe community. The

certificate will be submitted later on.” It could not even be ascertained which tribal community he claimed to belong to.) However, eligibility does not depend entirely on

claiming membership in an application, but depends on whether a candidate supplies

information which, if verified, would show membership in the listed group (P. Susila v.

Group Membership

283

tions and certificates remain unchallenged in the great majority of

cases. Virtually all of the cases in which membership has been disputed

involve election contests in which a successful candidate for a reserved seat is challenged on the ground that he is not a member of the group for which the seat is reserved. The election situation offers unique incentive to disprove membership of a particular individual: there is a single affected party for whom

the stakes are high enough

to justify the

expense of litigation and who is likely to have the resources (and backing) to undertake it.?

The cases raise many puzzling questions: Is membership in a caste or

tribe to be determined solely by birth? or by allegiance? or by the

opimionz of its members or of the general neighborhood? Does one lose

(or gain) caste membership by joining a sect or another religion? by excommunication? by assimilation? Does one lose tribal membership by claiming or achieving caste status? Who is a communicant of a particular religion? those born into it? those who have been converted to it? those who adhere to its precepts? What is the effect of unorthodoxy? of excommunication? Are the tests used for the application of personal law appropriate in the area of preferences? Before taking up these questions, which involve instances in which membership ina listed group is rendered problematic by the existence of an additional affiliation or by some factor which appears to vitiate membership in the group, there is the preliminary question of the meaning of the lists themselves. A. READING THE SCHEDULES

Just which groups are on the lists of Scheduled Castes and Scheduled Tribes is not as simple a matter as it might appear. The nomenclature of communal groups in India is equivocal and ambiguous, fluid and shifting. A high-level official committee noted that State of Madras, A. 1. R. 1970 Mad. 399, 400). Inconsistent assertions on other occasions

will not necessarily defeat a claim of membership (Bhaiya Ram v. Anirudh, A.1.R. 1971S. C. 2533).

2. Most challenges of claimed group membership that reach the courts are ini-

tiated by disappointed rivals. It is, of course, possible for such challenges to originate with governmental authorities monitoring the-distribution of preferences (for an apparent example, see Kajari Saha v. State, A. 1. R. 1976 Cal. 359) or with organizations acting

on behalf of genuine recipients. The Evening News [New Delhi] of Jan. 19, 1974, reports

that police registered a charge of cheating against two doctors who were alleged to have obtained admission to medical college on false certificates of being members of Scheduled Castes. The case was registered following a complaint lodged by the Society for the Welfare of Children and Women.

284

Identifying the Beneficiaries

the names by which certain groups and sub-groups are known vary from

district to district and sometimes from area to area within a district. Moreover, there is a tendency for castes and sub-groups to give themselves new and fanciful names from time to time, and to resent being called by any

other name.3

The lists of castes and tribes that make-up the Scheduled Castes and Tribes Orders make a valiant attempt to grapple with this by including some synonyms and specifying some sub-groups, but other synonyms

and sub-groups are omitted— by design or otherwise. To make matters worse, many of the names used havea shifting denotation, being used at

different levels of generality—e.g., the same term may be used to refer to an endogamous caste group and toa cluster of related castes.4

Sometimes

more

than one community

bears the same name.$ The

“‘name”’ of one group may be used as titles or surnames in another.6 A

group’s name may connote a

profession followed by others as well.”

The lists also include many entries that seem to refer to no group at all—some are merely general appellations (e.g., Chandal, Panchama), others are groups that cannot be found by the census-takers in the area specified.8 The confusion is compounded by problems of inconsistent translation, transliteration, and spelling. The problems of nomenclature and enumeration are detailed by the Superintendant of Census

Operations in Madras:

Synonyms used by certain castes are the names of certain other main castes. For

instance,

the

Chakkiliyans

in certain

areas are

known

under

the

synonym Dombars. Dombara is also a different Scheduled Caste with the synonym of Dombars. Again in Thanjavar district, some Pallars are known by the synonym of Panikkars. But Panikkars in other areas always do not

belong to Pallars. Similarly, Pandaram is a synonym of Valluvan, but all Pandarams are not Valluvans. Again, Andi Pandarams do not belong to the

Scheduled Castes. Jogi is a synonym of Kuravan. But it is also a synonym of Dom or Dombara in Kanyakumari area. Uralis and Pulayans are found both in the Scheduled Castes and Scheduled Tribes list. In spite of clear instructions issued by us, generic names like Harijans, Scheduled Castes and Scheduled Tribes have been used in the Census return.?

3. Department of Social Security 1965: 20. 4. Cf. Fox 1969. 5. Depatment of Social Security 1965: 13. 6. RCSCST 1961-62: I, 5. 7. The Punjab Evaluation Committee Welfare Department (1966: 141) reports that

in such cases higher caste “persons not acutally belonging to Backward Classes try to

avail of the benefits by displaying their professions and making a caste out of it.” 8. Department of Social Security 1965: 15. 9. Census of India, 1961, Vol. IX, Madras, Part IV-A (1), Scheduled Castes and

Tribes (Report and Tables), p. 13.

Group Membership

285,

And, of course, the very existence of the lists acts to induce movement

across the line into favored categories by manipulation of equivocal nomenclature.1© In view of this confusion, it is sometimes a problem merely to know whether a group to which somebody belongs is a group that is listed on one of the Schedules. A review of the Supreme Court’s experience gives some idea of the intractability of these problems to judicial solution.

In Basavalingappa v. Munichinnappa," election ofa candidate to a seat reserved for Scheduled Castes was challenged on the ground that he

was not a member belong to the Bhovi but his challengers which there was no

of a Scheduled Caste. The candidate claimed to caste, a group listed in the Scheduled Caste Order, claimed he was a member of the Voddar caste, for listing in the Order. The Election Tribunal found

that Bhovi, as listed in the order, was a sub-caste among Voddars, and

since the candidate did not belong to this sub-caste he was not eligible for election. The High Court reversed, finding that the Bhovi caste mentioned in the Order was no other than the Voddar caste. The Supreme Court, cautioning that it was not ordinarily open toa person to offer evidence to prove that his caste A is part of a caste B listed in the Order, decided to evaluate the evidence in view of the peculiar circumstance—i.e., that caste B is otherwise nonexistent. There had been no

Bhovis anywhere in Mysore before the States’ Reorganization of 1956. But the 1950 Order listed Bhovis in Mysore, and it must have been intended to include some group. Thus the Court felt justified in examining the evidence to determine which caste was meant by this term. The evidence showed that a Voddar conference had in 1944 resolved to change their name to Boyi and had communicated this to the Mysore Government, which had in 1946 passed an order that the community be called Boyi in all government communications and records. The caste was

also sometimes

spelled

Bovi

or Bhovi.!2 The

Voddars,

found, were indeed the Bhovis referred to in the Order.'3

it was

A few months later, in Bhaiya Lal v. Harikishan Singh,'* the Supreme Court made good its warning that it would not ordinarily permit members of group A to show that that group was a part of listed caste B. 10. A dramatic instance of such movement appears in the case of the Sunris and Sahas discussed below at nn. 18 ff. 11. A.LR. 1965 S.C. 1269. 12. Cf. Konkani v. Shankar Rao, 3 E. L. R. 409 (1953), where the Election Tribunal!

found that the Konkni community, the Konkna caste, and the Konkna tribe were in fact the same. 13. More than a attempted to have Bhovi (Shri G. Y. 1976)).

decade later, when the lists were revised, a Karnataka M.P. vainly “‘Waddars” included in the Scheduled Castes list as a synonym for Krishnan at LSD [5th Series], Vol. 64, No. 16, cols. 103-4 [2 Sep. 14. A. 1. R. 1965 S.C. 1557.

286

Identifying the Beneficiaries

A candidate elected to a Scheduled Caste seat had declared in his nomination paper that he was a Chamar. (The Scheduled Caste Order contained an entry “‘Chamar, Jatav or Mochi”.) The Election Tribunal and the High Court declared his election invalid on the ground that he was not a Chamar but a Dohar—a group not on the Scheduled Caste list in that district. The candidate claimed that he was a Dohar Chamar (and that this group was also called Mochi.) The Court refused to allow him to offer evidence that Dohars were a sub-caste of Chamars on the

ground that it could only determine whether or not he was a member of the listed ‘caste—i.e., whether he was a Chamar.15 Evidence showed

that Dohars did not interdine or intermarry with Chamars, that they lived in different mohullas, that they had their own Dohar samaj. The

Court concluded that he was not a Chamar. These cases perturbed the authorities in charge of administering preferences, who had always acted on the assumption that every listing included any sub-groups and that sub-groups need not be specified separately.1° The Lokur Committee, pointing to the need for accurate listing of all synonyms, phonetic variations, and sub-groups, predicted that it would be “impossible to include all sub-groups.”!7 The elusiveness of the facts in these cases and the Supreme Court’s despair of ascertaining them is demonstrated in the next case to reach the Supreme Court, Abhoy Pada Saha v. Sudhir Kumar Mondal.*® The Scheduled Caste Order list for Bengal included an item “Sunri excluding Saha.” One Mr. Saha, a successful candidate whose nomination per described him as a member of the Sunri caste, was challenged on the ground that he was not a Scheduled Caste. The Election Tribunal thought that Sahas, though originally a group within the Sunris, had developed into a distinct caste that had severed all connections with the Sunris.!9 According to the Tribunal, Mr. Saha belonged to the Sunris 15. But cf. Balchandv, Laxminarain Mateh, 8 E. L. R. 465 (1953), where Jatavas were

held to be a part or group within the Scheduled Caste of Chamars, where the item included “‘Jatia or Jatav Chamar.”

16. Department of Social Security 1965: 19. The Ministry of Law had opined that all

parts ofa listed group were entitled to benefits, even sub-divisions which were not listed

(RCSCST 1958-59: I, 10). The establishment of the Lokur Committee is attributed to

the pressure on government of the Supreme Court’s requirement of comprehensive

listing (RCSCST 1965-66: 65). 17. Id., at 20. 18. Abhoy Pada Saha v. Sudhir Kumar Mondal, A. 1. R. 1967 S.C. 115.

19. Census figures suggest that the Tribunal may have been on the right track. The reported Sunri population declined steadily from 114 thousand on the 1901 census to

45,000 on the 1951 census, presumably because members of this group were reporting

themselves as Saha. But the Sunri population rose to 106,000 in 1961 and 166,000 in

1971. See West Bengal Tribal Welfare Department Bulletin of the Cultural Research

Institute 11 (1): 43, for figures to 1951; Census of India, 1961, Vol. XVI, Part V-A (i), p. xliv, for 1961; Census of India, 1971, Series 1, Paper 1 of 1975, Scheduled Castes and Scheduled Tribes (1975) : 92, for 1971.

Group Membership

287

and therefore was not a member of the independent Saha caste, which

had, because of their historical association, been excluded by the Order

as a precaution. The High Court, on the other hand, thought that the

effect of the listing was to exclude all Sunris with the surname of Saha, a procedure it found within the President’s power under article 341 to exclude parts of castes. The Supreme Court rejected both views and expressed extreme dissatisfaction with the Tribunal’s reading of the Order. The Tribunal was wrong in thinking Saha an independent caste. Since Sunris are a caste, the Court deduces from the phrasing of the listing, Saha must be a sub-caste. The error of the Tribunal lay in interpreting the Order in the light of the evidence before it. There was no justification for doing that. After all, the evidence in a case may be imperfect. . . . A method of interpreting a statutory provision which might lead to such uncertainty cannot be correct.

. . . Evidence cannot alter the

natural interpretation of the words in the Order.”

Sahas then were a smaller caste within the Sunri caste. Since there was

evidence that the candidate was within the larger, but none that he was in the smaller, he is a Sunri and the election is declared valid.24 In its understandable reluctance to become embroiled in the

evidentiary quicksand, the Court attempts to escape by a rigid formalism which assumes that the “‘text” (i.e., the Order) contains ans-

wers which can be elicited by proper interpretation. The “text” here is admittedly a very crude approximation to the social actualities. Yet the Supreme Court rejects the empirical finding of the Tribunal about

the actuality, because it does not conform to the logical structure of the

Order, and substitutes a finding in accordance with that logic. In Laxman Siddappa Natk v. Kattimani Chandappa Jampanna, the Supreme Court attempted to formulate some standards for dealing with these troubling cases. A successful candidate for a seat reserved for the Scheduled Tribes was challenged on the ground that he was not a member of the Nayaka tribe, as declared in his nomination paper. The applicable list of Scheduled Tribes, which was adjusted after the transfer of the district from Bombay to Mysore in the States’ Reorganization

in 1956, had an entry for “Naikda or Nayaka including Cholivala Nayaka,

Kapadia Nayaka,

Mota Nayaka and Nana Nayaka.” The

20. A. 1. R. 1967S.C. 115 at 118.

21. A further episode in the Sunri-Saha story is found in Kajari Saha v. State, A. 1. R. 1976 Cal. 359, where a medical college applicant who claimed to be Sunri with the

surname Saha was found by welfare department investigators to be a member of the

Vaishya Saha caste and not a Sunri at all. The disappointed applicant's father was a government servant, hired against a reserved place, of more than 20 years service. 22. A. T. R. 1968S. C. 929,

288

Identifying the Beneficiaries

challenger claimed that the candidate was not a Nayaka, but a member

of the Bedar caste. The candidate claimed that “he was not a Naikda but a Nayaka and that Nayakas were also known as Bedars.”23 The High Court, after examining census reports and other documentary evidence, concluded that there were no Nayakas in this districtand that the candidate was a Bedar. The Supreme Court read the earlier cases as making the determination of membership a question of fact to be proved by evidence—a reading not immediately apparent. Where membership was contested, the onus lay on the challenger to prove that it was not as claimed. The

challenger might disprove the claim to belong to a group by evidence of

different customs or by showing social acceptance by the other community. Here, in spite of suspicious circumstances surrounding the candidate’s certification as a Nayaka, the challenger had failed to discharge this burden. “There was no evidence one way or the other. In these circumstances, the election petitioner could not succeed because

of the weakness of the appellant’s case.””24 Here the Court moves from statutory interpretation to the rules of evidence. But if an individual’s membership in a group is a question of. fact, are not the relations that obtain between groups— whether they are co-extensive, distinct, overlapping, one within the other—also questions of fact which cannot be answered by deductive reasoning? Apparently not, for a few months later the Supreme Court encoun-

tered yet another case of this kind. In Pars Ram v. Shivchand, a candi-

date’s nomination paper was rejected on the ground that he was not, as claimed, a Chamar, but was a Mochi. (Chamar was on the Punjab’s Scheduled Caste List, Mochi was on the state’s list of Backward Clas-

ses.) The High Court found that Mochis and Chamars were distinct groups and that apparently “Hindu Mochis have developed into a higher caste or caste-group than that of the Chamars, particularly in the matter of social status.”25 Following Bhaiya Lal, the High Court held that since Mochi was not mentioned as a synonym of Chamar, it is not the same

caste; since it is not mentioned

in the Order,

it is not a

Scheduled Caste and the candidate was not eligible to stand for a Scheduled Caste seat. On appeal the Supreme Court affirmed, holding that the claim that both appellations referred to the same caste “tis not open

23. Id., at 932. This dispute appears to be an artifact ofa freak in the process of

delimiting Scheduled Tribe constituencies. It was discovered too late that the four

districts in the constituency contained only 2% Scheduled Tribes population. There are

no members of the Nayaka tribe in the area, but the Bedars, who predominate in the

area,

are

also called

Nayakas.

(Personal

communication

knowledgeable Mysore lawyer.) 24. Id. 25. A. 1. R. 1968 P. & H. 331, 335.

to the author

from

a

Group Membership

289

to agitation by evidence.”26 Since the determination of the lists lies within the exclusive power of the President, “it is not for us to examine it

and come to the conclusion that ifa person was in fact a Mochi, he could

still claim to belong to the Scheduled Caste of Chamar....”?7 Thus

the Supreme Court reasserts its unwillingness to review the factual

determination of the relation between groups.?8 It assumes an ordering

of groups that corresponds to the ordering of names in the presidential

list.

Yet, two years later, in Bhaiya Ram Munda v. Anirudh Patar,?9 the

Supreme Court did not hesitate to review evidence and conclude that Patars, although not mentioned among the sub-tribes of Mundas included in the Scheduled Tribes Order, “are it appears regarded as the lowest in the social order amongst the Mundas but they are still Mundas.”3° The fact that some sub-tribes are enumerated in the Order does not give rise to an inference that those not enumerated are not Mundas. The Court refused to read earlier cases as foreclosing a reading of the list to match what the evidence disclosed to be the factual situation. These cases do not involve any large general questions; the facts are elusive and

obscure

and

difficult for a court

to elicit.

(Indeed,

the

benefits of appeal on the facts are not readily apparent.) Exposed to such an imponderable situation, there is a temptation to impose order 26. A. I. R. 1969 S. C. 597, 600. 27. Id., at 600. Mochi was listed as a Synonym of Chamar in the Scheduled Castes lists for 11 states—all but Assam, U.P., and Punjab. When Punjab was divided in 1966, the

item remained unaltered for Punjab and Haryana, but Mochis were listed along with Chamars in both Delhi and Himachal Pradesh. The Supreme Court cites this difference in treatment as evidence “that the question of inclusion of Mochis in the Scheduled Castes was considered” by the President as recently as 1966 (id., at 599).

28. In striking contrast to this timid retreat from the tangled facts is the sensitivity to the fluid and ambiguous character of these identities displayed in another case involving the same cluster of groups in the Punjab. In Didar Singh v. Sohan Singh, A. 1. R. 1966 Punj. 282,288, a candidate who claimed to be a Ramdasia

by caste was declared

unqualified by the Election Tribunal on the ground that he was a Rehtia. (The Scheduled Caste Order for the Punjab included an item specifying Chamar, Jatia

Chamar, Rehgar, Raigar, Ramdasi or Ravidasi— but Rehtia was not listed.) The court declined to take Rehtia as mutually exclusive of Ramdasia. One could not exclude the

possibility of “a Ramdasi being also called a Rehtia. In this part of the Country [the

position of groups like Chamars] . . . seems to have been traditionally in a fluid state, in so far as their description goes. The expressions Ramdasi, Ranidasia or Ravidasia .. . do not seem to possess or convey any clearcut or crystallized distinctive characteristics.

. .. [Nor do] these expressions respectively carry precise, exact and distinctive meanings excluding the possibility of some individuals . . . using any one of these epithets interchangeably. . . . [T]he position has all along been somewhat confused and certainly imprecise.” 29. A. 1. R. 1971S. C. 2533.

30. Id., at 2538.

290

Identifying the Beneficiaries

on the factual muddle by adopting formal deductive solutions. The choice between grappling with a very elusive and muddled social actuality or invoking a formal logical order to encompass it is a theme

we shall meet again as we explore the judicial treatment of various

membership problems. But they point to a general phenomenon that deserves reflection. These cases involve a contest about some coveted prize—usually an electoral contest, sometimes a medical admission—and the candidate

seeks to establish his eligibility or to disqualify a competitor. These efforts highlight the opaque and malleable qualities of the lists, qualities

that are present even when less-exalted benefits are being distributed

and there is less incentive for anyone to mount a challenge (or defense) of a particular assertion of group identity. Manipulative practices that we see in bold reliefin cases involving important prizes may exist in less visible but widespread form, availed of routinely and normally without challenge.31 The adequacy of the judicial response to questions of group membership cannot be assessed solely by the incisiveness and subtlety with which it addresses litigated disputes. Such an assessment must also consider the framework that judicial treatment provides for routine administration of preferences. B. CASTE AND SECT The tests for determining group membership were first given sive consideration by the Supreme Court in Chatturbhuy Vithaldas v. Moreshwar Parashram.3? An Election Tribunal had rejected the nation papers for a reserved seat submitted by a Mahar who had

extenJasani nomijoined

the Mahanubhava Panth, a Hindu sect which repudiated the multipli-

city of gods and the caste system. Reversing the Tribunal, the Supreme Court held that the candidate remained a Mahar and was thus entitled to stand for the seat reserved for Scheduled Castes. To determine whether adherence to this sect made the candidate cease to be a Mahar,

the Court specified three factors to be considered: “‘(1) the reactions of the old body, (2) the intentions of the individual himself and (3) the rules of the new order.”33 31. E.g., the Tamil Nadu Backward Classes Commission (1971: I, 160) notes that

“some forward or social higher castes have sometimes a sub-section whose caste name is more or less similar or identical with a name found in the list of Backward Classes. Without compunction such loopholes are freely made use of.” Comparable routine

misrepresentation of income is discussed in §G, below. 32. (1954) S. C. R. 817. Hereafter referred to as Jasani.

33. Id., at 838.

Group Membership

291

The candidate was admitted to all Mahar caste functions and had been allowed to marry within the community. He twice married Mahar girls, neither of whom were Panth members at the time of the marriage. He always identified himself as a Mahar. The Panth, in spite of its doctrinal repudiation of caste, had not penalized him for his adherence to the caste. The Supreme Court concluded that “conversion to this sect imports little beyond an intellectual acceptance of certain ideological tenets and does not alter the convert’s caste status.” It is clear that the primary

consideration was the second test—i.e., the intentions of the convert

himself— intention not in the sense of mere declaration but as evidenced by a consistent course of conduct and dealings.35 The Court applied the “broad underlying principle” of Abraham v. Abraham,?6 decided a century before by the Privy Council in determining the law of inheritance applicable to a Hindu convert to Christianity: “{h]e may renounce the old law by which he was bound, as he has renounced his old religion, or, if he thinks fit, he may abide by the old law, notwith-

standing he has renounced his religion.”37 Applying this principle, the Supreme Court found that “‘if the individual . . . desires and intends to retain his old social and political ties,” and if the old order is tolerant of

the new faith and does not expel the convert, the conversion has no

effect. “‘On the other hand, if the convert has shown by his conduet and

dealings that his break from the old order is so complete and final that

he no longer regards himselfas a member of the old body and there is no

reconversion and readmittance to the old fold . . . [he cannot] claim

temporal privileges and political advantages which are special to the old order.”38 Although the test is primarily one of the convert’s intention or behavior, this intention must be confirmed through acceptance by the old group. The inclusion of this additional test is important because, as 34. Id., at 840. The Panth is a devotional sect, founded in the 1 1th century, which not only eschewed caste and polydeism, but also challenged the validity of the Vedas, image worship, and the system of asramas (stages of life). On the role of this sect in Mahar tradition, see Miller 1966; Zelliott 1969: 26 ff.

35. Thus unsupported contentions that “I am belonging to the Scheduled Tribe

community” from which “it cannot be ascertained to what tribal community he claims

to have belonged” are insufficient to establish membership (Mahendra Nath Pathak v.

State of Assam, A. 1. R. 1970 A. & N. 32, 34). 36. Abraham v. Abraham, 9 M. 1. A. 199, (1863). The rule was subsequently overturned by the Indian Succession Act, 1865 (now Indian Succession Act, 1925, §58). But the

courts are still divided over whether the Hindu rule of survivorship is applicable to Christian families who continue to be joint after conversion. 37. Reporter's note at 9 M. I. A. 196. Cf. id., at 242-44.

38. (1954) S.C. R. at 838.

292

Identifying the Beneficiaries

the Court says, “the only modification choice which must be taken into account whose religious tenets he has renounced, considering is the right of the old body,

here is that it is not only his but also the views of the body because here the right we are the right conferred on it as a

special privilege to send a member of its own fold to Parliament.”39

The third test, “the rules of the new order,” is of minor significance.

Since it is the legal and political rights of the old body that are being considered, “the views of the new faith hardly matter.” “The new body is free to ostracise and outcaste the convert from its fold if he does not adhere to its tenets but it can hardly claim the right to interfere in matters which concern the political rights of the old body, when neither the old body nor the convert is seeking legal or political favours from the new body as opposed to purely spiritual advantages.”’*1 If this test has to be taken into account at all, it is only as indirect evidence of the intentions and conduct of the convert. For example, continued acceptance by a new group which was notoriously intolerant of the retention of the old ties might well evince an intention to break with the old group. But here the Court found it “evident that present day Mahanubhavas admit to their fold persons who elect to retain their old caste customs.””42

In Shyamsundar v. Shankar Deo Udgir,* the principles of the Jasani case were applied to decide whether a candidate for a reserved seat had lost his membership in the Samgar caste by joining the Arya Samaj, a Hindu sect which rejects idolatry and ascription of caste by birth. He had been accepted for membership in a local Arya Samaj organization, had paid membership dues, had married a girl of Sonar caste in accordance with Arya Samaj rites, and had reported himself as an “Arya” in the 1951 census. The Mysore High Court, citing the Jasani

case, announced

that there would be no deprivation of caste unless

there was either (1) expulsion by the old caste or (2) intentional abandonment or renunciation by the convert. There being no evidence of expulsion or ostracism by the old caste, the question was whether there had been a break from the old order “‘so complete and final that . . . he no longer regarded himself as a member of the Samgar caste.”45 The 39. Id., at 839.

40. Id., at 838.

41. Id.

42. Id., at 840. For a situation in which the rules of the new order played a more

prominent role, see Rhagava Dass v. Sarju Bayamma, A. 1. R. 1942 Mad. 413, where by

Joining the Byragi sect a person ceased to belong to his original caste. 43. A. I. R. 1960 Mys. 27.

44. On the Arya Samaj, see Lajpat Rai 1915. The caste referred to as “Samgar” is evidently the same as that listed as “Samagara” in Census of India 1961 (Paper No. 2: 84). The Sonar caste, mentioned below, are traditionally goldsmiths and of a higher social standing, usually associated with Vaishya status. See below, n. 54. 45. A. 1. R. 1960 Mys. at 32.

Group Membership

293

court found that his activities evinced that he regarded himself as a Samgar, as did his testimony that he believed in idolatry and in texts repudiated by the Samajists. The court found no evidence that he could not have married the Sonar girl “in the ordinary way” and thus the marriage was not inconsistent with his membership in the caste, nor was the census report, since the court refused to accept Arya as equivalent to Arya Samajist. Almost as an afterthought, the court notes that the Samaj did not expel him for departure from their tenets. Such expulsion would only have reinforced the court’s conclusion, where the

absence of it (if. there ever were expulsions) might indicate an acceptance inconsistent with his remaining in the caste. The test that emerges,

somewhat inchoately, is that so long as the person identifies himself with the old caste and is accepted by the caste—no matter if he is accepted by the new group or not—he remains a member of the caste for purposes of eligibility for nomination to a reserved seat. C. TRIBE In Kartik Oraon v. David Munzni,*® the challenged candidate was a

member of an Oraon family that had been Christian since his grandfather’s time. His eligibility to stand for a seat reserved for Scheduled Tribes was challenged on the ground that as a Christian he was no

longer an Oraon, since he had abandoned the animistic faith, did not

follow the manners and customs of the tribe, and had no affinity of interests or aspirations with the tribal people. The High Court found that the candidate’s active participation in the civic life of the tribe belied the charge of lack of common interests and aspirations. As to manners and customs, there was evidence that although Christian tribals did omit certain observances of tribal religion, they retained such practices as exogamy based on totemistic lineage, certain harvest rites,

ceremonial eating of first fruits, birth and marriage observances, and style of writing surnames. Even if he omitted to observe certain festivals or observed some in a manner different than other tribals, the court

concluded, the “most importnt thing ... is that the non-Christian tribals treat the converted Oraons as tribals, calling them “Christian Oraons.’’47 Christian and non-Christian Oraons intermarry and their descendents are treated as full members of the tribe. Christian Oraons are invited to feasts and participate in them. Applying the Jasani tests, which it found “fully applicable,” the court concluded that conversion did not extinguish membership in the tribe.48 46. A. I. R. 1964 Pat. 201. On the Oraons, see Rov 1928.

47. Id., at 203.

48. A similar approach was applied by an Election Tribunal which had to decide

294

Identifying the Beneficiaries

In Wilson Reade v. C. S. Booth,*? an Election Commission had rejected

nomination papers for a Scheduled Tribe seat from a candidate whose

father was English and whose mother was a Khasi. In pre-Independence days he had accepted for himself and his children (his wife was a Khasi) privileges restricted to Anglo-Indians. But he was accepted asa Khasi by the tribespeople, the group being matrilineal and anyone born

of a Khasi mother being regarded as a member of the tribe; he had

followed “‘the customs and the way of life of the tribe,” was treated by them as one of themselves, and had been active in Khasi politics. The Assam High Court found that even though he was an Anglo-Indian within

the constitutional definition,S° this did not prevent him from

being a member of this tribe or some other community. Whether he was in fact a Khasi depended not on purity of blood but on his conduct and

on acceptance by the community.

More recently the Supreme Court, in Horo v. Jahan Ara,5? extended

this line of analysis to permit assimilation into a tribe by one without any tie of descent. A prominent Scheduled Tribe politician married a Tamil Christian woman in 1954. Upon his death in 1970 she sought to fill his seat in Parliament, from a constituency reserved for Scheduled

Tribes. Her nomination papers were rejected by the Returning Officer on the ground that she was not, as she claimed, a member of the Munda

tribe, since the status of a Munda could be acquired only by birth and not by marriage. In a careful assessment of the evidence, the Supreme Court concluded that although Mundas are normally endogamous and marriage with non-Mundas is often accompanied by excommunication

of the Munda partner, such marriages may be sanctioned and those excommunicated and their spouses admitted to the tribe upon the performance of certain ceremonies. Here there was abundant evidence whether a Konda Dora who had converted to Christianity at the age of ten “for the

purpose of his education” was a member of the Tribe and thus qualified to stand for a

reserved seat. There was no evidence that the convert had been excommunicated by the Tribe. Finding that converts observe the same customs and habits, intermarry, and are treated as members of the tribe, the Tribunal held that “mere acceptance of Christi-

anity is not sufficient to make him cease to be a member” (Gadipalli Parayya v. Boyina Rajayya, 12 E. L. R. 83 [1956]). 49. A. 1. R. 1958 Ass. 128.

50. Art. 366(2) defines an Anglo-Indian as “‘a person whose father or any of whose other male progenitors in the male line is or was of European descent but who is

domiciled within the territory of India and is or was born within such territory of parents habitually resident therein.” Some preferences for Anglo-Indians on a declining scale for a ten-year period were provided byy the Constitution (Arts. 331, 333, 336, 337). On the Anglo-Indian community, see Grimshaw 1959: 227-40, On this tribe, consult

Gurdon 1914. 51. A. I. R. 1972S. C. 1840.

Group Membership

295

that the marriage was accepted as valid and was approved by the elders of the community assembled in the Parha panchayat. Upon such acceptance she became a member of the tribe and was entitled to stand for the reserved seat.52 The situation in the Jahan Ara and Wilson Reade cases is the reverse of that in the Jasani, Shyamsundar, and Kartik Oraon cases. In the former, it is

the “‘new” group that is the politically privileged group rather than the “old” one. The question is not whether the new identificaiton precludes the old, but whether the old one precludes the new. In Wilson Reade and

Jahan Ara, as in the other cases, the courts looked to the views of the

privileged group to confirm the individual’s claimed membership. It was the Khasis who were entitled to special representation, and it was their acceptance of Reade that was determinative. Neither birth nor the possibility that the Anglo-Indian community might also have accepted him was considered relevant. Similarly, in Jahan Ara it was acceptance by the Mundas that was determinative. All of these cases permit overlapping and multiple group affiliations. The possibility that an individual might be accepted by a second group is not taken to automatically remove him from the first. D. TRIBE AND CASTE

The empirical approach of the Jasani case has not been applied in the problem of tribals attaining caste status. The case of V.V. Giri v. D. Suri Dora‘? arose out of an election to a seat in Parliament reserved for a

member of a Scheduled Tribe. The candidate was born a Moka Dora,

and his family had described itselfas such in all documents from 1885 to 1928. Since that time they had described themselves as Kshatriyas.54 There was evidence that the family had adopted Kshatriya customs,

celebrated marriages in Kshatriya style, was connected by marriage to

Kshatriya families, employed Brahmin priests, and wore the sacred

52. There is some uncertainty about the application of the Jahan Ara analysis in the Scheduled Caste setting. A provision to accord Scheduled Caste status to women who marry Scheduled Caste men was recommended by the Joint Parliamentary Committee on Revision of the Scheduled Castes and Tribes Orders, but was not included in the revision of the lists that was enacted in 1976. On the Scheduled Caste situation, see text at n. 255, below.

53. A. I. R. 1959S. C. 1318. Hereafter referred to as the Dora case. 54. Classical Hindu legal and social theory divides society into four great varnas (literally “‘colors””) or orders: Brahmins (priests and scholars); Kshatriyas (rulers and

warriors); Vaishyas (merchants and agriculturists); Sudras (menials). Varna distinc-

tions are influential but bear no direct relationship to existing social divisions (Srinivas 1962; chap. 3; Fox 1969).

296

Identifying the Beneficiaries

thread in the manner of Kshatriyas. His election was challenged on the ground that he was no longer a Moka Dora and was therefore ineligible to stand for the seat. Applying the tests set forth in the Jasani case, the Election Tribunal concluded that the candidate was no longer a Moka Dora, finding that “he has expressed unequivocal intention of drifting away from the clan, has totally given up feeling himself to be a member of the Moka Dora tribe and considers himself a Kshatriya.”55 Apparently, the candidate’s family. was one of a number of families of Mokasadars or large landowners who, according to the Tribunal, “would not like to be called

Moka Doras but considered themselves Kshatriyas.”56 The Tribunal

found support for its finding in the observation that “persons of [this]

type . . . who have drifted away from their old clan and renounced the tribal customs and manners and chosen to adopt the prevailing practices of the higher caste [sic] of the Hindu community could not be

entrusted with the task of representing the genuine grievances” of the

tribal communities; to do so would amount to a denial of the benefit of special representation conferred on the tribals by the Constitution.57

Since the Tribunal had found against the candidate on the question of his intention to remain a Moka Dora, it was not necessary to go into the question of acceptance by either the tribe or the Kshatriyas. So far the case, although reaching the opposite result, proceeded along the lines laid down in the Jasani case. However, the case took a radically different turn when it reached the High Court of Andhra Pradesh,5® which addressed itself, not to the question of whether he had remained a Moka Dora, but to the quite distinct question of whether he had become a Kshatriya. Starting with the principle that caste is a matter of birth rather than choice and that higher caste cannot be

gained,59 the court conceded that

it is possible that a member of a Scheduled Tribe may in course of time

adopt certain customs and practices in vogue among the Hindus, but in

order to bring them within the fold of Hinduism it would take generations.

Even if they came within the fold of Hinduism, [a] question would arise

55. 15 E. L. R. 1 at 38 (1957).

56. Id. 57. Id.

58. A. I. R. 1958 A.P. 724.

59. The cases cited by the High Court in support of this point (id. at 735) are readily distinguishable. Sahdeonarain v. Kusumkumari, A. 1. R. 1923 P.C. 21, and Chunku Manjhio.

Bhabani

Majhan,

A.1.R.

1946

Pat.

218,

are concerned

with

whether

tribals are

governed by Hindu personal law. Maharajah of Kolhapur v. Sundaram Iyer, A. 1. R. 1925 Mad. 497 is concerned with vama status for purposes of finding the applicable rule of inheritance.

Group Membership

297

whether they have formed a separate sect among themselves, or [whether] they would belong to the [Sudras] or to the twice-born class.

Having thus indicated the severe limitations of possible mobility, the court proceeded to lay down, as requirements for proving that such movement

had taken place, its version of the tests employed in the

Jasani case. These tests—intention, reaction of the old group, rules of the new group—are used in the Jasani case to test whether the indi-

vidual has remained in his old group. But here the court uses them as tests of assimilation to the new group. They become in effect a set of binding requirements, which must all be satisfied in order to prove a case of successful mobility. The High Court found no evidence of the reaction of the old tribe (which, by the Jasani approach, would have been irrelevant had he failed on the test of intention) and

no evidence as regards the reaction of the new fold, except that some of the

Kshatriyas recognize appellant as a Kshatriya. We can understand this if

this had been the result of generations, but the acceptance of the appellant

as a Kshatriya by one or two families would not . . . be sufficient.64

Since he thus failed to attain Kshatriya status, the High Court assumed that he therefore remained a Moka Dora and found him eligible for nomination to the reserved seat. The Supreme

Court, rather than reasserting the Jasani tests and

disengaging them from the High Court’s theories about caste mobility, took a third tack. Where the Election Tribunal had addressed itself primarily to the “intention” test laid down by Jasani, and the High

Court had insisted that all three factors mentioned in Jasani were required to prove mobility, the Supreme Court fixed its attention only on the third—and originally least important—of the Jasani tests: the reactions of the new group. The Supreme Court found the evidence insufficient to demonstrate that the candidate was a Kshatriya, since “the caste status ofa person in this context would necessarily have to be determined in the light of the recognition received by him from the members of the caste into which he seeks an entry.’”’6? Finding no evidence of such recognition, the

Court said “unilateral acts cannot be easily taken to prove that the claim for the higher status which the said acts purport to make is established.”63 The Court concluded that the candidate had not be60. A. I. R.

1958 A.P. at 735. The “twice-born” are the three higher varnas—

Brahmins, Kshatriyas, and Vaishyas—who have supposedly undergone a second or intellectual birth upon investiture with the sacred thread. The Sudra vama is only

once-born. 61. Id., at 736. 62. A. I. R. 1959S. C. 1318 at 1327.

63. Id.

298

Identifying the Beneficiaries

come a Kshatriya and had therefore remained a Moka Dora, eligible for the reserved seat. In spite of the similarity in outcome, the course of reasoning taken by the Supreme Court here is in sharp contrast to that in the cases discussed earlier. The Dora case agrees with them that neither birth nor mere intention is determinative of group membership; the conduct of the individual and the attitudes of the groups must be considered. But which groups? and their attitudes about what? In the cases discussed

above, when the question was whether X had, by joining a new group

B, ceased to be a member of privileged group A, it was the group A whose reactions were consulted—the Mahars in the Jasani case, the Samgars in the Shyamsundar case, the Oraons in the Kartik Oraon case. But when the question was whether a person had become a member ofa privileged group B, then the views not of the old group A but only of group B were pertinent—the Khasis in the Wilson Reade case. It was irrelevant what the Anglo-Indians might have thought of Reade, just as the views of the Mahanubhava Panth, the Arya Samaj, and the Christians received only a passing glance. Had the courts in these cases seriously considered acceptance by the non-privileged group as incompatible with membership in the privileged group, the cases would most probably have had different outcomes.

Following on these lines, one

would have expected the Supreme Court in the Dora case to address

itself to the views of the Moka Doras. But, like the High Court below,

they consider only the views of Kshatriyas. True, the Jasani case did mention “the rules of the new order” as one of the factors to be considered. But it is clear that this was not only the least important factor, but was intended to mean the rules of the new order respecting the retention by X of his membership in the old group. It was not the views of the Bs as to X’s membership in the Bs that counted, but their views as to his membership in the As. In the Dora case, when the Supreme Court consults the attitudes of the new group,

it is on the question of X’s membership in the Bs. The practical effect of these divergent approaches can be easily seen in tabular form (table 22). Let us imagine that X, a member of privileged Group A, has somehow aspired to membership in Group B. Table 22 merely restates the requirement of Jasani that to be an A (for the purpose of filling a 64. Another election tribunal dealing with a similar Mokasadar family reached the same conclusion by anticipating the approach followed by the Supreme Court. The

candidate remained a tribal since “there was no evidence that the Kshatriya community as a whole recognized him as belonging to their class” ( Gadipalli Parayya v.

Boyina Rajaya, 12 E.L.R. 93 (1956]). One may wonder who are “the Kshatriya community as a whole” —whether this is defined in terms of the locality, the district, the state, or all-India?

Group Membership

299

reserved seat) X must fulfill both tests I and I]. This occurs only in cases

1 and 2. It is clear that according to the Dora court the yes on test IIT in

case | would make the answer “no.” And in cases 4 and 6 and 8 the Dora

method would make him an A by virtue of his not being a B (i.e., failing test III), whereas the Jasani method would find him not an A because of

failing either test I (case 6), test II (case 4), or both tests (case 8).®

Since the Supreme Court never discusses the question of the candidate’s membership in the Moka Doras, one can only gather that there is implicit in the Court’s view a logical incompatibility between member-

ship in the two groups. Had his Kshatriya status been upheld, he would

ipso facto have not been a member of the Moka Doras for the purpose of standing for the reserved seat. Faced with the question of whether X remained an A, the Court addressed itself to the question of whether he

had become a B. But this course of reasoning is only plausible if it is assumed, first, that the two memberships exhaust the possibilities, and second, that they are mutually exclusive.

In assuming that they exhausted the possibilities the Court seems to deny the possibility that the candidate’s family had, although failing in some sense to become Kshatriyas, so separated themselves from the tribe as to lose acceptance as members. Such an intermediate possibility was considered by the Election Tribunal and to some extent by the TABLE

22

Contrastinc Resucts or DirFERENT Tests OF MEMBERSHIP Case No.

Test 1 Did X intend toremain an Aand so conduct himself?

1, 2.

yes yes

yes yes

yes

no

3.

4.

5.

6. 7.

8.

yes

Test II Did the As accept X _asan A?

Test IIT Did the Bs accept X asa B?

Jasani tests

Dora test

yes no

yes yes

no yes

no

no

yes

no no

yes no

no

yes

no

yes

yes

no

no

no

no no

yes no

Is X an A?

no yes

no

no no

no

no

yes

65. It should be noted that using the approach of the High Court in the Dora case the answer would be yes in every instance except for case 7 and then only upon the additional condition that the As accepted X asa B.

300

Identifying the Beneficiaries

High Court. Such splitting off is one of the classic and welldocumented methods by which new castes are formed.§7

The Jasani line of cases had allowed overlapping and multiple affilia-

tions. It was possible to be simultaneously a Mahar and a member of the Mahanubhava Panth, a Samgar and an Arya Samajist, a Christian

and an Oraon, an Anglo-Indian and a Khasi. Why is it not possible to

be both a Moka Dora and a Kshatriya? The Court does not indicate the source of its notion that these affiliations are mutually exclusive. But it

seems that this incompatibility is felt because Indian society is vis-

ualized as consisting of groups with unique corporate ranks in some definite rank ordering. Thus membership in one such group entails occupying such a rank and is inconsistent with membership in another group, which would mean simultaneously holding a lower rank in the

same system of ranks. The Court refers to the claim here as one for

“higher status,” which presumably cannot be achieved without giving up membership in the group with lower status. In part, this notion of Indian society as consisting of mutually exclusive groups ranked in a definite and unique order is a carry-over from

the area of personal law. The courts have long applied to members of

different religious communions their respective laws in matters of marriage,

divorce,

inheritance,

succession,

and

religious endowments.

Since in Hindu law there were some differences in the rules applicable

to the three higher varnas on the one hand and Sudras on the other, the

courts had from time to time to determine which rules were applicable

to particular persons or groups. For this purpose Hindu society was

visualized as if it consisted of four ranked compartments—the lowest being residual—and any of the actually existing caste groups would be assigned, if need be, to one of these theoretical compartments. Since 66. 1G E. L.R. at 38; A. I. R. 1958 A.P. at 735. 67. See e.g., Muthusami Mudaliar v. Masilamani, 1. L. R. 33 Mad. 342 (1909); Hutton

1961: 50 ff. 68. These differences are concisely summarized by Derrett 1958: 383-85. 69. In order to make such assignments, the courts evolved various tests: lists of

diagnostic customs (see e.g., Gopalv. Hanmant, I. L. R. 3 Bom. 273 [1879] where the tests of Sudra status are widow remarriage and admission of illegitimate sons to dine and

marry within the caste and to inheritance), or alternatively, tests of reputation (see ¢.g., Subrao v. Radha, 1. L. R. 52 Bom. 497 (1928), where it is held that vama depends on the consciousness of the caste as to its status and the acceptance of this estimate by other castes. According to the theory of vama, Hinduism comprised the four varnas and every

caste group could be assigned to one of these; caste and vara were co-extensive with Hinduism.

But departures from the symmetry of this scheme are found in many

instances where courts modified it to account for the actualities of the situations before them. Thus it is possible to have vara standing without belonging to a caste group (Sunder Devi v. Jheboo Lal A. 1. R. 1957 all. 215 [convert to Hinduism); Upoma Kuchain v. Bholaram 1. L. R. 15 Cal. 708 [1888] [daughter of outcaste]; cf. Ratansi v. Administrator

Group Membership

301

the Constitution, the courts have continued to make such varna assignments when necessary.”° Now that the various Hindu Code acts of 1955-56 have eliminated almost all of the instances in which vara might make a difference in applicable law,”! the courts can look forward to the day when they will no longer be faced with the task of making these imponderable and often fictitious varna identifications. In personal law cases the question was not whether an individual was

or was not a member of some existing social group, but whether he

should be assigned the status of one or the other varna.72 Ordinarily, the individual was indisputably a member of some actual caste or group, and

the proceedings took the form of determining the varna of this

group. It was assumed for this purpose that all caste groups could be assigned to one or the other of the vamas. Since the purpose of determining vama was to ascertain the appropriate rule of law, and since vamas clearly stood in a ranked order, the whole object of the proceeding was

to arrive at a unique determination of status.

The question before the courts in these election cases is quite different in kind. It is whether in fact a person is for a particular purpose to be

considered a member of some existing group. There is no necessary

relationship between membership in such a group and the holding of General, A. 1. R. 1928 Mad. 1279). Caste and vama may apply to persons who are not strictly Hindus (Inder Singh v. Sadhan Singh, 1. L. R. [1944] 1 Cal. 233 (Sikh Brahmins)). Caste groups have been recognized which have no vara and are not Hindu in any sense (Abdul Kadir v. Dharma 1. L. R. 20 Bom. 190 [1895]). Again, members of the same caste may hold different varna statuses (Subrao v. Radha, above).

70. After the advent of the Constitution the administration of separate personal law to the respective religious communities was challenged as discriminatory and Ultra vires

Art. 15. Although Art. 44 directs the eventual elimination of separate personal laws, the

continuing validity of disparate rules of personal law and the power of the state to create new rules applicable to particular religious communities has been upheld (e.g., Stateof Bombay v. Narasu Appa, A. 1. R. 1952 Bom. 84). The assignment ofa community toa sama has been held not to constitute a deprivation of rights to equality before the law, nor is it religious discrimination (Sangannagonda v. Kallangonda, A. 1. R. 1960 Mys. 147). 71. Le., the Hindu

Marriage Act of 1955, the Hindu Succession Act of 1956, the

Hindu Minority and Guardianship Act of 1956, and the Hindu Adoptions and Maintai-

nance Act of 1956. Derrett (1958: 383-85) suggests that the only instances in which

parna might continue to have effect are succession to sanyasis and determination of the maximum age for adoption. 72. The judicial treatment of the relation between varna and caste was plagued by

confusion, engendered in part by the use of “‘caste” to refer at once (1) to the four great classes or varnas into which Hindu society is theoretically divided by the Sanskrit

lawbooks, (2) to the multitude of existing endogamous groups or jatis, and (3) to

intermediate clusters of jatis known by a common name (cf. Fox 1969). The listings in

the Scheduled Castes Order and the various state lists of Backward Classes are a mixture of the second and third types.

302

Identifying the Beneficiaries

vara status.73 Apparently, the candidate in the Dora case—and his family and possibly the whole group of Mokasadars—claimed to be Kshatriyas. It is unclear whether they were merely asserting Kshatriya vara status or whether they were claiming membership in some particular endogamous group of Kshatriya families. It would seem possible to achieve varna status without necessarily becoming effective members

of a caste with that status.” In any event, it is unclear whether attain-

ment of the vama status claimed in the Dora case would have been felt by

the tribe to be incompatible with continued membership. Whatever

theoretical incompatibility there may be in belonging to two varnas, it is

not impossible to be accepted as a member of two actual social groups. But varna theory is not the sole source of the Supreme Court’s notion of mutually exclusive group membership.’5 This notion is supported by its picture of Indian society. The Court indicates that it bears in mind “the recognized features of the hierarchical social structure prevailing amongst the Hindus” and the “inflexible and exclusive nature of the caste system.”76 The Court is, of course, only giving its view of the conditions that obtain; it expresses the hope “that this position will change, and in course of time the cherished ideal of a casteless society ... will be attained.”77 Nevertheless, in its anxiety not to be “un-

realistic and utopian,” the Court seems unnecessarily to give currency to the view that all groups in Indian society are ranked in some unique and definite order.” Thus the Court is impelled to formulate a general 73. Cf. Mulai v. Lal Dan Bahadur Singh, 9 E. L. R. 9 (1952), where an Election Tribunal

found that the proclamation ofa former prince had not transformed the Gonds of Rewa State into Kshatriyas. “But even if the Gonds . . . could be deemed to be Kshatriyas, they would not cease to be members of Scheduled Tribes.” Even “their aristocratic sub-division known as Raj Gonds, still continue to be Gonds and . . . belong to the Scheduled Tribes.” 74. See n. 69, above. 75. Ironically, the Supreme Court judge who delivered the opinion in Dora was aman of outspoken reformist views who opposed caste distinctions and had in his writings eloquently denounced varna determinations as a blot on the administration of Hindu personal law (Gajendragadkar 1951:516 ff.) The continuing pull of ama notions on judges of the most austerely anti-hierarchic views is displayed in Justice Krishna lyer’s suggestion that untouchables are those “outside the four-fold Hindu division” (State of Kerala v. N. M. Thomas, A. 1. R. 1976 S. C. 490 at 535). 76. A. 1. R. 1959 S. C. at 1327.

77, Id.

78. This tendency to picture Indian society as a series of graded corporate ranks seems congenial to the courts, perhaps because of the felt necessity of having some conceptual means for reducing the immense variety of Indian society to terms which could be applied without extensive investigation in each individual instance. Cf. Marriott’s suggestion that urban and educated Indians (and foreigners) tend to conceive of caste in terms of criteria which constitute or imply a scale of Hindu ritual values, rather than according to the structure of interaction among various groups (Marriott

1959: 104).

Group Membership

303

staudard for assigning standing in such a rank ordering. Noting that whatever may have been the case in ancient times, “status came to be based on birth alone,” the Court says “‘it is well known that a person who belongs by birth to a depressed class or tribe would find it very

difficult, if not impossible, to attain the status of a higher caste amongst

the Hindus by virtue of his volition, education, culture and status.”’79

Thus Hindu society is not only hierarchic but inflexible as well. If the Court meant literally that caste status was determined by birth alone, it would of course, be redundant to consider the views of any group in order to determine it. But the Court apparently means that birth is determinative in the first instance and that this can be varied—not by an individual’s conduct but only by the unanimous recognition of his

claims by members of the higher group. J. L. Kapur, J., dissenting, vigorously rejected the primacy of birth and put forward another general theory of assignment of rank. Holding that caste varies as a consequence of the gunas, karma, and subhavana and is dependent on actions, he found that the candidate had “‘by his action raised himself to the position of a Kshatriya and he was no longer a member of the Scheduled . . .Tribe... .”8° Either of these general theories may prove embarrassing. Acceptance of the dissenting judge’s theory that caste (or varna) status may be gained wholly by individual action could expose the courts to a torrent of litigation in which they would be faced with the necessity of setting up legal tests for caste standing and assigning it to individuals and communities.®1 The majority’s acceptance of birth as the primary determinant of group membership avoids this difficulty and is no doubt accurate in the overwhelming majority of instances. But by making membership in the old group dependent on failure to achieve the purported membership in the new group, the majority’s theory may disincline the courts from giving legal recognition to existing patterns of mobility, which ordinarily involve a period of conflicting claims and overlapping 79. A. 1. R. 1959S. C. at 1327. 80. Id., at 1331. Gunas, etc., means roughly material nature, deeds, and temperament.

R. Dhavan (1977: xxix) points out that this is an Arya Samaj theory of caste.

81. Of course, such litigation occurs without judicial encouragement. See ¢.g.,

Sankaram Namboodri v. Madhavan, A. 1. R. 1955 Mad. 579, where one section ofa family of

Embrandiri Brahmins changed its name and claimed that they were Nambudiri

Brahmins and therefore governed by the special personal law of the latter. The court

was willing to concede that had they voluntarily effected a partition in accordance with Nambudiri law, this would help establish their status as Nambudiris for purposes of personal law in the future, but held that the change of name was a mere unilateral

declaration of no effect on their status in the absence of evidence of renunciation by the whole family of their present status or of their acceptance by the Nambudiri community asa whole. .

304

Identifying the Beneficiaries

identifications. Successful separation from an old group may be overlooked, with the result of imposing on a privileged group a candidate

who is not an accepted member of it. More generally, existing channels of mobility may be discredited. For if acceptance by a new group removes one from the old, the Hinduization of tribals and the formation

of new sects would be accompanied by the danger of disqualification for receipt of preferences. Since the system of preferences is designed to increase flexibility and mobility within Indian society, there seems little reason to make abandonment of older and slower methods of mobility a

condition for the utilization of the new ones. The dilemma posed by these opposing theories can be solved by

eschewing any general theory of assignment and deciding questions of group membership by the pragmatic approach of the Jasani case. Since this permits

multiple

identifications,

the courts

would

not pose for

themselves the kind of either/or puzzles that have no satisfactory ans-

wer. They would be concerned only with whether for the purpose of the particular measure the individual concerned ought to be counted a member of the privileged group.®? As far as it is compatible with the particular legislative policy, membership would depend entirely on the

voluntary affiliation of the individual®3—as confirmed, where neces-

sary, by voluntary acceptance by the group’s members. Consonant with the constitutional principles of freedom of association and the auton-

omy of social groups, State imposition of standards of membership would be minimized,® and judicial determination of the relative status of groups would be eliminated. 82. In Jankilal v. Jabarshingh, A.1.R. 1957 Nag. 87, the court found that “‘a person can be a Hindu and also be deemed a member of an aboriginal tribe” for the purposes of the provisions of the Central Provinces Land Alienation Act, which outlawed conditional sale provisions in mortgages made by listed tribes. 83. A right to benefits is not dispelled by casual claims to “forward” status. Where

some applicants “have chosen to mention [in application forms] that they belong to the forward community, while making it clear that they were Nadar Christians from

Kanyakumari Dt., belonging to the Church of South India . . . the classification of the

petitioners as backward

did not depend

upon what they stated in the application

against the relevant column, but whether they had mentioned in the application that they belonged to the Church of South India. Once that appeared on the record, it would

be obvious that they belong to the backward class . . . .”" (Susila v. State of Madras, A. 1.

R. 1970 Mad. 399).

.

84. It has been argued with some persuasiveness that the Constitution withdraws all governmental power to determine whether an individual is a member of a particular caste. ‘Can a secular government force a citizen to belong or not to belong to a particular caste?” queries the chairman of the Backward Classes Commission in considering whether census clerks may put down the caste ofan individual according to their conception of it or whether the individual’s conception is determinative. The chairman contends that if caste is to be a voluntary affiliation, government must refrain from assigning it (Report of the Backward Classes Commission 1956: I, xviii). But once

Group Membership

305

E. RELIGION The Constitution is openly and determinedly secular.85 Freedom of

religion is guaranteed.® Religious discrimination on the part of the State is forbidden.8? The courts have been vigilant in invalidating

governmental measures framed along religious lines.®8 Nevertheless, in some instances religion has been made a qualification for preferential

treatment. The President’s 1950 Order specifying Scheduled Castes provided that “no person professinga religion different from Hinduism shall be deemed a member of a Scheduled Caste.’’89 In 1956 this was broadened to include Sikhs.9° Who meets this religious qualification? The legal definition of Hinduism, developed for the purpose of applying appropriate personal

law, was neither a measure of religious belief nor a description of social behavior as much as a civil status describing everyone subjected to the application of “Hindu law” in the areas reserved for personal law.91

caste and tribe have been accepted as appropriate units for the distribution of prefer-

ences, some governmental determinations, in order to prevent abuses, are unavoidable.

It would seem possible to have more or less objective standards, even for determining a

voluntary affiliation and to refrain from any assignments other than the determination of whether or not an individual is within the group to which he claims to belong. 85. On Indian secularism generally, see Smith 1963; Luthera 1964; Seminar 1965; Smith1965; Sharma 1966; V.K. Sinha 1968; Derrett 1968; Galanter 1972. Fora

discussion of the problems treated in this section, see Smith 1963:322 ff.

86. Art. 25, 26, 28, 30(1).

87. Art. 15, 16, 29(2), 30(2), 325.

88. State of Rajasthan v. Pratap Singh, A. 1-R. 1960 S.C. 1208; Nain Sukh Das v. State of

U.P., A.1.R. 1953 S. C. 384; State of Jammu and Kashmir v. Jagar Nath, A. 1. R. 1958J. &

K. 14. But cf. n. 70, above, and n. 113, below.

89. Constitution (Scheduled Castes) Order, 1950, Para. 3. An exception was included for Sikh members of 4 of the 34 Scheduled Castes listed for the Punjab. Cf. the Government of India (Scheduled Caste) Order, 1936, para. 3, which provided that

“[nJo Indian Christian shall be deemed a member of a Scheduled Caste.” The Con-

stitution (Scheduled Tribes) Order, 1950, contains no religious provision.

90. The Scheduled Castes and Scheduled Tribes Orders (Amendment) Act, 1956 (63 of 1956), para. 3. The main provision now reads “‘No person who professes a religion different from the Hindu or Sikh religion shall be deemed a member of a Scheduled Caste.” The inclusion of Sikhs has an interesting history going back to dissatisfaction with their exclusion from the Scheduled Castes under the 1936 Order. In the Con-

stituent Assembly, Sikhs gave up their demands for political safeguards for minorities in return for assurances that their backward classes would be recognized and listed as

Scheduled Castes

(there was some opposition

to this inclusion from among the

Scheduled Castes). See the speech of Sardar Hukum Singh at IX CAD 235; remarks of Sardar V. J. Patel, 1X CAD 247. Sardar Hukum Singh moved an amendment to this

purpose in the Constituent Assembly, but it was withdrawn with the understanding that the Sikhs would be considered in the making of the lists of Scheduled Castes (VII

CAD 552 ff.).

91. Or, more accurately, all who would be subject to Hindu law in the absence of Proved special custom or of a contingency such as marriage under the Special Marriage

306

Identifying the Beneficiaries

Heterodox practice, lack of belief, active support of non-Hindu religious groups,°? expulsion by a group within Hinduism93—none of these

removed one from the Hindu category, which included all who did not

openly renounce it or explicitly accept a hostile religion. The individual could venture as far as he wished over any doctrinal or behavioral borders; the gates would not shut behind him if he did not explicitly adhere to another communion.™ The same negative definition remains today for purposes of application of personal law. In Chandrasekhara Mudaliar v. Kulandaivelu Mudaliar,5 the Supreme Court had to decide on

the validity of a consent to adoption by one who disavowed belief in the

religious efficacy of adoption, in Hindu “atma”

rituals and scriptures, in

[soul] and salvation. The Court found that “the fact that he

does not believe in such things does not make him any less a Hindu. . . . He was born a Hindu and continues to be one until he takes to another religion . . . whatever may be his personal predilections or views on Hindu religion and its rituals.” In the post-constitutional cases involving preferences the same latitudinarian conception of Hinduism has been carried over from the area of personal law. To “profess” Hinduism merely means to be a Hindu by birth or conversion.27 Unorthodox practice or lack of personal belief in its tenets does not mean lack of profession for this Act (IIT of 1872, now 43 of 1954). But persons who are not Hindus may attraet-the

application of Hindu law. These include some Christians (see e.g., Chinnaswammy v.

Anthonyswamy, A.1.R. 1961 Ker. 161. Cf. n. 36, above). Tribals may be subject to Hindu law although they are not Hindus. See Mira Devi v. Aman Kumari, A. 1. R. 1962 M.P. 212.

Hindu personal law was applied to some Muslim groups until the passage of the Muslim Personal Law (Shariat) Application Act (XXXVI of 1937). 92. Bhagwan Koer v. Bose, 30 1.A. 249 (1903).

93. Ratansi D. Morarji v. Admr. General of Madras, A. 1. R. 1928 Mad. 1279. 1283. 94. The concept of “Muslim” is treated somewhat differently. There are expressions of the same negative (no conversion) test. See Bhagwan Bakhsh Singh v. Drigbijai Singh, 1. L. R. 6 Luck. 487 (1931). Generally it is held that adherence to some minimumof beliefs

(the unity of God, the mission of Mohammed as his prophet, and the authority of the Koran) is necessary and sufficient to make one a Muslim (Narantakath v. Parakkal, 1. L. R. 45 Mad. 986 [1922]; Jiwan Khan v. Habid, 1.L.R. 14 Lah. 518 [1933}). Repudiation of

these beliefs, even without conversion to another religion, makes one not a Muslim (Resham Bibi v. Khuda Bakhsh, 1. L. R. 19 Lah. 277 [1938}). Compare the test for Sikhs, n.

100 below.

95. A. I. R. 1963S. C. 185. 96. Id., at 200. See also the definition of “Hindu” set out in the Hindu Marriage Act, 1955, Section 2, and discussion in Derrett 1963, §§17-20. A similar latitudinarianism

may be observed in the tests for whether a tribe is sufficiently Hinduized to attract the application of Hindu law. Orthodoxy is unnecessary; it is sufficient that the tribe acknowledge themselves as Hindus and adopt some Hindu social usages, notwithstanding retention of non-Hindu usages (Chunku Manjhi v. Bhabani Majhan, A. 1. R. 1946 Pat. 218), 97. Michaelv. Venkataswaran, A. 1. R. 1952 Mad. 474.

Group Membership

307

purpose—one may eat beef and deny the authority of the Vedas. In effect the test seems to amount to a willingness to refrain from calling oneself something else. Thus where the election to a reserved seat of an

active supporter of Dr. Ambedkar’s Buddhist movement® was challenged on the ground that he was not a Hindu, the court found that “it has to be established that the person concerned has publicly entered a religion different from the Hindu . . . religion.”99 Mere declarations falling short of this would not be sufficient. The candidate had supported the movement for mass conversion by serving on the reception committee, editing a newespaper supporting the movement, and attending a rally where an oath, “I abandon the Hindu religion and accept the Buddha religion,” was administered by Dr. Ambedkar. When those who wished to convert were asked to stand, the candidate stood. But there was no evidence that he did in fact take the oath; the

court held that in the absence of such a declaration, he remained a

Hindu.1 The same test of public declaration was subsequently upheld by the Supreme Court in Punjabrao v. Meshram,1 where there were eyewitnesses that the candidate did convert to Buddhism at a mass ceremony. In spite of their political enmity to the candidate, the Court finds the testimony of these witnesses credible in view of the fact that the candidate was a prominent follower of Dr. Ambedkar and “‘it would be highly improbable that [he] would have remained aloof” from the Buddhist

movement.?°2 This direct evidence was corroborated by a later written declaration and by certain acts and omissions by the candidate. The “strongest circumstance” corroborating the evidence of the conversion is that as chairman ofa local committee, the candidate presided over the

conversion ofa small Shiva temple into a Buddhist temple. “(H] owever

great the admiration or regard that a Hindu may have for Lord Buddha, he would shudder at the idea of desecrating a Shiva Linga in this manner or even of converting what was once a Shiva Temple intoa 98. On this Buddhist movement, see text at n. 163, below.

99. Karwadi v. Shambharkar, A. 1. R. 1958 Bom. 296 at 297. 100. /d., at 299. The vagaries of the declaration test are illustrated in Rattan Singh v. Devinder Singh, 7 E. L. R. 234 (1953), 11 E. L. R. 67 (1955), where the candidate had at various times described himself as a Mazhabi Sikh, a Harijan Hindu, a Balmiki, and a Balmiki Hindu. The Tribunal, holding that the minimum qualification for being a Sikh is willingness to declare “I solemnly affirm that I believe in the ten gurus and that I have

no other religion,” found him to be a Balmiki Hindu in 1953 and a Mazhabi Sikh in 1955. Any objective evidence was rigorously excluded since “the question of ... religion . . . is a matter of personal faith and cannot be the subject of any evidence ofa third party.” 101. A. I. R. 1965S. C. 1179.

102. /d.,‘at ‘1181: The witnesses, who were also Buddhist converts, were members ofa tival faction within Dr. Ambedkar’s Republican Party (id., at 1180).

308

Identifying the Beneficiaries

Buddhist Temple.” Second, the candidate willingly married his daughters to Buddhists, upon which the Court reflects that “‘ifhe werea Hindu . . . it is unlikely that he would have reconciled himself with the idea of giving his daughters in marrige to non-Hindus, more particularly when the bridegrooms’ side insisted on following the Buddhist ritual.” 1% Third, the wedding announcements substituted a picture of the Buddha and an invocation of his blessing for the usual picture and invocation of the Kuladaivata (household deity). Rejecting the candidate’s contention that he was treating Buddha as “‘the 1 Ith incarnation,” the Court observed that he was not of that “‘sophisticated” class that have discarded altogether the picture and the blessing, and “had [he] considered himself to be a Hindu, he would have followed the usual practice.””105 All of the above indicate to the Court that a conversion did in fact occur. The test of conversion is public declaration. Entry into another religion must be so public “that it would be known to those whom it may interest.’’°6 Therefore, if one publicly declares that he has ceased to belong to his old religion and has embraced a new one, he will be accepted as professing the new one. There is no religious test for membership in the new group beyond the declaration. “In the face of such an open declaration it would be idle to enquire further whether the conversion

A decade

. . . was efficacious.””107

later, in its next encounter with the Buddhist issue, the

Supreme Court coupled a more latitudinarian view of Hindu practice with a condescending if sympathetic view of the Buddhist movement. In Ganpat v. Presiding Officer,* the eligibility of several candidates for a Scheduled Caste seat in the Maharashtra Assembly was challenged on grounds of their conversion to Buddhism. The challenged candidates claimed to be Hindus, and the question was whether this claim was belied by other evidence. For example, the alleged departures of one candidate from Hindu practice consisted of his being listed on a distinctively Buddhist invitation to his brother’s daughter’s wedding and 103. Id., at 1183.

104. Id., at 1182.

105. fd., at 1182. The incarnations (avataras) of Visnu are commonly accounted to

number ten, Buddha being the last to have appeared. However, in general Buddha

receives little attention from Vaishnavite Hindus. Apparently, the candidate here either had a different list or was unaware that Buddha was already included.

106. Id., at 1184.

107. Id., at 1184, The formalities of conversion are to be tested by the new religion

itself, not by external standards. In the case of these Buddhists it appeared that officiation by a Bhikku (priest) was not required and recitation of three vows and five Precepts was considered sufficient (see id. at 1184). On the problem ofa declaration test,

see n. 100, above.

108. A. I. R. 1975 S. C. 420..

Group Membership

309

participating in the wedding, during which pictures of Dr. Ambedkar

and Buddha were garlanded, a Buddhist Bhiku officiated, and mantras

were recited in Pali. The Court was unpersuaded that these instances

established that the candidate, a born Hindu, had ceased to profess

Hinduism.

The Court’s skepticism is based in part on its view that religious

identification is ‘‘a highly personal matter’ and that “open assertion,” especially by an educated person, should be given greater weight than the interested testimony of others based on “stray instances.” 1 But it also finds support in the Court’s views of Hinduism and of the Buddhist

movement. The Court remains unpersuaded that there is any signifi-

cant departure from Hinduism here.

Hinduism is so tolerant and Hindu religious practices so varied and eclectic

that one would find it difficult to say whether one is practicing or professing Hindu religion or not. Especially when one is born a Hindu the fact that he goes to a Buddhist temple or a church or a durgah cannot be said to show that they are no more Hindus unless it is clearly proved that they have

changed their religion from Hinduism to some other religion.11°

Hinduism’s tolerance of syncretism combines with the persistence of

custom to make it difficult to abandon. “[I]fa certain community in a spirit of protest says they would like to give up Hinduism and adopt Buddhism it is not likely to make much change either in their beliefs or in their practice.” If in circumstances of extreme oppression, “some members of the Scheduled Castes in their protest resort to desperate measures

to erase the indignity of untouchability one cannot blame

them. But whether it produces any result is a different question however well meaning such efforts may be.””!11 Hindu tolerance of Buddhist practices raises the question of whether Buddhism is “‘a religion different from the Hindu . . . religion” within the meaning of the Scheduled Castes Order. Converts to Christianity

and

Islam are, of course, non-Hindus.112 But Buddhists, Jains, and

Sikhs are treated

as Hindus

for many

purposes.1!3

Hindu

is an

109. fd., at 425. 110. Id., at 424. 111. Td., at 424. “112. Michael v. Venkataswaran, above n. 97. But one may remain a Hindu even while belonging to a religious sect that includes non-Hindus among its members (State of Bombay v. Yagna Sastri Purushadasji, 61 Bom. L. Reporter 700 [1960]). 113. These groups are Hindus for purposes of personal law. See, e.g., Hindu Marriage

Act, 1955, §2. They are specifically included among Hindus for purposes of authorizing the state to open Hindu temples to untouchables. See the explanation II to Art. 25(2) of the Constitution. But their separateness is recognized in some contexts; e.g., Jains were not of “the same religion” as Hindus for the purposes of temple-entry legislation (State v. Puranchand, A. 1. R. 1958 M.P. 352; Devarayiah v. Padmanna, A. I. R. 1958 Mys. 84).

310

Identifying the Beneficiaries

equivocal term, sometimes used with reference to adherents of more or

less “orthodox” vedic and brahmanical communions and at other times used to embrace the full array of “heterodox” sects, including Sikhs,

Jains, and Buddhists. Yet Jains and Buddhists are considered non-

Hindus for purposes of preferences, and Sikhs are mentioned separately.114 The candidate in Meshram argued that “Hindu” in the Scheduled Caste Order should be given a broad meaning that would include Buddhist, in line with much other official usage. The Supreme Court finds this clearly negatived by the language of the Order: the separate mention of Sikhs would otherwise be redundant. This reading

of the Order clearly accords with legislative intention to exclude the major “heterodox” communions. But the Court’s observation that “the

word ‘Hindu’ is used in the narrower sense of the orthodox Hindu religion, which recognizes castes and contains injunctions based on caste distinctions,”115 is troubling. It comes close to suggesting a positive definition of Hinduism which goes beyond merely excluding Sikhs, Jains, and Buddhists to the point of throwing doubt on the inclusion of

all sects which in one fashion or another repudiate caste distinctions in

doctrine or practice. This is particularly discomfiting because the Court, while inquiring what

it means

to “profess”

a religion

“different

from

the Hindu,”

reverses the question to examine what it means to profess Hinduism. It reads the Order as requiring that to be treated as a member of a Scheduled Caste ‘‘a person ... must be one who professes either Hindu or Sikh religion.”116 And it notes, ‘‘the word ‘profess’ in the Presidential Order appears to have been used in the sense of open declaration of practice by a person of the Hindu (or the Sikh) religion.””117 The Court is obviously not attempting to establish a positive religious test of eligibility for preferences. But the unfortunate juxtaposition of these dicta on Hinduism and its profession with the evidentiary examiDifferences in regard to the organizatin of religious trusts may be recognized by having

a Hindt religious trusts law which covers Jains and Buddhists but not Sikhs (Moti Das v.

S. P. Sani, A. 1. R. 1959 S. C. 942).

114. They were excluded until 1956, see note 90 above. (Gurmukh Singh v. Union of India, A.1. R. 1952 Pun. 143; Rattan Singh v. Devinder Singh, n. 100, above). 115. A. I. R. 1965 S. C. at 1184. The Supreme Court has more recently enunciated views of Hinduism which seem to conflict with its description here. In Sastri Yagnapurushdasyi v. Muldas Bhundardas Vaishya, A. 1. R. 1966 S. C. 1119, it was held that

the Swaminarayan sect was, in spite of its protestations, part of Hinduism and subject to Bombay's temple-entry act. The Supreme Court observed that the sectarians’ belief in caste pollution “is founded on superstition, ignorance and complete misunderstanding of the true teachings of the Hindu religion. . . .” 116. A. TR. 1965 S.C. at 1184.

117. d., at 1184.

Group Membership

311

nation of conventional Hindu practice and attitude with regard to intermarriage, invitations, and respect for idols comes uncomfortably close to suggesting that these Hindu proprieties are a requirement for preferences. Not only converts, but anti-caste sectarians, atheists, and

militant iconoclasts would feel the pinch of such a test. It would not only be extremely difficult to apply, but it would no doubt violate the understandings of many Hindus, impede innovation and change within Hinduism, and present serious problems of religious discrimination and freedom of religion.178 In spite of the Court’s remarks, it is clear that the Scheduled Caste Order itself does not establish or sanction such a positive religious test. It does not require anyone to profess Hinduism, much less practice it. It merely requires that he does not profess a different religion. Professing Hinduism and professing a non-Hindu religion may be mutually exclusive categories, but they are not exhaustive.119 The question is whether one professes a non-Hindu religion; the profession and practice

of Hinduism need be considered, as they are in Meshram, only for their

evidentiary value in this inquiry.

It is clear, than, that Buddhists are excluded by the presidential Order. The Meshram Court merely notes that when a person “has ceased to be a Hindu he cannot derive any benefit from that Order.””120 The Court never reaches the question of whether the “Hinduism” test for recipients of preferences infringes the constitutional ban on religious discrimination by the State.124 118. While the Order’s religious distinction itself is open to these objections, it might

conceivably correspond

to some difference in conditions. Any relevant differences

between those who observe Hindu proprieties and those who do not is even more implausible. 119. One

wonders

whether someone

who repudiated

Hinduism,

or an avowed

athiest, would be held to “profess a religion other than Hinduism.” Cf. Chandrasekhara Mudaliarv. Kulandaivelu Mudaliar, A. 1. R. 1963 S.C. 185. 120. A. I. R. 1965 S. C. at 1184. 121. The failure to reach the constitutional issue was due to the preoccupations of

Meshram’s counsel in the High Court. The Election Tribunal had decided the factual question of a conversion against the candidate, much as the Supreme Court did later.

On appeal to the High Court, Meshram’s counsel never raised the question of the

constitutionality of the religious classification. Instead, counsel were eager to argue that

Buddhism was in fact not a religion different from Hinduism. The High Court held that this was a factual matter which should have been pleaded and proved by evidence and refused to allow the argument to go forward. Assessing the evidence, the High Court

found no evidence of authority to eflectuate a conversion to Buddhism and a predominant political rather than a religious motive. The court concluded that the ceremony was a protest against caste and demonstrated petitioner’s sympathy with Buddhism but was not tantamount to the requisite profession of non-Hindu religion. On appeal, the Supreme Court re-assessed the evidence once more and re-instated the view of the

Election Tribunal that the conversion was effectuated.

312

Identifying the Beneficiaries

The constitutional challenge had been raised in several earlier cases. In Gurmukh Singh v. Union of India,22 a Bawaria Sikh protested his exclusion from the Scheduled Castes, in which the presidential Order had included Hindu Bawarias. A Full Bench of the Punjab High Court

found that Article 341 empowered the President to select those “parts of castes” which he felt should be included and that he could select those

parts on the basis of religion without violating Article 15(1), the ban on religious discrimination, since Article 15(4), which authorized prefe-

rential treatment for the backward, operated as an exception to the prohibitions of 15(1). The Court conceded that Scheduled Castes were

to be designated on the basis of their backwardness. But, finding that

the Constitution vested in the President the entire power to make such

determinations, the Court refused to review his order by considering whether the Sikh Bawarias were in fact sufficiently backward to be included. In Michael v. Venkataswaran,!23 the religious requirement was upheld against a Paraiyan convert to Christianity who wished to stand for a reserved seat. Even if there are cases in which both the convert and his caste fellows consider him as still being a member of the caste, the court found, “the general rule, is [that] conversion operates as an expulsion from the caste . . . a convert ceases to have any caste.””!24 The presidential Order, according to the court, proceeds on this assumption and takes note of a few exceptions. 125 The court declined to sit in judgment on the President’s determination that similar exceptional conditions do not prevail in other instances. Thus the presidential Order was upheld not because ofan absence of judicial power to review it but because ofits accuracy in the general run of cases. In In re Thomas,'26 another bench of the Madras Court considered a

convert case which did not involve the presidential Order. The Madras Government

had extended school-fee concessions

to converts from

Scheduled Castes “provided .. . that the conversion was of the .. . student or of his parent.” A Christian student whose grandfather had converted could not, it was held, complain of discrimination on grounds of religion. The basic criterion was not religion but caste, and converts did not belong to the Scheduled Castes. By conversion they had “ceased to belong to any caste because the Christian religion does not recognize 122. A. I. R. 1952 Punj. 143.

123, A. I. R. 1952 Mad. 474. The continuing efforts of Christian groups to secure the extension of benefits to recent converts and to established Christians are usefully

compiled in National Council of Christian Leaders (1978).

124, Id.,at 478. 125. Le., the four Sikh groups listed in the 1950 Order. See n. 90, above. 126. A. I. R. 1953 Mad. 21.

Group Membership

313

a system of castes.’’127 The concessions to recent converts were merely an indulgence, and the State could determine the extent of this indulgence. These cases indicate the three distinct grounds on which the religious disqualification has been upheld: (1) the inappropriateness of reviewing the President’s Order; (2) the constitutional authorization for using religious criteria in the designation of beneficiaries; (3) the theory that no religious discrimination is involved, since non-Hindus have no caste. 1. Is the Scheduled Castes Order unreviewable? Executive action, even in pursuance of expressly granted and exclusive constitutional

powers, is not immune from judicial review for conformity with constitutional guarantees of fundamental rights. The special status of the

Order as an act of the President is no bar (if it ever was),}28 for since

1956 the religious requirement has been promulgated by Parliament. In any event, the posture of the Gurmukh Singh and Michael cases was one of judicial restraint rather than judicial powerlessness. 129 Is such restraint appropriate? Since those cases were decided, judicial power to review i government’s designation of beneficiaries of preferences has been firmly established. Since 1959 the courts have subjected the standards used by state governments to designate backward classes to close and

detailed scrutiny.43° There is no indication that the delineation of

Scheduled Castes by the President and Parliament enjoys an immunity from similar judicial scrutiny where challenged as contravening funda-

mental rights. 31

127. Id., at 22. The exclusion of Buddhists from the preferences for Scheduled Castes has been similarly justified. See discussion below in text at nn. 163 ff

128. Art.12, 13. In Karkare v. Sheode A. 1. R. 1952 Nag., 330, the court found that the

immunity conferred on President and state governors by Art. 361 “does not place the actions of the Governor purporting to be done in pursuance of the Constitution beyond the scrutiny of the Courts. . . . Unless there is a provision excluding a particular matter

from the purview of the Courts [as in Arts. 122, 212, 263, 329 (a)] it is for the Courts to examine how far any act done in pursuance of the Constitution is in conformity with it.” But cf. Biman Chandra v. Governor, A. 1. R. 1952 Cal. 799, holding that Art. 361 removes

the acts of a governor from judicial review unless there is evidence of dishonesty or bad faith. 129. Art. 12 provides that for purposes of the Fundamental Rights provisions of the Constitution, ‘the State’ includes the Government and Parliament of India and the Government and the Legislature of each of the States and all local or other authorities. . . .”” In Gurmukh Singh the court concedes that the President is included and that therefore his action is governed by the requirements of the chapter on fundamental rights.

130. See chaps. 7 and 8, above. 131. See Art. 12, n. 129 above. The Hinduism requirement is an expression of the

power conferred by Art. 341 to select “‘castes, races or tribes or parts of or groups within

castes, races or tribes” (emphasis added ). Is the power of the President and of Parliament (which has exclusive power to modify the list) subject to review on the

314

Identifying the Beneficiaries

2. Article 15(4) permits special provisions in favor of Scheduled

Castes

notwithstanding

the provisions of Article

15(1), which

ban

discrimination by government on grounds of caste or religion. Does Article 15(4) authorize the use of religious criteria in selecting the Scheduled Castes? Caste and religion are normally forbidden bases of classification. 132 Constitutional authorization for special provisions to

disadvantaged groups has been held to authorize the use of such other-

wise forbidden classifications.433 However, their use is still subject to

the standards applicable to any governmental classification of citizens.

The classification “must be founded upon an intelligible differentia

which distinguishes persons or things that have been grouped together

from those left out . . .and . . . the differentia must havea rational rela-

tionship to the object sought to be achieved.’’!4 In selecting Backward Classes,caste units may be employed, but caste standing may be used as

a criterion only to the extent that it is useful in identifying social back-

wardness and cannot be the sole or dominant test for that purpose.95 Religious groups, too, may presumably be the units used in designating

Backward

Classes,

but

it is clear that

religion

cannot

be the sole

criterion of their backwardness.136 Non-Hindus cannot be made the recipients of preferences on the basis of purely religious classification in

the absence of other evidence of backwardness. One wonders why they can be excluded from preferences solely on the basis of religion in spite

of the presence of such evidence. Perhaps the State’s power to define

Backward Classes is not as broad as the President’s (and Parliament’s) to define Scheduled Castes. And perhaps the relevance of religion to

untouchability, notwithstanding its inappropriateness in marginal cases, is more apparent than to defining backwardness. But it cannot be

question of whether the delineation of “parts” or “groups” is reasonably related to the object of the clasification? There is no indication that this power is exempt from such review. See Art. 13(2) and (3). 132. On the religious classification, see cases noted at 88 above. On caste classifica-

tions. see Sanghar Umar v. State, A. 1. R. 1952 Saur. 124; Bhopalsingh v, State, A. 1. R. 1958 Raj. 41; State of Madras v. Champakam Dorairajan, A. 1. R. 1951 S. C. 226. 133. The power to designate “classes” in Art.

15(4) and Art. 16(4) operates as an

exception to the prohibition on the use of all grounds of classification in Arts. 15(1) and 16(2) respectively. E.g., Ramakrishna Singh v. State of Mysore, A. 1. R. 1960 Mys. 338, 349.

See chap. 7, §D, abov 134. A. TR.

application

1960 Mys. at 346-48, A. I. R. 1961 Mys. at 229. These cases represent the

to the field of preferences of the general standards for the constitutionality of

classifications, firmly establsihed by numerous rulings of the Supreme Court. See, ¢.g..

Budhan Chaudhry v. State of Bihar, A. 1. R. 1955S. C. 191; Bidi Supply Co. v. Union of India, ALT. R. 1956 S.C. 479. ji v. State of Mysore, A. 1. R. 1963 S. C. 649 at 658; Chitralekha v. State of Mysore, *. 1823 at 1833. For full discussion, sce chap. 7, §\. above.

Group Membership

315

inferred that this exempts the power from review in those cases where the use of such standards is challenged as inappropriate.

The religious requirement is an expression of the power which Article

341 confers on the President and Parliament to determine which “caste,

race or tribe or part of or group within any caste, race or tribe” shall be included in the list of Scheduled Castes.137 If the standards of reasonable classification are to be applied, the crucial consideration would be whether the division into Hindu and non-Hindu corresponds to some

_difference in conditions, resources, or the incidence of disabilities so

that the division is rationally related to the object of the preferences.

Existing precedents would not seem to foreclose such an approach,

since the cases explicitly upholding religious tests were all decided

before judicial review was firmly established in the preference area and before the Jasani case and others had developed an empirical approach

to questions of group membership.198 At least one court seems to have

moved along these lines. In an unreported 1967 judgment, the Madras

High Court warned the state government that the claim ofa person to belong to the Backward Classes could not be rejected merely on the ground of his recent conversion toa religion like Christianity .139 3. The

argument

that

religious

criteria

are

appropriate

rests

squarely upon the third ground—that acceptance of a non-Hindu religion operates as loss of caste. It is castelessness which constitutes the

crucial difference of non-Hindus from Hindus. In Rajagopal v. Armugam

a two-judge bench of the Supreme Court subscribed to this notion of the de-castifying effects of conversion.#° Upon his youthful conversion to

Christianity, the candidate

lost his membership of the Adi Dravida Hindu caste. The Christian religion does not recognize any caste classifications. All Christians are treated as equals and there is no distinction between one Christian and

another of the type that is recognized between members of different castes

belonging to Hindu religion. ...The tenets of Christianity militate against

persons professing Christian faith being divided or discriminated against on the basis of any such classification as the caste system. It must therefore be 137. Italics added. 138. Punjabrao v. Meshram, A. 1. R. 1965S. C. 1179, clearly does not foreclose such an approach, since it decides only the disputed fact of whether there was a conversion and never reaches any of the constitutional arguments against the religious classification. 139. Meera Bai v. Director of Medical Education (High Court of Judicature at Madras, 20 September 1967. Anantanarayanan, C. J., and Veeraswami, J.). 140. A. I. R. 1969 S. C. 101. In the litigation between these parties their names are

variously spelled Rajgopal and Rajagopal, Armugam and Arumugam. In referring to the reported cases, [ have used the spellings found in the reports; in the narrative account, I have used the longer spellings.

316

Identifying the Beneficiaries

held that, when the appellant got converted to Christianity in 1949, he

ceased to belong to the Adi Dravida caste. 141

Does acceptance of Christianity, Islam or Buddhism invariably evidence a loss of membership in the caste group to which the convert belonged at the time of conversion?142 This can be read as either a question of fact or a question of law—a question about his observable interactions with others or about his legal status. The finding of castelessness in Rajagopal I is based, not on any evidence about the candi-

date’s social interactions with others after his conversion, but solely on

Christian doctrine—and oblivious of common knowledge that Christian practice often falls short of these aspirations.43 The Court acknowledges that in Jasani the sect’s anti-caste views were “‘only ideological and involved no change of state.”44 But it omits to subject the candidate’s experience as a Christian to the same test of operative effect, repeating only that Christianity “is a religion which militates against

recognition of caste.” 145 Christian refusal to “recognize” (in the sense of sanction or legitimate) caste identities among Christians is taken to foreclose the possibility that the court might “recognize” (in the sense of perceive or identify) caste identities that may exist among Christians.

Why in such cases do the courts insist that the act of conversion has as

a matter of law deprived the convert of membership? In other instances

of conversion by members of privileged groups, the courts have addressed themselves to the factual side of the question. In the Jasani case the Supreme Court set out to determine “‘the social and political consequences of such conversion . . . in a common-sense practical way rather

than on theoretical grounds.”4° Thus in the Jasani and Shyamsundar

cases the courts addressed themselves to whether the adherence to sects within Hinduism had actually removed the person from his caste. A similar empirical approach has been applied to conversions among Scheduled Tribes.147 The Scheduled Tribes Order has never contained any religious disqualification, but some tribal politicians have 141. Id., at 107.

142. One of the paradoxes of the religious test as now applied is that an individual

convert from Hinduism to Sikhism or vice versa automatically remains within the Scheduled Castes without any consideration of his intention to remain in the caste,

continued acceptance by his caste fellows, or change in his condition. He is accounted a

member of the caste even though, according to the loss of caste by conversion theory, he has renounced it.

143. In a sequel to this litigation, another bench of the Supreme Court displays a thorough acquaintance with the frequency of caste identity among South Indian

Christians, but comes toa similar conclusion that converts do not suffer from disabilities

(Arumugam v, Rajgopal {II}, A. I. R. 1976 S. C. 939, 949). See discussion in §F, below. 144. A. I. R. 1969 S. C. at 108.

146. (1954) S.C.

J. at 326.

145. Id., at 108.

147. See §C, above.

Group Membership

317

pushed for a bar on converts. Thus, Mr. Kartik Oraon (who back in 1963 failed to disqualify his Christian opponent on the ground that he was no longer an Oraon)" cited as a favorable precedent the protection of Scheduled Castes from ‘‘a more aggressive, vocal and advanced force, i.e., those converted to other religions.”149 The 10% of converts,

he claimed, took 90% of the benefits intended for the tribals. In 1969 the Joint Committée on the Scheduled Castes and Scheduled Tribe Orders Amendment Bill proposed a provision that no person who has given up tribal faith or faiths and has embraced either

Christianity

Tribe.15°

or Islam,

shall

be deemed

a member

of any

Scheduled

This proposal provoked heated controversy during the 1970 debate on the Amendment Bill, since it proposed a glaring distinction between Hindu tribals and those who converted to Christianity and Islam. (It was pointed out that only 4% of the Scheduled Tribes had declared themselves in the 1961 census to be following a tribal religion.) Ironically, the government opposed the proposal on the ground that it constituted unconstitutional discrimination “between converts to Christianity and Islam on the one hand and converts to other religions on the other.”451 Mr. Kartik Oraon expressed puzzlement that the government

found

a religious

test unconstitutional

in the case of

Scheduled Tribes but permissible in the case of Scheduled Castes.15? Similarly, government was alert to the fact that conversion did not

automatically change the conditions of tribals, but insistent that conver-

sion by Scheduled Castes did effectuate a fundamental change.153 In any event, the version of the bill introduced and passed in 1976 did not contain any such provision. Like the government, the courts have dealt differently with questions of religious change among the Scheduled Castes than among the Scheduled Tribes.15¢ Why, in dealing with religious classification among the Scheduled Castes, have the courts forsaken the empirical approach they found congenial in other settings? Why, rather than looking to the facts of the individual case, have they chosen to apply this theoretical rule about loss of caste? As the Supreme Court observed in the Jasani

148. See Kartik Oraon v. David Munzni, A.1.R. 1964 Pat 201, discussed in § C, above. 149. LSD (4th series), Vol. 16, No. 50, col. 3147 (25 Apr. 1968). 150. RCSCST 1969-70: 93. 151. Shri K. Hanumanthaiya, Minister of Law and Social Welfare, LSD (4th series), Vol. 45. No. 6, cols. 318-19 (17 Nov. 1970). 152. Shri Kartik Oraon at place cited in n. 149; also at LSD (5th series). Vol. 44, No. 16, col. 53 (2 Sep. 1976).

153. See Shri B.K. Daschowdhury, LSD (4th Series), Vol. 45, No. 10, col. 265 (23 Nov. 1970). 154. See §C, above.

318 Identifying the Beneficiaries case, “conversion . . . imports a complex composite composed of many

ingredients. Religious beliefs, spiritual experience and emotion and intellectual conviction mingle with more material considerations such as severance of family and social ties and the casting off or retention of old customs and observances. The exact proportions of the mixture vary from person to person.” 155 It is surprising that the courts both in old cases like Michael and Thomas and in recent ones like Rajagopal have accepted a picture of conversion which corresponds more closely to missionary aspirations than to observable consequences. This acceptance reflects the continued force of a view of caste groups which sees them as units in an overarching sacral order of Hinduism. To abandon the whole is to abandon the part. According to this view, which prevailed prior to the Constitution, caste groups were visualized as occupying a unique place in an integrated but differentiated religious order in which the different parts enjoyed rights and duties, privileges and disabilities as determined by their position in this order.456 From this view of caste derived the long-standing reluctance of the courts to give legal effect to caste standing among non-Hindu communitigs.157 But the courts have always recognized castes among non-Hindus when the claim was not an assertion of standing in this sacral order, but was a

claim regarding the corporate autonomy or internal regulation of the

group.15® For the latter purposes “caste” included “any well-defined

native community governed for certain internal purposes by its own rules and regulations” and was thus not confined to Hindus.159 In the preference area the question of non-Hindus having caste arises in two sorts of factual situations: first, those involving individual con-

verts; second, those involving a caste group or a section of a caste made up of members who are non-Hindus. In the case of individual converts, the question facing the court would seem to be whether the individual’s acceptance of Christianity, Islam, or Buddhism evidences a loss of membership in the caste group to which he belonged at the time of conversion. There is evidence that in at least some cases of conversion the convert continues to regard himselfas a member of the old caste and 155. (1954) S.C. J. at 326.

156. For an analysis of this and competing conceptualizations of caste, see Galanter

1968. : 157. See, e.g., Michael Pillai v. Barthe, A. 1. R. 1917 Mad. 431, where a claim of

high-caste Christians for restoration ofa wall separating them from low-caste Christians in church was rejected on the ground that such a claim for precedence could not be

enforced among Christians, since it was based on Hindu notions of pollution. 158. Abdul Kadir v. Dharma, 1. L. R. 20 Bom. 190 (1895); Yysuf Beg v. Malig, A. 1. R. 1927 Mad. 397; Inder Singh v. Sadhan Singh, 1. L. R. 1944 (1) Cal. 233. 159. Abdul Kadir v. Dharma, I. L. R. 20 Bom. 190 at 192 (1895).

Group Membership

319

to be so regarded by others.‘ The unwillingness of the courts to credit such membership derives not from an examination of the facts but from an identification of castes as components in the sacral order of Hinduism. Caste and Hinduism are regarded as coterminous, when Hinduism is abandoned, so is caste membership. The second type is somewhat more troublesome. There is little dispute that such persons as, e.g., the Sikh Bawarias in the Gurmukh Singh case are, in fact, members of a caste group of the kind whose existence among non-Hindus has been well known and long recognized by the judiciary.61 There is evidence that some non-Hindu castes or parts of castes are in circumstances equivalent to those of the Hindu untouchables.162 To refuse to recognize caste membership in such cases implies that the ‘“‘caste” to which the court is addressing itself is not caste in the sense of a body of persons bound by social ties, but

caste in the sense of a body of persons which occupies a given place in the ritual order of Hinduism. A more troublesome case is presented by the adherents of the Buddhist movement initiated in 1956 by Dr: Ambedkar. More than three million persons, mostly Mahars from Maharashtra but including

members

of other

Scheduled

Castes

in many

states,

had

become

Buddhists by the time of the 1961 census. In addition, a large but un-

known number have refrained from, or concealed, conversion in order

to remain

Castes.1

eligible for the preferences

they enjoyed

as Scheduled

,

Exclusion of Buddhists is of course mandated by the “Hinduism” requirement of the Scheduled Castes Order: 160. The reports are replete with cases in which converts have lived so indistinguishably with their caste fellows that the courts retrospectively infer a tacit reconversion without either formal abjuration of the new religion or formal expiation and

re-admittance to Hinduism (Durgaprasda Rao v. Sundarsanaswami, A. 1. R. 1940 Mad. 513; Gurusami Nadar v. Irulappa Konar, A. 1. R. 1942 Mad. 193). And cf. Muthusami Mudaliar v0 Masilamani, 1. L. R. 33 Mad. 342 (1909), where Christian wives were accepted as members ofa Hindu caste. The “indulgence” extended by the State in the Thomas case, n. 126 above, seems to reflett an awareness that recent converts, if not effective members of their old castes, are at least subject to similar disabilities. Cf. the Kerala fee concessions in education for Christian converts from the “Backward Classes” (RCSCST 1960-61 : 320). 161. See I BCC 28 ff; cases cited n. 158, above.

162. See e.g., the observations of the Commissioner regarding Christian and Muslim sweepers, who do not receive special treatment (RCSCST 1960-61 : 316). 163. On this Buddhist (or neo-Buddhist, as it is sometimes called) movement, see Zelliott 1966 and 1969; Keer 1962; Isaacs 1965; Miller 1966.

164. “The majority [of Buddhists] consider the special privileges their right on grounds of past suffering and present economic restrictions, and do not list their

children as Buddhists until they have made their way through the educational system and into government jobs” (Zelliott 1969:292).

320

Identifving the Beneficiaries

As Buddhism is different from the Hindu religion, any person belonging to

the Scheduled Castes ceases to be so if he changes his religion. He is not, therefore, entitled to the facilities provided under the Constitution specifically for the Scheduled Castes. 165

This is reinforced by the notion that converts have no caste since “Buddhism [does] not recognize castes.”16 conversions the Home Secretary announced:

Shortly after the first

The legal position is fairly clear; namely, that if the fact of conversion is proved, then obviously the persons concerned cease to belong to any caste

much less to the Scheduled Castes. Where a right, privilege or preference

has been specially recognized under the Constitution specifically for members of the Scheduled Castes, such treatment cannot obviously continue if the persons have ceased to belong to the Scheduled Castes by reason of the conversion. *67

Recognizing that conversion was of itse!f unlikely to improve the condition of the converts, the central government recommended that the state governments accord the Buddhists the concessions available to the Backward Classes. Such preferences, less in scope and quantity than those for Scheduled Castes, have been granted in some states, but

others have withdrawn all preferences.16 Persistent efforts of Buddhists to regain the preferential treatment that they formerly enjoyed have achieved only partial success. The State of Maharashtra, where most of the Buddhists are concentrated,

restored to them those of the benefits enjoyed by Scheduled Castes

which are conferred by the state. But the central government (and other 165. RCSCST

1957-58:1, 25. This ruling is based squarely on the “Hinduism”

requirement of the President’s Order. See the statement of Pandit Pant, Times of India,

Aug. 21, 1957, p. 12, col. 3. Compare the observation of the Supreme Court that “Hindu” for purposes of the Scheduled Caste Order includes that religion which “recognizes caste and injunctions based on caste distinctions” (Punjabrao v. Meshram,

A.LR. 1965 S.C. 1179 at 1184). This finding, along with the non-inclusion of Buddhists

among Hindus, was gratifying to most Buddhists, who felt that the decision confirmed both the efficacy of their conversion ceremony and the torrectness of their conviction that they are no longer Hindus (Zelliott 1966: 207).

166. Statement of Mr. B. N. Datar in the Rajya Sabha, reported in Times of India, Aug. 27, 1957, p. 10, col. 1.

167. Letter of Home Secretary to Mr. B. K. Gaikwad, M.P., of June 13, 1957, reported in U. P. I. dispatch from Poona, June 22, 1957. The Home Secretary went on to

say that, as to “‘the actual circumstantial need of these converts to Buddhism for special treatment toward their educational and economic advancement . . . the issue is one of pure fact and would be whether they are sufficiently backward to be eligible for special

assistance as some of the weaker sections of the people.” That is, they would be treated as backward classes but not as Scheduled Castes. 168. RCSCST 1957-58:1, 25; II, 60.

Group Membership states)

have

remained

adamant

in their unwillingness

321

to include

Buddhists on a parity with Scheduled Castes. The major benefits outside the purview of the state governments are reservations in legislative

bodies, post-matriculation scholarships, and reservations in central government employment. Since most Buddhists oppose the continua-

tion of legislative reservations, the dispute has centered on the latter benefits. 1© After several years of vacillation the central government in 1969 acceded to Buddhist demands in regard to post-matriculation scholarships, leaving the other practices unchanged.17° When in 1976 the Scheduled Castes Order was amended by Parliament, Buddhists were not included.171 It is clear that although Buddhists are not members of a “‘caste” in

the sense of a component of the Hindu sacral order, they constitute a

“caste” or community in the minimal sense of an identifiable group who share religious tenets and social practices and who marry and socialize chiefly among themselves. If this community can be shown to comprise persons who were subject historically to the deprivations suffered by the Scheduled Castes and who suffer at present from disabilities and invidious treatment, what distinguishes them from the Scheduled Castes for purposes of preferential treatment designed to overcome these disadvantages? (There is some evidence that they do suffer comparable

169. Most Buddhist spokesmen contend only for economic, educational, and employment benefits. See Republican Party 1964: 20. However, Professor Bhandare (1965) argues that Buddhists are entitled to all benefits, including legislative reservations. Other Buddhist spokemen have expressed opposition to any reservations—especially in government service because it removes educated members of the community from participation in political activity (interview with B. P. Maurya in Chicago, 28 June

1963). 170. A bill that would have restored benefits to Buddhists was defeated in the Lok

Sabha (New York Times, Aug. 30, 1961, p.2, col. 6). Restoration of benefits was one of the demands in the great Buddhist agitation of 1964-65, which resulted in over 300,000

arrests (Republican Party 1964; Khobrogade 1966). Subsequently, the central govern-

ment appeared to move close to this position. In 1966, Mr. Y. B. Chavan, then defense

minister, told a rally that the central government had accepted in principle the extension of concessions given to Scheduled Castes to neo-Buddhists also (Indian Express, April 15, 1966). However, the Congress agreement in principle fo drop the

Hinduism test for Scheduled Castes was later explained as a proposed shift from caste to economic criteria to be implemented by 1981 (Rudolph and Rudolph 1967: 149). Educational concessions at the college level were finally extended to Buddhists in November, 1969 (Phadnis 1969: 1897).

171. The proponents of inclusion of the Buddhists took the tack of trying to incorporate the language of Explanation IT (to Art. 25 of the Constitution) into the Scheduled Caste Order—i.e., that “that reference to Hindu shall be construed as including a

reference to persons professing Sikh, Jain or Buddhist religion” LSD (5th Series), Vol. 64, No. 17, cols. 147 ff. (2 Sep. 1976). Cf. ProfessorN. M. Kamble, RSD, Vol. 97, No. 18, cols. 17 ff. (3 Sep. 1976).

322

Identifying the Beneficiaries

disabilities and that the condition of most converts has not changed appreciably.)172 In effect, their disqualification lies in their failure to acknowledge their low ritual status by the acceptance of caste identity. Such acceptance and acknowledgment is made a condition of receiving benefits designed to dispel their depressed social and economic

conditions. Exclusion of Buddhists, then, involves a curious transposition of ends and means. Historically, low ritual status was employed to

identify the recipients of preferences directed toward secular betterment because of a presumed correlation between ritual impurity and depressed socio-economic conditions.173 Here, evidence of the exis-

tence of those conditions which preferences are designed to dispel is never considered. Acknowledgment of low ritual status is transformed

from an instrument for identifying the deserving into an independent prescriptive requirement. As a result, the preference policy is transformed into an instrument for extracting such acknowledgment. The argument has been advanced that existing constitutional provisions foreclose extension of preferences to the Buddhists and that an amendment would be required to empower the government to do this. But Articles 15(4) and

backward classes.17*

16(4) clearly contemplate a plurality of

There would seem to. be nothing to prevent

the government from denominating the Buddhists as Backward Classes and conferring upon them preferential treatment in education,

government employment, distribution of land, etc., at the same level

that it is conferred upon the Scheduled Castes—as Maharashtra has done.

Since legislative reservations alone are constitutionally specified ex-

clusively for the Scheduled Castes, the inclusion of Buddhists would depend upon whether they could be encompassed within the Scheduled Caste category. The history of the Scheduled Caste classification militates against the notion that the present definition of Scheduled Castes represents the farthest constitutional limit of that category. The first Scheduled Castes Order of 1936 did not contain a Hinduism test (although Christians were excluded).175 The Hinduism test was not specified in the Constitution.176 Its subsequent formulation can 172. On the continuing disabilities experienced by Mahars who have become Buddh-

ists, see Karve 1968: 97. Nair (1962: 187) reports of Sewagram (Gandhi's ashram and

model village), where the Mahars have become Buddhists: “There has been no improve-

ment in the status of the Harijans since they embraced a new religion. They continue to

be treated as untouchables.” 173. Dushkin 1961: 1667. 174, See chap. 13, §§C, D. 175. See n. 89, above. « 176. At one point in the drafting of the Constitution, Mr. K. M. Munshi moved an amendment which described the Scheduled Castes as “‘a section of the Hindu community” and which was accepted but subsequently omitted in accord with the agreement to include some Sikhs among the Scheduled Castes (Shiva Rao 1968: 760).

Group Membership

323

hardly be considered an unalterable expression of the constitutional

mandate.?77 It was certainly not so regarded in 1956 when the govern-

ment expanded the requirement to include Sikhs.178 The power of Parliament is presumably wide enough to designate as Scheduled Castes any “‘castes” or parts of castes. It is not clear that this is confined to “caste” in the sense of a group that is a ranked component of the Hindu sacral order or whether “caste” in the sense of community would suffice. The argument for the more restrictive interpretation is that the classification is based upon disabilities consequent upon the occupation ofa low position in that sacral order and applies only to those within it. But if the group is identifiable as one whose present disabilities stem from former membership in that order, the welfare objectives of the preference policy would seem to recommend that “caste” be interpreted widely enough. to include all who had been victimized by that order. 177. The Order was issued in August 1950 (P. T. I. Dispatch from New Delhi dated August 16, 1950).

178. See n. 90 above on the process of including Sikhs. The formidable difficulties which their inclusion poses to the argument against inclusion of the Buddhists is nicely displayed in the following exchange from the Rajya Sabha debate over the 1976 revision

of the Scheduled Castes and Scheduled Tribes Order: Shri Shyam Lal Yadav (Uttar Pradesh): Sikhs are also a different community, a

different religion. Why the Sikhs are included? Shri K. Brahmananda Reddi [Home Minister]: They were included long ago. Shri Shyam Lal Yadav: . . . Sikhism is also a different religion than Hinduism as Buddhism is a different religion. Shri K. Brahmananda Reddi: Sikh religion is treated as coming under Hinduism.

Shri Shvam Lal Yadav: How? Sikhism is quite different.

The Vice-Chairman (Shri Lokanath Misra): You cannot expect the Hon. Minister to go into the philosophy of Hinduism at this stage. Shri Kalp Nath Rai (Uttar Pradesh): I want to know whether Hinduism includes

Jainism, Buddhism and Sikhism. . . .

(Interruptions)

Shri K. Brahmananda Reddi: I have just now read what it means.

The Vice-Chairman (Shri Lokanath Misra): It is because of the Presidential Order that the effect has been given and, therefore, that is the finality in the matter.

Shri Shyam Lal Yadav: They can change it. Shri K. Brahmananda Reddi: Whether Parliament can change it or not, is a different issue altogether. .. . Prof. N. M. Kamble: . . . Now, the quesion is whether the Buddhists are governed by the Hindu Law. Hon. Minister knows it. Shri K. Brahmananda Reddi: I am not going into that question. I have just now answered on the basis of information that we have that on consideration these Buddhists do not form a part of them so as to be entitled to be considered as Scheduled Castes.

Shri K. K. Madhavan: What about Sikhism?

(Interruptions) The Vice-Chairman (Shri Lokanath Misra): Order, order, please sit down. RSD Vol. 97, No. 18, cols. 49-51 (3 Sep. 1976).

324

Identifying the Beneficiaries

The exclusion of non-Hindus from preferences appears to give expression to a view of caste that is at variance with the constitutional and statutory “dis-establishment” of the sacral concept of caste since Indian Independence. Varna distinctions have been abolished in the personal law;!79 untouchability has been abolished; imposition of caste disabilities have been prohibited;1® all support for claims of precedence based on caste standing has been withdrawn.181 Caste groups are, of course, free to entertain any views they wish concerning

their station within (or without) Hindu society. But they may no longer call on the government or the courts to confirm and enforce these views. The law recognizes caste groups as corporate entities, bound by mani-

fold ties of association. The nature of these ties, religious and social,

varies from group to group in accordance with its own internal order. But the Constitution bars governmental recognition of hierarchical

ordering, the apportioning of sacramental honor among caste groups, or the imposition of religious tests for recognition of a group as a

caste. 182 Since preferences for Scheduled Castes are meant to alleviate the disadvantages historically associated with low standing in the Hindu sacral order, an attenuated recognition of this hierarchical order is utilized for the purpose of deciding who is to receive these preferences. Scheduled Castes are indentified-in part by considerations of vama, pollution, and ritual distance.*®3 But it is clear that these otherwise discredited legal categories are appropriate only to the extent that they serve to identify the needy and deserving. They are appropriate only as convenient descriptions of practical conditions, not as prescriptive requirements. Conversion, both to sects within Hinduism and to religions outside Hinduism, has traditionally been and continues to be one of the common expedients of those at the bottom of the caste hierarchy seeking to improve their position.*8 Historically, the disadvantages associated 179, Hindu Marriages Validity Act, 1949 (XII of 1949) and the Hindu Code Acts, n.

71, above. 180. Art. 17, Art. 15(2), Untouchability (Offenses) Act, 1955 (XXII of 1955) and

kindred state legislation. On the operation of these prohibitions, see Galanter 1972. 181. See Galanter 1968.

182. It is open to a caste group itself to impose religious requirements for membership. See Deochand Totaram v. Ghaneshyam, A. 1. R. 1935 Bom. 361.

183. In fact, attempts to identify the Scheduled Castes in terms of criteria of pollution, impurity, and disabilities proved inadequate to isolate the groups which local administrators felt deserving of inclusion. Additional criteria of poverty and illiteracy had

to be added. Thus the Scheduled Castes were determined by an uneven mixture of ritual and socio-economic criteria. See chap. 5, above. 184. Titus 1959; Pickett 1933; Hazari 1951; Keer 1962; Isaacs 1965.

Group Membership

325

with their low position have often followed converts across religious line; they continue to do so in at least some cases in the present¥##5—a fact not unknown to the government. 18 How effective or ineffective a means of escape conversion has proved to be in a particular case should be evaluated in empirical terms. To deem conversion to non-Hindu communions an automatic disqualification for aid violates the constitutional command of equal treatment for different religions. It also restricts freedom of religion,187 which might be thought to require that government refrain from administering its welfare schemes so as to put a heavy price-tag on its exercise.18° And, apart from its dubious constitutionality, it is at variance with the stated welfare objectives of the policy of preferences for the disadvantaged.19 The Hindu requirement seems to retlect a hostility toward conversions which is anachronistic.19° Conversions do not carry the same import of political and national division that they did in pre185. For some recent and pungent observations on the failure of conversion per se to alter the condition of low caste converts, see Andhra Pradesh Backward Classes Commission 1970: 202-3. 186. Before the Buddhist movement was on the scene, Home Ministry spokesmen

were fully cognizant of the continuing incidence of disabilities on members of nonHindu religions. Thus in the Rajya Sabha debate on the scope of the Untouchability

(Offences) Act, Mr. B. N. Datar observed that untouchability is carried across religious lines and pursues converts (Times of India, Sep. 19, 1954). Cf. the assurances to Parliament of the then Home Minister, Pandit Pant, that the Untouchability (Offences)

Act would apply among Christians and Muslims as well as to Hindus (quoted in Banerjee 1960: 100). 187. Apparently, the Hinduism requirement has curtailed conversion; Lelah Dushkin (1957: 97) reports that “the fact that it was only the Hindu lowest castes that were included (in the Scheduled Castes lists) virtually stopped the movement toward conversion to other religions, in most areas.” 188. Cf. the American case of Sherbert v. Verner, 374 U.S. 398 (1963).

189. In Ganpat v. Presiding Officer, A.1.R. 1975 S.C. 420, a two-judge bench of the

Supreme Court advances a curious argument that the religious test impedes the perpetuation of preferences. “The attempt of persons that have changed their religion from Hinduism to Buddhism” to retain benefits as Scheduled Castes “only shows that

otherwise these persons might get a vested interest in continuing to be members of the Scheduled Castes” (id., at 425). The religious test of the Scheduled Castes Order is said

to be salutory “from the point of view of discouraging that tendency” (id.). Of course anything that cuts off benefits can be said to prevent those excluded from developing a

vested interest in them. But if denial of benefits to non-Hindus may discourage conversion, it is difficult to see how it decreases the urge to retain benefits (except in

circles where they might be a stigma of rejected Hindu identity). Indeed, it is instances of the non-occurrence of such an effect that the court cites as evidence that it obtains. 190. Cf. the disabilities imposed upon converts from Hinduism in the various Hindu

code acts. See, e.g., Sundarambal v. Suppiah, A. 1. R. 1963 Mad. 260. Derrett (1958: 393)

remarks that it is “‘strangely inconsistent with the claim to be a secular state” that a

Hindu who changes his religion is liable to be divorced by his wife, may forfeit an

326

Identifying the Beneficiaries

Independence India,’9" and anxiety inspired by the threat of separate

electorates is similarly out of date. The test is certainly not designed to

bolster Hindu orthodoxy, since one can remain a Hindu while embracing the most heterodox beliefs and practices. Not only is the test

inconsistent with the ban on religious discrimination but, ironically, it

seems to impose on Hinduism the notion ofa hard and fast line between

creeds and communions. This is not a characteristically Hindu notion, but one more consonant with the exclusivist creeds of the West, which

require the convert to abjure his previous faith. It is rather surprising that Indian jurisprudence should give currency to this notion rather than to views more reflective of the Indian tradition of religious

tolerance.19?

Preferences are designed to remedy certain conditions— particularly to offset or dispel the disadvantages created by the imposition of social disabilities and the lack of economic and educational resources. The usefulness of the theory that converts are casteless in describing these conditions is questionable. Where the preference is one that devolves on the members of the group as individuals, surely the appropriate question is whether the disabilities and disadvantages have, in the particular case, been effectively dispelled by the conversion. Where the preference is a political one that devolves on the group corporately, the appropriate question would be whether the group still accepts the convert as a member and is willing to have him as a representative. In neither

instance would it be necessary to decide whether he is a member of the

group for all purposes or whether he is a Hindu. F. GROUP

INTEGRITY: ADMISSION, EXPULSION, AND RE-ADMITTANCE

In the cases discussed earlier, where the tests of affiliation (i.e.,

individual intention and conduct) and group acceptance were em- existing claim to maintenance, may lose the right to give his child in adoption, may lose the right to be the guardian of his own issue, and may have his issue deprived of the right

of inheritance from unconverted relations. See also D. Smith 1963, chap. 6.

191. It should be recalled that the “Hinduism” requirement was formulated in the aftermath of the controversy produced by Dr. Ambedkar’s 1936 threats that untouchables would convert from Hinduism to an unspecified religion. Hindu leaders were particularly eager to get assurances that it would not be Islam. There was even talk of assurances that ifit were Sikhism the converts might retain the benefits accorded under the Government

of India Act,

1935. See Zelliott

1969: 214 ff; Keer

1962: 2774,

Anonymous 1935-37. 192. For an example of judicial expression of Hindu cosmopolitanism, see Kolandei v. Gnanavarum, A. 1. R. 1944 Mad. 156, where a gift by the manager ofa Hindu joint family

to a Roman Catholic church was upheld as binding on his heirs, since “the very idea

Group Membership

327

ployed to decide questions of group membership, their conjunction was sufficient to confirm a previously existing membership.’ But can affiliation and acceptance operate to create a new membership? or to restore membership that has been lost? The latter problem has arisen as an extension of the loss-of-caste-byconversion theory employed in Rajagopal v. Armugam.’™ After his youthful conversion to Christianity, candidate Rajagopal had effected a reconversion to Hinduism bya series of acts “clearly amount[ing] to a

public declaration of his professing the Hindu faith.”195 In line with the requirements of the Meshzam case, he had made “‘a complete public declaration” in such a way that the change would be known to those whom it might interest.1% Therefore, at the time of the contested election, he indisputably met the “Hinduism” requirement of the Scheduled Castes Order. However, since conversion to Christianity had divested him of caste membership, 1% the question arises whether upon reconversion he had again become a member of the Adi Dravida caste and thus regained his eligibility to stand for a reserved seat.19® To answer this, the Court that pious and charitable purposes should . . . mean objects that can be said to promote

a particular faith, creed, or dogma is foreign” to the ideals of Hinduism (id., at 157). Cf.

State of Bombay v. Yagna Sastri Purushadasji, 61 Bom. L. Reporter 700 (1960), where admission of non-Hindus to membership was held not incompatible with a sect being “Hindu.” While such cosmopolitan tolerance should not be imposed on groups which find it uncongenial, certainly the courts should not impose a narrower view on ups which are willing to accept members of different religions.

193. §§ B, C, above.

195.

194. A. I. R. 1969S. C, 101.

Id., at 107. The acts of the candidate included marriage to a Hindu woman,

bringing up his children as Hindus, entering them as Hindus on school records, applying for alteration of his service record to show he was a Hindu. 196. See text at n. 106, above.

197. Without the interposition of the loss-of-caste-through-conversion theory, the question can be addressed in straightforward factual terms. In a similar case where a candidate had converted to Christianity for purposes of marriage, and later reconverted, a single judge of the Andhra Pradesh High Court, applying the Jasani tests,

concluded: “He continued to retain his caste, though he lost his religion. . . . After his reconversion to Hinduism ... he regained his religion and became a full-fledged ‘bindla’ . . . qualified to stand for election [from a reserved constituency)” ( Narasimha

Reddy v. Bhupathi, 31 E. L. R. 211, 223 [1967]).

198. This two-level inquiry into (1) reconveysion and (2) caste membership was

unknown to earlier administrative policy. For purposes of eligibility for preferences, reconverts to Hinduism who were born in Scheduled Castes have been deemed members of the Scheduled Castes (RCSCST 1953: 132). See also the statement of the Union

Home Minister, reported in the Times of India, Sept. 12, 1959. However, a draft directive suggests that the Rajagopal approach has been accepted for administrative purposes:

“a convert or reconvert to Hinduism or Sikhism may be accepted as a member of the

Scheduled Castes if he has been received back and accepted as a member by the concerned Scheduled Caste” (CWSCSC [4th] 2: 59 [1969]).

328

Identifying the Beneficiaries

turns to a doctrine developed in the application of the personal law that reconversion might be effectuated by acceptance by a Hindu community upon undergoing whatever formalities it saw fit to impose—even

none at all.199 In these cases, acceptance by the caste was employed as evidence of successful reconversion. In Rajagopal, however, the reconversion is conceded: the question is whether he regained his caste membership. The old cases’ test for reconversion is transformed into a test for caste membership. Assuming, without deciding, that re-entry into a caste is possible, the Court finds no evidence that it had occurred. If re-entry were possible, it would be by virtue of acceptance by the caste. In matters affecting the well-being or composition of the caste, the caste

itself is the supreme judge. It was on this principle that a reconvert to Hinduism could become a member of the caste, if the caste itself as the

supreme judge accepted him as a full member of it.

Here, the Court found no evidence to satisfy this requirement. The various acts of the candidate that amounted to his public declaration that he was a Hindu do not show that the Adi Dravidas re-admitted him or started to recognize him as one of them. Since his marriage (to an Adi

Dravida woman) “was not performed according to the rites observed by

members of that caste,”’201 it was no proof of admission into the caste by “members of the caste in general.”202 There was no evidence that at any stage “‘any step was taken by members of the caste indicating that ... the candidate] was being accepted as a member of this caste.”203 Acceptance

by his immediate family and a few relatives

“cannot be held to be equivalent to a recognition by the members of the caste in general.”20 This opinion raises many puzzling questions. By “members of the caste in general” does the Court refer to the endogamous jati (kula) to which the convert previously belonged? or to all who call themselves Adi Dravidas—a numerous group spread over much of South 199..No proof of formal abandonment of his new religion is necessary for the convert to effect a successful reconversion to Hinduism. While a mere declaration is not sufficient to restore him to Hindusim ( Ramayya v. Josephine Elizabeth, A. 1. R. 1937 Mad. 172), acceptance by a Hindu community with whatever formalities it deems proper—

even none at all—is sufficient. Durgaprasada Rao v. Sundarsanaswami, A. 1. R. 1940 Mad. 513; Gurusami Nadar v. Inilappa Konar, A.1.R. 1934 Mad.630. However, cf. Marthamma

v. Munuswami, A.1.R. 1951 Mad. 888 at 890, where the primary test is the “intention” of

the reconvert; the court says “the religious persuasion of a man now-a-days depends on his ‘subjective preference’ for any religion.” Cf. the declaration test discussed in n. 100 above. 200. A. I. R. 1969S. C. at 109. 201. Id., at 110. 202. Id., at 110.

203. Id., at 110.

204. Id.

Group Membership

329

India?20s or to the Adi Dravidas of the constituency in question? or to some other group larger than his endogamous jati but smaller than all Adi Dravidas? The Court does not specify what it would consider adequate evidence of acceptance: would a pronouncement by caste dignitaries or a resolution of a caste association suffice? Nor is there any indication

that the caste (whatever unit is involved)

possesses any

organs through which it might express itself on the question.2 Unlike the small communities involved in the personal law reconversion cases, Adi Dravidas are a widespread composite group. Whatever adequate evidence of acceptance might be, it is clear that the Court places on the “re-convert” the burden of proving it.-Since the Supreme Court had earlier ruled that the burden of proof rests on the challenger to disprove

a candidate’s claimed membership

in a group,?°7 it appears that

conversion, even when followed by reconversion, shifts the burden of

proof to the candidate. Rajagopal did not tarry long in discharging this burden. Three months after his appeal was dismissed by the Supreme Court, a conference was held under the auspices of the Kolar Gold Fields Adi Dravida Scheduled

Castes Conference. This

conference,

which

Noranha, J.,

later referred to as Rajagopal’s “magnum opus,”208 attracted at least

205. According to the 1961 Census Adi Dravida is one of several names used by parts

of the vast Parayan

cluster of Madras

(others are Paraiyan, Samban,

Panchama,

Vettiyan), whose total population in 1961 was 3.4 millions, over 56% of the entire

Scheduled A, pp. 13, Dravida” the 1900s. 1921

Caste population of Madras 19). There were also 37,439 (lit., original Dravidian) was The census return under this

(Census of India 1961, Part IX Madras, Part V Adi Dravidas listed in Andhra. The term “Adi adopted by many Tamil untouchable castes in heading increased from nil in 1911, to 49,978 in

to 1,546,520 in 1931, when it was an accepted census classification. In 1922 the

Madras Government undertook to adopt the terms Adi Dravida and Adi Andhra for future record-keeping purposes. (G. O. No. 817 (Law General), dated 35th March 1922, reprinted in People’s Herald 2(3):5 (July 15, 1969). As the name suggests, its use

implied that this group were the original inhabitants of South India who were subsequently reduced to their present low estate by Aryan invaders. For an exposition, see Rajah 1925, chap. 4. Thus, Adi Dravidas are a composite group, the components of

which retain their distinctiveness to somé extent. Gough (1960: 17, 244) describes Adi

Dravidas as not “a single organized group” but “rather a named caste category of similar culture and occupation comprising a number of endogamous groups each in turn divided into local communities.” In the village she studied these groups “retained separate administrative organization for each sub-caste community.” Similarly Béteille

(1965b: 92) observes that although higher castes and the government often treat Adi

Dravidas as a single unit, “there are divisions and subdivisions among them.” 206. Indeed, there is some question whether a composite group like Adi Dravidas

may exercise the traditional prerogatives ofa caste. Cf. Ellappa v. Ellappa, A. 1. R. 1950 Mad. 409. 207. Laxman Siddappa Naik v. Kattimani Chandappa Jampanna, A. 1. R. 1968 S. C. 929 (discussed in §A, above). 208. Rajagopal v. Armugam [II] 1973(2) Mys. L. J. 240 at 250.

330

Identifying the Beneficiaries

a thousand persons, mostly Adi Dravida Hindus, and addressed vari-

ous questions of interest to Scheduled Castes. At the conference

a religious ceremony was performed ...it was a “Shuddh” or cleaning

ceremony to remove the stigma of [Rajagopal’s] being considered as a Christian pursuant to the judgment of [the Mysore High] Court and the

Supreme Court and their publication in some newspapers and it was not a

ceremony to convert him, already a Hindu, into Hinduism.2

Accordingly, the conference resolved to set right the doubt arising in [Rajagopal’s] election petition, recognize and also admit him to our Caste, Sri S. Rajagopal, who was born in our Caste,

brought up amidst us. and who represented our Caste in Mysore Legisla-

ture for the past 6-7 years, serving efficiently and honestly.21°

As the -1972 general elections approached, Rajagopal and Arumugham squared off for another electoral contest and another round of litigation (which we shall call Rajagopal I]). Arumugham objected to Rajagopal’s nomination papers on the ground that upon his reconversion, Rajagopal had not re-obtained his membership in the caste. The Returning Officer rejected Rajagopal’s papers. Arumugham was duly elected and Rajagopal challenged the election. The Returning Officer’s rejection, according to the High Court, was “hasty, unwarranted and erroneous.”2!1 The Adi Dravida conference was ‘Vox Populi . . . abundant proof that he belongs to the mainstream of that community.”’212 He showed that he was a member of the caste “without a semblance of reasonable doubt.”?%3 On appeal the Supreme Court seemed about to reopen the fundamental question of the relation of religion to caste. Aware that the correspondence of caste and Hinduism is imperfect, it observes that caste identity among Christians is “not an infrequent phenomenon in South India.”2#* Whether converts to Christianity remain members ofa caste “ultimately . . . must depend on the structure of the caste and 209. Id., at 251. 210. Id., at 253.

211. Ud., at 255.

212. Jd., at 253. Arumugam argued that the conference was an improper body for such a manifestation of caste assent and that the only legitimate arbiters of acceptance

into the Adi Dravida fold in the Kolar Gold Fields were the Nattamdars [headmen] of

each block in the mining area and the Adi Dravida Jana Sabha. After inspection of the

evidence the court concludes that “the imstitution of Nattamdars and the existence of the Adi Dravida Jana Sabha in KGF can be dismissed as belated inventions for the

purposes of defense” (id., at 254). (The court observes that panchayats sponsored by the

mine management and “unrecognized panchayats” formerly functioned in the blocks but are now reported to be defunct [id., at 254-55.])

213. /d., at 255.

214. Arumugam v. Rajgopal [11] A. 1. R. 1976 S. C. 939 at 945.

Group Membership

331

its rules and regulations.” 215 Does the caste expel converts or does it tolerate them within its fold? The Court suggests that such toleration seems to prevail among Adi Dravidas. There are both Hindus and Christians, and intermarriages between them, so it seems prima facie that on conversion to Christianity Rajagopal “did not cease to belong to

the Adi Dravida caste.”?!6 This line of argument, which appears to undermine the “loss of caste” notion, would bring the Court squarely to the

question

of the constitutionality

of the Hinduism

test of the

Scheduled Caste Order. But the Court veers away, saying it need give no final opinion on whether he remained in the caste, since it finds he was not only reconverted but re-admitted.

Taking the other path, the Court characterizes a caste as a “social

combination of persons governed by its rules and regulations,” which

need not be written or formalized.217 The caste may, “‘if its rules and regulations so provide, admit new members just as it may expel an

existing member.” It may specify particular rites or ceremonies for re-admission, but if it does not, then “the only thing necessary for readmission would be the acceptance of the person . . . by the other members of the caste.”218 The question then is whether the Adi Dravidas accepted Rajagopal after his reconversion. Of the battery of evidence he had assembled, the Court found his participation in Hindu weddings and funerals unpersuasive. Nor was it impressed by the fact that he was looked upon as a leader and peacemaker by the Adi Dravidas of the locality. It was most impressed by the 1968 conference, which had been held with the object, inter alia, of readmitting [Rajagopal]

. . . into the fold of Adi Dravida caste and not only

was a purificatory ceremony performed

... with a view to clearing the

doubt which had been cast upon his membership . . . by the decision of this

court [in Rajagopal I] but an address . . . was also presented to [Rajagopal] . . . felicitating him on this occasion.??9

The Court also found persuasive his laying the foundation stone for construction

of a new

wall of the Hindu

temple,

and

his honorific

participation in temple ceremonies was “strongly indicative of the fact that he was accepted and treated as a member of the Adi Dravida

community.”’220

Yet the Court’s estimation of this array of evidence is puzzling. Recall

that

it was

Hinduism

earlier established

that Rajagopal

and indisputably met the Hinduism

215. Id., at 945. 217, Id., at 48. 219. Hd, at 951.

216, Ud, at 47, 218, Id., at 48. 220. Id., at 950.

had

reconverted

to

requirement of the

332

Identifying the Beneficiaries

Scheduled Caste Order; the only question was of his caste membership. But since the Court indicated its awareness that Christians were accepted as members of the Adi Dravida caste, it is not clear what

remained to be proved with this evidence. Its emphasis on Hindu ceremony suggests that the Court was interested in caste acceptance of

him as a Hindu—for if he were accepted as a member of the caste but not viewed as a Hindu, the problem of the Hinduism requirement

would be unavoidable. (Of course, he might be deemed a Hindu for

purposes of the Scheduled Caste Order even if he was not recognized as one by his former caste people. Judicial deference to caste autonomy on the question of caste membership has not been extended to the question of whether someone is a Hindu.) The tendency to fuse the caste mem-

bership question with that of Hinduism need not derive from such deliberate avoidance of the constitutional question; more likely it flows from the same. view

of caste that underlines the “‘loss of caste” theory.

The Court not only avoids the question of the constitutionality of the Hinduism requirement, it supplies a rationale for that test which elaborates on the “loss of caste” theory. It is because orthodox Hindu society. . attaches social and economic disabilities to a person belong to a Scheduled Caste ... that . . . favorable treatment is

given to him by the Constitution. Once such a person ceases to be a Hindu

and becomes a Christian, the social and economic disabilities arising be-

cause of Hindu religion cease and hence it is no longer necessary to give him protection and for this reason he is deemed not to belong to a Scheduled Caste. But when he is reconverted to Hinduism the social and economic disabilities once again revive and become attached to him because there are disabilities inflicted by Hinduism. A Mahar or a Koli or a Mala would not be recognized as anything but a Mahar or a Koli or a Mala after reconversion to Hinduism and he would suffer from the same social and economic disabilities from which he suffered before he was converted to another

religion.224

In this version, it is not that caste is lost by conversion but that the

disabilities attendant upon it are dissipated. This may be as much a fiction as is the “loss of caste” theory. But merely to set it out raises some puzzling questions about the Court’s combination of caste autonomy

doctrines with preference policy. For if disabilities impinge because of the response of “Hinduism,” then why should the revival of protections be dependent on acceptance by the caste group? The caste autonomy doctrine serves nicely where individual intention and conduct and group acceptance are reinforcing. But suppose 221. Id., at 949.

Group Membership

333

they diverge? Should group acceptance outweigh intention and conduct? In the Jasani case the test for retention of membership in the privileged group was primarily one of the convert’s intention. However, it was necessary that this intention be confirmed by acceptance by the old group. The inclusion of this additional test is important because, as the Court says, the only modification here is that it is not only his choice which must be taken into account, but also the views of the body whose religious tenets he has renounced, because here the right we are considering is the right of the old body, the right to send one of its own fold to Parliament.?22

Since the general electorate choose from among the candidates stand-

ing for the reserved seat, the only way that representation of the privileged group can be assured is by seeing that only those accepted by it are permitted to stand for these seats. But since seats are reserved not for particular Scheduled Castes but for the Scheduled Castes as a whole, one might question the relevance of looking specifically to his

own community to see whether he was accepted. For example, if the Scheduled Castes comprised As, Bs, and Cs, and X’s membership in

the As is in doubt, whether X is a member of the Scheduled Castes

might depend equally not only on whether the As accept him as an A but on whether the Bs and Cs do so as well or instead. Unlike legislative representation, such benefits as scholarships, housing, and agricultural loans do not devolve on the group in any corporate

or collective capacity; membership in the group is merely a convenient

device for identification of deserving beneficiaries. Thus, if the preference in question devolves not on the Scheduled Castes collectively but

on their members individually, the rationale for giving greater weight to group acceptance would be absent. The primary test of the intention

and conduct of the individual would apparently be conclusive, and it would not be necessary to confirm this by acceptability to the group.

Much the same may be said of the situation in which affiliation and group acceptance are most directly in conflict—that of expulsion or

excommunication. What effect should expulsion or excommunication from the group have 6n ‘eligibility for preferential treatment? The

Constitution leaves unchanged the power ofa caste group to excommunicate those who offend it.?23 In religious denominations—and 222. (1959) S.C. R. at 839.

:

223. Section 9 of the Civil Procedure Code forbids the courts to take cognizance of

“caste questions.” It is still a good defense to a criminal action for defamation to assert the privilege of communicating news of an excommunicaton to one’s caste fellows

(Panduram v. Biswambar, A. 1. R., 1958 Or. 259; Varadiah v. Parthasarathy, 1. L. R. 1964 [2] Mad. 417. Cf. Hadibandhu v. Banamali, A. 1. R. 1960 Or. 33). The power to excommuni-

334

Identifying the Beneficiaries

these include at least some caste groups?24—this right of excommunication is constitutionally protected in so far as it is used for the enforcement of essential religious discipline.??5 In spite of the language in

Rajagopal that the caste is the “‘supreme judge” of its membership, it seems doubtful whether ostracism or excommunication would effec-

tively deprive one of membership in the group for purpose of qualifying for benefits that were not directed to the group collectively. In such a

case the relevant question would be whether the expulsion was associated with any conditions which effectively dispelled the disabilities

or other backwardness which had caused the group’s members to be

singled out for preferential treatment. Presumably, if the individual still suffers these disadvantages he would remain a member of the privileged

group in the sense of being a legitimate recipient of these preferences. So

while the group’s opinion of the membership of the individual should be decisive in the case of preferences that devolve on the group collectively, where the preferences are distributive (i.e., devolve on the mem-

bers of the group individually), the State may legitimately intervene to protect the individual from group power.?26 The appropriate question would be whether his condition has changed from that which caused

members of the group to be given preferences.

Is membership merely a convenient device for identifying deserving

beneficiaries? It might be argued that preferences are directed to members of particular groups not only to benefit the individual members directly, but to utilize the ties of kinship, loyalty, and cooperation

among

them

to share and multiply these benefits. But while group

cohesion is an instrument of welfare policy, it is not necessarily an objective of that policy or a necessary condition for its effectiveness.

Preferences which include among their objectives the promotion of the

cohesion and integrity of the group must be distinguished from those cate is subject to two important restrictions. It is an offense to undertake any disciplinary action directed toward the enforcement of “untouchability” (Untouchability

Offenses Act, §7). Itis a corrupt election practice to threaten a candidate or elector with

excommunication (Representation of the People Act, 1951, §123 [2]). 224. See e.g., Sri Venkataramana Devaru v. State of Mysore, A. 1. R. 1958S. C. 255. 225. Saifuddin Saheb v. State of Bombay, A. 1. R. 1962 S. C. 853. Some caste groups among the Scheduled Castes or Backward Classes might regard themselves as religious

denominations and invoke their constitutionally protected right to ¢xcommunicate for the purpose of upholding religious discipline. However, since preferences to these groups are for the purpose of ameliorating their condition rather than promoting their

religious unity, such expulsion should not outweigh the right of the expelled individual

to share in these preferences—at least so far as his enjoyment of them does not detract from the religious freedom of the group.

226. This would seem analogous to the notion that excommunication is protected only so far as it withdraws religious rights and does not extend to deprivation of civil tights (except where consequent upon the former). See Saifuddin Saheb v. State of Bombay, A. 1.R. 1962 S. C. 853.

Group Membership

335

where group cohesion is used only as an instrument. It is clear that preferences for the Scheduled Castes and Backward Classes are for the purpose of ameliorating the deprivations of the lowly; they are not

designed to promote the cohesion and integrity of the groups that make up these categories. On the other hand, the preferences for Scheduled Tribes include among their purposes the promotion of the cultural intergrity

and

autonomy

of the tribal groups.?27

While

expulsion

should not outweigh the right of the individual to enjoy preferences

among the Scheduled Castes and Backward Classes, in the case of the

Scheduled Tribes the group’s view of the individual’s membership might be given greater weight.

If, as we have seen, individual affiliation and group acceptance may

determine who

remains a member or is restored to membership,

the

question arises whether affiliation and acceptance may determine who becomes a member. May someone who was not a putative member become a member of a Scheduled Tribe, a Scheduled Caste, or a caste

or religious group listed as a Backward Class and receive the preferential treatment extended to its older members?

In Jahan Ara we saw an instance of assimilation of an outsider into a

Scheduled Tribe.?28 In Rajagopal I the Supreme Court suggested that a caste might admit to its fold even those not born within it.229 Subsequently, a Constitution Bench of the Supreme Court faced the question in Guntur Medical College v. Mohan Rao.°° The aspiring medical student

was the son of Madiga parents who had converted to Christianity before he was born. Converts to Christianity from Scheduled Castes were treated by the State of Andhra Pradesh as members of the Backward Classes. When the applicant did not succeed in securing admis-

sion

to medical

college as a member

of the

Backward

Classes,

he

underwent conversion to Hinduism and, claiming to be a Madiga, applied for admission to another medical college as a Scheduled Caste. His provisional selection was cancelled on the ground that he did not meet this state requirement:

No candidate other than a Hindu including a Sikh can claim to belong to

Scheduled Castes. No candidate can claim to belong to the Scheduled Caste except by birth.231 227. See chap. 5, §C, above. 228. See discussion in §C.

229. See A. I. R. 1976 S. C. 939 at 948.

230. A. I. R. 1976S. C. 1904.

231. The Principal of the Medical College had misinterpreted this to require that the applicant be a Hindu by birth, an interpretation which the Supreme Court deemed “plainly erroneous” before proceeding to hold the requirement void as property interpreted. The grounds for this voiding seem to be that the power to define Scheduled

Castes is exclusive with the President and Parliament, and it is not open to other

authorities to apply more restrictive definitions.

336

Identifying the Beneficiaries

The Supreme Court found the “birth” requirement void as beyond the provisions of the Scheduled Castes Order.

Hence, it was open to the

applicant to show that he had become a Madiga upon his conversion to Hinduism. He claimed that his parent remained Madigas even after

their conversion, Court is willing conversion is not of the caste and

so to an its

he was in fact a Madiga by birth—a possiblity the countenance, since it says that loss of caste upon absolute rule, but is dependent upon “the structure rules and regulations.”32 But the Court is equally

willing to contemplate the possibility that even if his parents lost membership in the Madiga caste upon conversion, and the applicant

was not a Madiga by birth, he might have become a Madiga.

It is for members of the caste to decide whether or not to admit a person within the caste. . . .It may if its rules and regulations so provide, admit a new member just as it may expel an existing member. The only requirement” for admission ofa person as a member of the caste is the acceptance of the

person by the other members of the caste.?33

-\lthough there is no finding that that is the case here, and althongh it is said in reference to a child of former members of the caste, the Court

does not limit this rule to those with a history ofdescent from the caste in

question. Presumably, it is applicable to someone who is by family history a total stranger to the caste.234

The rejection of the “birth” requirement raises the question of how preference policies will be protected against exploitative assertions of group identity. It is not surprising that there have been attempts to effect changes in personal status for the purpose of establishing eligibility for preferential treatment.?35 Earlier we noted some instances of deceptive

self-characterization

by individuals and groups, often in

marginal and ambiguous cases.236 Here, however, we are dealing with 232. Id., at 1907. 233. Id., at

1908. The Court

gives no indication of whether the caste’s “rules or

regulations” allowing admission need be long established or whether freshly minted

ones would suffice. 234. Cf. the further developments noted below in text at notes 251 ff. 235. Although documented cases of such deception are not plentiful, it is perceived in

many quarters that such deception is widespread. hus, a venerable member of the Rajya

Sabha complained that “there are many IAS and IPS officers. They change their names or add some Scheduled Caste or Scheduled Tribe names and come into this list and get jobs thereby, cheating the Government.” Specifically, he complains, in Andhra Chris-

tians with superior access to education, “by calling themselves Mala or Madiga or

Chamar are getting all the facilities and they are coming into power. His father is a

Christian; his grandfather is a Christian; his religion is Christianity; he goes to Church everyday; he marries a Christian girl. But for the purposes of these benefits, he calls

himselfa Scheduled Caste man. Many such cases are there. Everyday it is happening” (RSD Vol. 97, No. 18, col. 44 [3 Sep. 1976]). 236. See §A, above.

Group Membership

337

deliberate and explicit change in status. This rather drastic stratagem has been adopted in regard to those very scarce and critical opportunities, medical and professional school admissions. It seems to take two

forms: (1) conversion to a religion whose members are entitled to reserved places (thus we have, in addition to reconversion to Hinduism,

instances reported of conversions to Buddhism in Maharashtra and to Islam in Jammu

and Kashmir);?37 (2) high caste Hindus securing

adoption in Scheduled Caste families.238 In Ashok v. Dean, three, higher caste students had embraced Buddhism and described themselves as Nava-Buddhas for purposes of

securing admission to a medical college.7° (The state’s list of Backward Classes included Nava Buddhas, which it defined as “Scheduled

Castes converted to Buddhism.”) The court found them ineligible and ordered their admissions withdrawn on the ground that Nava-Buddhas meant Scheduled Castes converted to Buddhism “and not just any one belonging to any caste whatsoever who has embraced Buddhism.”2

Here the court managed to resolve the matter handily, with an assist

from the Government Order, but it might well have reached the same result unaided. The adoption stratagem has proved more difficult to resolve. In 1968 there were reports that students in Delhi and elsewhere in nothern India were securing admissions by having themselves adopted by

237. In September 1966 it was reported (Organizer, Sept. 11, 1966) that eleven young Kashmiri Hindus had converted to Islam on the ground that their religion was an

obstacle to obtaining technical education and suitable jobs. According to the Organizer (Oct. 30, 1966), “Thousands of letters poured in on the converts from far and near. Arya

Samaj, RSS, Dogra Sabha, Kashmiri Pandit Sabha leaders met the eleven misguided youths and impressed on them the impropriety of changing their religion for worldly

considerations. Union Home Minister Nanda is . . . believed to have written to the Kashmir Government that the students should not be given admission just because they

had changed their religion. After all this persuasion the eleven youngsters came back to the Hindu fold on Oct. 23. The shuddhi ceremony was performed in the Arya Samaj Mandir Srinigar.” 238. These are courses that go beyond the mere procurement and use ofa false “caste

certificate.” Such certificates, attesting to membership in a Scheduled Caste or Scheduled Tribe, are issued by various officials—upon production of a document purporting to be from a

caste organization (attesting to the membership of the applic-

ant) and a certificate from a special executive magistrate. The three metropolitan

magistrates courts in Bombay, for example, are reported to issue “nearly 10,000 caste

certificates every year [but] do not have the machinery to verify their authenticity”

(Times of India, 24 June 1977). There are numerous reports of the fraudulent use of such

certificates to secure admissions and jobs, promotions, loans, and other benefits—e.g., CWSCST (5th) 21: 33-35 (1973). The practice is parallel to the false procurement of income certificates discussed in §G, below.

239. Ashok v. Dean, 69 Bom. L. Reporter 603 (1967). 240. Id., at 612.

338

Identifying the Beneficiaries

Scheduled Castes.%41 The Elayaperumal Committee reported that in Rajasthan, non-Scheduled Caste students had themselves adopted by Scheduled Caste persons, then obtained certificates from magistrates, on the basis of which they were admitted to engineering and medical colleges against seats reserved for Scheduled Castes. The Committee

reported that in 1968-69, of the 28 Scheduled Castes’ seats in medical colleges, 16 were occupied by non-Scheduled Caste candidates on adoption certificates.242 One response to these abuses is a “‘birth” test, an option now closed by the Supreme Court.43 What alternative

means are there for checking exploitative acquisition of membership in

beneficiary groups? In response to inquiries, the Department of Social Welfare provided “legal advice” notable for its formalism and lack of resourcefulness:

[T]he legal status of a high caste child adopted by a Scheduled Caste family

would make him eligible to treatment as a member of the adopters’ caste if the adoption satisfied the conditions laid down in the Hindu Adoptions and Maintenance Act, 1956. In the case of Scheduled Tribes, however, the said enactment does not apply and the guiding factor in such cases will be

whether the customs of the concerned tribe permit the adoption of a nontribal into the tribe and whether the particular adoption in question is covered by the customs.244

This advice seems at variance with the Rajagopal notion that the caste is the ‘‘supreme judge” of its membership, which would seem to imply that group acceptance rather than the formalities of adoption law would be decisive. Castes could admit or confirm new members according to their lights. But while Rajagopal requires some group confirmation, the 241. RCSCST 1967-68: 33; Dept. of Social Welfare 1969: 207.

242. Department of Social Welfare 1969: 207. Most reports of fake adoptions concern

fessional admissions, but such practices are not unknown in regard to posts. See the report ofa Brahmin who had himself adopted by a Scheduled Caste and appeared in the I.A.S. examination (RCSCST 1970-71: 127).

243. E.g., the Ministry of Labor instructed employment exchanges to register as Scheduled Castes reconverts to Hinduism but not converts who were born ina different religion “whatever may be their family connections” (CWSCST [4th] 5: 27 [1970]. See, generally, the Circular sent by the.Govenment of India to all state governments that they should admit to reserved seats in medical colleges only those “born in a Scheduled

Tribe/Scheduled Caste family” (CWSCST [4th] 14: 19 [1970]).

These “birth” requirements have been around long before the recent concern with exploitative acquisition of membership in beneficiary groups. E.g., in 1953 the Govern-

ment declared that upon conversion to Hinduism, those born in another religion (whose Parents were converts) should not be treated as members of the Scheduled Castes

“whatever may be their original family connections” (RCSCST 1953: 132). Reconverts born as Hindus, on the other hand, have been treated as Scheduled Castes. 244. RCSCST 1967-1968: 33.

Group Membership

339

departmental advice seems to accept what amounts to delegation to a single individual of power to confer membership— often, one assumes, an individual who is consulting his own interest rather than group

interests.

Like Rajagopal, this advice reflects the heavy sway of the notion of unitary group membership, according to which an individual is a

member of the group either for all purposes or for none at all. Such a notion is not only a deceptive guide to the social realities; it is an obstacle to rational policy-making. Why should the group’s decision for its purposes (much less an individual Scheduled Caste member’s for his) necessarily bind the government in its administration of its policies of preferential treatment? The government should not be allowed to attach conditions to inclusion which are not relevant to the carrying out of these policies.245 But surely it may enquire whether inclusion of a new member would effectuate the particular policy involved. Where dealing with such collective preferences as political representation (or

with Scheduled Tribes, where group autonomy is an aim of government policy) group acceptance might well prevail. On the other hand, in dealing with such benefits as scholarships, housing, etc., the government might well enquire whether the new member had such an affinity of conditions and resources that it would carry out the aims of the particular program to include him. The Mysore High Court employed such a policy analysis to fashion an incisive solution to the adoption problem in Shantha Kumar v- State of Mysore,?46 a case which reminds us that exploitative manipulation is at least as possible with “neutral” categories as with conmmunal ones.

Mysore had adopted an “income cum occupation test” for Backward Class status. Medical admissions were reserved for candidates whose parents were in manual occupations and had incomes below a specified amount.47 Shantha Kumar, the son of a supervisor in the National Extension

Service, was given in adoption

to a coolie (his mother’s

sister's husband?). The adoption supposedly took place three years earlier when the boy was sixteen, but the deed of adoption was executed

just three months before the application to medical college. The Selection Committee refused to regard him as a member of the Backward 245. E.g., the “birth test” for reconverts described in n. 243 above. Since reconverts

born as Hindus are treated as Scheduled Castes, Hindu birth is made a condition for governmental recognition of membership in a Scheduled Caste. 246. 1971 (1) Mys. L. J. 21. Another bench of the Mysore court came to a similar result in Nataraja v. Selection Committee, 1972(1) Mys. L. J. 226 [NIC]. There a candidate

for medical college, the son ofa health inspector earning Rs. 2,000 per year, had been adopted by his mother’s sister. a tailor with an annual income of Rs. 780. 247. The Mysore income-cum-occupation test is disussed at chap. 8, §G, above, and in §G, below.

340

Identifying the Beneficiaries

Classes, and he petitioned the High Court to overturn that decision. The court deliberately suppressed its doubts about the genuineness of the adoption and decided to “proceed on the assumption that there was a genuine and valid adoption of the petitioner’’2*8 at the age of sixteen.

The court observed that the rationale for reservations for Backward

Classes was to offset the unfavorable environmental conditions of such persons. But the applicant had not suffered any environmental disadvantage while living with his natural father for his first sixteen years, and the environmental conditions of the last three years ‘‘cannot be said to destroy or nullify the advantage”49 conferred by his earlier up-

bringing. In the case of a late adoptee who “had all the while imbibed

the

.. . environmental advantage of his natural father’s income and

occupation,"’259 it is the income and occupation of the natural father rather than

of the adoptive

father that is relevant in determining

eligibility for benefits for the Backward Classes. Otherwise, adoption could be used to divert benefits from the intended beneficiaries.

This line of analysis produces a similar result even where the affiliation with a beneficiary group is in undoubted good faith. Thus in Urmila

Ginda v. Union of India, a woman of high caste origin married to a Scheduled Caste man applied for a reserved post and claimed that by the Hindu theory of marriage she had become his Sapinda and caste

fellow.251 The court put aside these “spiritual .. . considerations” and held that as one who was not personally handicapped she could not take advantage of the special provisions contemplated by Articles 15(4)

and

who

16(4) ‘because she is not one of them—she is a high caste Hindu

was

not subject

to any such

educationally.”25? To permit her of such provisions by diverting neficiaries and open possibilities Jahan Ara as a case of assimilation

backwardness

either socially or

to compete might defeat the purpose benefits away from legitimate beof abuse. The Court distinguishes into a tribal group‘253 Urmila Ginda is

devoid of any assertion of assimilation, affinity of conditions, or accept-

ance by the husband’s group.

248. 1971(1) Mys. L. J. at 22. 249. Id., at 23. 250. Id., at 23. 251. A. I. R. 1975 Del. 115. During the Lok Sabha debate on the revision of the Scheduled Castes Order, one M.P. opined that it would be redundant to legislate Scheduled Caste status for intermarrying wives, since this would “happen automatically without anybody's intervention” because women in India took on their husband's social identities (Shri G. Kuchelar, LSD [4th Series], Vol. 45, No. 9, col. 270 [20 Nov. 1970}).

25), Id., at 116. An attempt to reverse this result by statute failed. See n. 52 above. 253. See the discussion of Jahan Ara in §C, above. The court here implies that the words “tribal communities” in Art. 340 provide more leeway for assessments of membership in cases of Scheduled Tribes than with Scheduled Castes (id., at 116).

Group Membership

341

Subsequently, the Punjab and Haryana High Court encountered a case in which the missing element of caste acceptance was supplied. Not

only had a medical college applicant been adopted by a member of the

Scheduled Castes, but the Baradari [Assembly] of the caste had indicated its acceptance of him as a member. The court declined to read the

Guntur Medical College case as extending the acceptance test to those with

no historic connection with the community, for to do so would ‘open

the floodgates for all sorts of undesirable practices to gain narrow

ends.”’254 The floodgates argument

is less persuasive in the Urmila

Ginda situation than in the adoption case, since it is established public policy to foster intermarriages with Scheduled Castes.?55 Although Government and Parliament declined to set a rule of automatic inclusion of wives,25° this does not imply that they are automatically excluded.

Instead, it leaves a question of fact, to be decided in each

instance according to the tests of intention, group acceptance, and the policy of the particular scheme of preference. One might imagine policies (like health care, drinking water, provision of agricultural loans)

in

appropriate.

which

the

inclusion

of such

,

wives

would

G. NON-COMMUNAL CRITERIA: INCOME OCCUPATION TESTS

clearly

be

AND

Since membership problems of the-kinds discussed above are a by-product of the use of communal units to identify the beneficiaries of preferences, it is sometimes anticipated that the replacement of com-

254. This case (Civil Writ Petn. No. 3970 of 1976) is described and an excerpt from the judgment reprinted at RCSCST 1975-1977: 19. 255. For examples of various state programs of gifts and grants to couples entering such marriages, see RCSCST 1973-74: 184; cf. the account of the public reception in

Ahmedabad at which Prime Minister Desai presided over the honoring of eighteen

couples entering mixed (Harijan-caste Hindu) marriages. The state gave each couple an award of Rs. 5,000 and undertook to give accorded various benefits to the offspring of these marriages (Overseas Hindustan Times, 8 June 1978).

256. The Joint Parliamentary Committee set up in 1967 to revise the Scheduled Castes

and Tribes Order recommended inclusion of wives. The provision in the 1970 bill: “Notwithstanding anything.contained in any law . . . or in any custom on usage to the contrary, a woman who marries a person belonging toa caste specified in any partof the Schedule shall be deemed to belong to the Scheduled Caste to which her husband belongs,” was greeted by a Government spokesman to be a “heartily welcome . . . encouragement for Savarna girls to marry Harijans” (Shri K. Hanumanthaiya, LSD {4th Series], Vol. 65, No. 6: 322 [17 Nov. 1970]). When the bill reappeared in 1976, this

provision was removed from the Government version, and efforts to restore it from the floor were unsuccessful.

342

Identifying the Beneficiaries

munal criteria by economic and educational tests of backwardness would eliminate such disputes. But even criteria like income, occupation, and place of residence, which may seem subject to rather mechanical determination, raise a host of problems that are quite analogous to the membership disputes that we have reviewed. Movement to noncommunal criteria will change the nature of the disputes, but it will not eliminate disputes about who is included and who is not. Income tests immediately raise the question of whose income is to be

measured. Typically, the requirement is in terms of family income—the income not of the applicant himself but of some larger unit. But what is

that unit? those in certain biological relationships? those who are legally a joint family? those who share a common pocket? those who live under one roof? In Jayaswal v. Principal,?57 an applicant for medical college admission declined to supply information about his father’s income and instead filed his own income certificate, showing that he was separated

from his father as a result ofa suit and that his annual income was below the maximum of Rs. 3,000. The Patna High Court upheld the rejection of his application because of omission of a certificate of his father’s

income.?58

In Shetty v. State of Mysore,259 the High Court held that even more

than the father’s income might be taken into account. Medical college

seats were reserved for those who met an occupational test and whose

annual income was less than Rs. 1,200. A candidate listed his father’s

income as Rs. 348.20—a one-thirteenth share of joint family income.

The court held that “therefore his family income was Rs. 4,526.60, which

exceeds

Rs.

1,200—the

maximum

prescribed

by

the

Or-

der.”26 The court does not explain why it regards all members of an undivided joint family as a “family” for purposes of the Order, but it hints that it is because ‘“‘the social status ofa member of a joint Hindu family’’261 is related to the income of the joint family. But surely not all who share a common status are members of the same family. The Order itself seems to visualize somewhat smaller units than the family in-

volved here (13 co-parceners and presumably a comparable contingent

of females), when it explains that the Rs. 1,200 income ceiling was determined on the gound that the annual per capita income in the state

in 1961 was Rs. 266 and that the average family consists of five members. It was clearly a different family that they had in mind!?6 257. A. I. R. 1968 Pat. 504.

258. The survey cited in n.265, below, indicated that false declarations of separation from a joint family were about a ninth of all false certificates.

259. A. I. R. 1969 Mys. 48.

260. Id., at 50.

261. Id., at 51.

262. The absence of any sliding scale for family size tends to favor small families,

perhaps a laudable policy but one never avowedly pursued in this setting.

Group Membership

343 In practice, the conceptual problems of identifying family income are overshadowed by the formidable problems of accurate assessment of income. The difficulties offered by joint families, income in kind rather than cash, and so forth are compounded

by the widespread

inclination to understate income. The absence of income and tax data on most families had led the government to ascertain income by requiring production of a certificate signed by a government officer. The officers vary from place to place and scheme to scheme,?® but generally they are highly placed and invariably have no way of ascertaining the income of the person for whom they sign.?6 There is no way of knowing how many of these certificates are false. In a 1964 sample investigation of 15,438 beneficiaries in one state, it was found that 4,491 (29.1%) had given false statements of income. False income statements ran as high as 58.9% in one of the four districts sampled; none had less

than 5.6% false statements.265 A few years after the survey just mentioned, the Tamil Nadu Backward Classes Commission remarked of that state’s experience with income certificates:

[I]t is impossible to check the income declaration . . . of people who derive their income from land or small business; and those who are self-employed succeed generally in filing satisfactory income certificates. Only those whose

opulence is known ... do not attempt to secure scholarships for their children by questionable means.?6 263.

Many such provisions are like that of Gujarat. which required a certificate

from “ta Revenue Officer not below the rank of Deputy Tahsildar” (Gujarat, Administration Report of the Social Welfare Department [Backward Class Section] 1963-64: 75)

Some states accept certificates from a wider range of officials. Thus in Madras the certificate must be “issued by a gazetted officer, an Inspector of Police or a non-gazetted officer of the Revenue Department not lower than a Deputy Tahsildar or SubMagistrate” (Govt. of Madras, 1962: 29). The anonymous study referred to below found that certification by revenue officials tended to be more accurate than that of gazetted officers, while certificates issue by honorary magistrates were least accurate.

264. Thus in Nataraja v. Selection Committee, 1972(1) Mys. L. J. 226, 227 [NIC], the

court noted that the machinery for determining income and occupation

“has given

ample scope for manipulations. We have noticed that certificates of backwardness have

been issued by Tahsildars indiscriminately without proper investigation into the bonafides of the claims.” 265. I am not at liberty to disclose the provenance of this study, but the state

concerned enjoys a reputation of having one of the most efficient administrations in

India. Cf. the observations of the Punjab Evaluation Committee (Welfare Department

1966: 141) that “with the adoption of economic criterion, there is a tendency among the

people to get themselves falsely declared as backward.”

266. Tamil Nadu Backward Classes Commission 1971 : 77. Irschick (1977) notes that

from the time poverty certificates were introduced in Madras in 1921, they provided abundant “‘room for local manipulation.”

344

Identifying the Beneficiaries

Such practices were so prevalent that [i]t is generally conceded that a very large percentage of people definitely above the income limit fixed are able to obtain scholarships by understating

their income.267

But of course

opportunities

for such dissimulation

are not spread

evenly. They are least available to those like schoolteachers and inferior

government

servants who receive fixed and known salaries.2 The

Tamil Nadu Commission argued in favor of special treatment of lowpaid government officers, noting that among other problems they “suffer from the handicap . . . that they cannot . . . suppress their monthly salary and produce false income certificates.”2

Where income is ascertainable and accurately reported, there may be

the additional problem of fluctuating eligibility, dramatically displayed in Shameem v. Medical College.27° The State of Kerala had established an

income cut-off of Rs. 6,000 on membership in the Backward Classes for purposes of medical admissions. The income of applicant’s father, a small businessman, was Rs. 7,020 for the most recent year and Rs. 3,710

for the previous year. Although the average for the two years was

Rs. 5,365, comfortably below the cut-off line, she was ruled not eligible.

Here fluctuations, surely not unusual, give the whole arrangement a curious lottery-like quality. A student might be excluded in a singularly good year; another, might gain inclusion as a windfall in a singularly

poor year. These problems could obviously be addressed by provisions

for averaging, at the expense of more record-keeping and more possiblities for manipulation. But even if information was accurate and the fluctuation problem

was solved, income tests have another peculiar arbitrary quality when applied to a non-divisible benefit like a medical admission. When applied to a divisible good like a stipend, the incremental nature of income makes it a measure that has the advantage of. allowing benefits to be calibrated to need or desert. Benefits can increase as income falls. But income has a very different character when used to distribute a

non-divisible good like medical admissions or reserved posts. An eligibility measure that is continuous and.incremental is subject to a single cutting point. The difference between Rs. 5,999 and Rs. 6,000 becomes the difference between inclusion and exclusion. One might, of course, 267. Tamil Nadu Backward Classes Commisson 1971: 160. 268. The anonymous study referred to above found that false statements of income

were more prevalent among agriculturists and the self-employed than among employed

ns. 269. Tamil Nadu Backward Classes Commission 1971: 77. 270. A. I. R. 1975 Ker. 131, rev'd A. I. R. 1976 Ker. 54.

Group Membership

345

import the incremental virtue of income as a measure into the selection process by according graduated weightage in proportion to low income. But so long as eligibility for benefits is an all or none affair, income tests create border lines that are difficult to justify as well as to ascertain. Like

communal tests, income tests import their own brand of arbitrariness

into the selection process. Occupation tests also may give rise to many membership problems.

A series of Mysore cases illustrate some of the complexities. The Mysore

order specified the following occupations as contributing to social back-

wardness: “(1) Actual cultivator; (2) Artisan; (3) Petty businessman; (4) Inferior service (i.e., Class IV in Government service and cor-

responding class or service in private employment) including casual labour, and (5) Any other occupation involving manual labour”.271 First, there is the simple problem of the definition of these broad categories. For example, who is an “actual cultivator” or an “artisan” is not always self-evident.272 What is an “occupation involving manual labor”? An applicant for admission to a medical college was rejected on the occupation

test when

she reported

her father’s occupation

as

purohit (i.e., family priest), who performed rituals manually.?”3 The court held that this was not an occupation “involving manual labor.” It was necessary to look at the predominant character of the occupation:

Though a purohit may use his hands in performing certain rituals and ceremonies the predominant character of his occupation is that it requires study and knowledge of scriptures and of the body of traditions and the performance of his work involves mainly chanting or recitation of “‘mantras” or scriptures.274

Second, there is the problem of the composite occupation where there are several sources of income, some from listed occupations and some

not. In Sridhara v. Director of Technical Education,2”5 the applicant’s father earned Rs. 800 per year, Rs. 660 as a part time clerk, and Rs. 140

by dealing in pencils and paper. He claimed that he was a “petty businessman” who “supplemented” his business income by working time as a clerk. In view of the figures, the court might have easily employed a rule of preponderance. But it chose to formulate a general

policy against inclusion of any person who engaged part time in un-

listed occupations. It finds the rationale of the Order in the notion that the environmental conditions which attend certain occupations contribute to backwardness, while others do not. Ifa person follows both,

271. The order is reproduced in full in Shetty v. State of Mysore, A. I. R. 1969 Mys. 48.

272. A “mechanic” is an “artisan” (Pasha v. Stateof Mysore, 1969(1) Mys. L. J. 79). 273. Sudha v. Selection Committee, A.1.R. 1967 Mys. 221

274. Id., at 222.

275. 1966(2) Mys. L. J. 774.

346

Identifying the Beneficiaries

it cannot be said that the advantage of following that latter category of occupations is destroyed or nullified by the same person following, in addition, an occupation of the former category. Having regard to the objects sought to be achieved by this classification, namely to help those who are at the bottom rung of the ladder of social and economic development . . . it is

only reasonable to construe the Government Order as providing that only

those who follow exclusively the listed occupations come within the Backward Classes. Any other interpretation will bring into the Backward Classes

a large number of persons who do not suffer from serious environmental disadvantages. . . .276

While this rule has the virtue of definiteness, it is difficult to see that the

object is effectuated, for example, in a situation the the Shridhara case. Suppose the father made Rs. 660 and added Rs. 140 from part-time clerking. Does prestigious occupation automatically carry with

reverse of the one in from selling pencils any income from a it the disqualifying

environmental advantages? Such a rule, if strictly applied, would be a

disincentive to take up part-time employment which might deprive one’s children of enhanced educational opportunities. Third, there is the problem of change over time.?77 In Shetty v. State of Mysore,278 the applicant’s father was a retired schoolteacher who was now engaged in agriculture. The High Court held that the occupations

contemplated by the order

are not casual or temporary occupations, but the habitual occupations of

families, otherwise a person may claim to be treated as backward one year and forward in another as suits his advantage.279

it was common knowledge, said the court, that a schoolteacher even after retirement was held in the same esteem, so he could not be

regarded as a member ofa Backward Class. Fourth and finally, there is the problem of whose occupation is to be measured. As in the case of income, it is generally assumed that it will be the father’s. But what if there is a joint family and some members are engaged in higher-status occupations? In the Shetty case, the court insists that the family is a unit for purposes of backwardness. “It would not be possible

to envision

a situation

in which one is treated as

backward and another is not.”289 But suppose you have a joint family 276. Id., at 777. 277. In what must

be regarded

as the reductio ad absurdum

of the mechanical

application of the occupation test, a student whose mother was a coolie was rejected

because his mother was ill and therefore had no occupation. This was reversed by the

Court in Viswanath v. Government of Mysore, A. 1. R. 1964 Mys. 132 at 144. 278. A. I. R. 1969 Mys. 48. 279. Id., at 50-51.

280. Id., at 51.

Group Membership

347

as in Shetty with thirteen co-parceners. Are all the members of a family disqualified from being backward if one of the thirteen earns some money in a higher-status occupation? With family boundaries and members’ occupations as variables, the possibilities for producing the

indicia of eligibility are not negligible.28 Unlike the manipulation of

communal identity, these expedients may be resorted to without exciting reproach by self or others. . What is striking in these cases is that, contrary to what might have been anticipated, judgments about status seem at least as tangible and determinate to the courts as do facts about economic standing. Where the economic facts appear doubtful or subject to interpretation, the

court resolves the ambiguity by turning to assessments of status, and it

experienced no difficulty in judging, e.g., the status ofa clerk, a purohit, a schoolteacher, or a joint family with Rs. 4,000 income. The economic tests tend to be resolved into selection-on the basis of status.?8? But if this is so, it is not clear that reported income/occupation is really a better measure of status than communal identification. Courts have been relatively indifferent to the disincentive effects of

these rules, or more precisely the choice they pose between disincentive to mobility and positive incentive to dissimulation. A family whose son aspired to be a doctor or who anticipated any other tangible boon from

governmental preference would of course be well advised not only to

keep reported income from rising above, e.g., Rs. 1,200, but even at

lower levels of income to avoid employment that would jeopardize their occupational standing. But these are largely fictional. In reality, the

difficulties have been of overinclusion rather than disincentives. Of course,

these problems

in administering economic

tests have

their parallel in the administration of communal tests. Apart from cases

of dissimulative identification, there were a host of borderline cases in

which issues of group membership were litigated when a challenger with a heavy personal stake sought to wrest a valuable prize by establishing his eligibility or disqualifying a competitor. Disputes about

economic tests are different in several ways: compared with community, income is not likely to be a matter of general knowledge; compared with community there is little compunction about or stigma attached to dissimulation of income; and finally the categories, malleable and fluid, not just at the border lines, are less established in common experience.

The cases that reach the courts are not unusual contests about dis-

281. For an unsuccessful but instructive attempt, see Shantha Kumar v. State of Mysore,

1971 Mys. L. J. 21, discussed in §F, above.

282. Cf. the observations of Sabberwal (1972: 74) on the tendency of administration to employ status as an indicator and rationing device in place of determinations which

would require resources and competence to make.

348 Identifying the Beneficiaries puted border lines, but cases of ordinary applicants testing the margins of routine administration. Were there serious attempts to monitor economic tests, there would be many more cases of this kind than there

have been in the communal group membership area. On the other hand, there is a greater possibility of avoiding such litigation by more careful, refined, and definite formulation of the administrative rules.

This would seem to be an area in which courts could contribute to solutions on the administrative level by pressuring the administrators to eliminate some of the gross inequities that pervade the rules. However, so far no litigant has found a way to provide the courts with a

handle for exerting such control.

H. CONCLUSION: PRAGMATIC AND FORMAL SOLUTIONS The judicial treatment of group membership problems displays two divergent tendencies, each with some support from the Supreme Court. The first, which might be called the empirical or pragmatic approach, is represented in the Jasani, Shyamsundar, Kartik Oraon, Wilson Reade, and

Jahan Ara cases.283 The second, which might be called the formal or fictional approach is represented in the Dora case and in the cases dealing with the “Hinduism” test for Scheduled Castes.28* Both approaches recognize the “‘compartmental” nature of Indian society. But the empirical approach is willing to give recognition to the areas of blurring and overlap that are found within it. The formal approach, emphasizing theoretical symmetry, tends to picture the society as one of mutually exclusive and hierarchically ranked compartments; where in fact individuals straddle compartments, the Court sees its task as assigning them to one or the other. The empirical approach does not share this notion of resolving the ambiguity into a single identification but is congenial to multiple and overlapping affiliations; it addresses itself to whether, in the light of the policy of the particular legislation involved, the individual can be said to be a member of the group concerned.

The formal approach concentrates on the theoretical consequences of

certain acts: one who attains caste status loses his tribal affiliation; one

who declares himself a member of a non-Hindu religion loses caste membership. The empirical approach pays less attention to such theoretical incompatibility and gives greater weight to the facts of intention and acceptance. Thus sect members can retain their caste membership, an Anglo-Indian or a Tamil Christian can become a tribal, a convert can remain a tribal.

283. See §§B and C, above.

284. See §§D and E, above. Cf. the Saha case, §A, above.

Group Membership

349

The empirical and formal labels are not entirely satisfactory, for both

approaches are formal in the sense of being selective and abstract, and both are empirical in the sense of rendering the most accurate picture of Indian social reality as it is experienced. But that reality is so vast and complex, and any individual experience of it so partial and fragmentary, that there is no single correct way to render it. The two approaches I have distinguished differ in which aspects of that experience they find most salient. What I have called the empirical approach emphasizes the group affiliations as experienced by (at least some of) the participants at the particular social locations in question; the formal approach tends to rely more heavily on perceptions of these affiliations by observers at distant (often hierarchically superior) locations in the society.

These perceptions, mediated by categories of hierarchy and exclusivehess, are more evident in dealing with the Hindu caste order than in dealing with tribal groups. Problems of acquiring or losing caste status tend to trigger bewildering and often unpersuasive conceptual acrobatics. In cases of membership in tribal groups, the courts seem to experience less difficulty in aligning entitlements with the factual con-

tours as experienced by the participants. As the Court in Jahan Ara says:

When a person in the course of time has been assimilated in the community

it is somewhat difficult to comprehend how that person can be denied the

rights .and privileges which may by constitutional provisions.7*5

be conferred on that community

...

Is one of these approaches preferable or constitutionally incumbent upon the courts? The empirical approach is clearly more consonant with current knowledge of Indian society. It sheds the encumbrance of theories of caste which reflect perceptions of Indian society current several generations ago. Modern students of Indian society have modified older notions of the caste system, which emphasized vara and

inflexibility, and have achieved a new understanding of its complexity, which includes its local variability, the ambiguity of caste-ranking, the existence of mobility, and the limitations of vara theory.28 The courts have long recognized the deficiencies of the older notions.287 Now that 285. A. I. R. 1972S. C. 1840 at 1850, The Court notes some additional textual support for this pragmatic approach in the tribal setting whenit i commends the view of the High Court that Art. 342’s mention of “tribal communities” as well as “tribes” evidences an awareness that those who are not members of a tribe in the strictest sense may by assimilation become members of the tribal community.

286. See Mandelbaum 1970; Fox 1969; Singer and Cohn, eds., 1968 Rudolph and Rudolph 1967; Srinivas 1962 and 1966; Karve 1961; Marriott 1959. 287. On the limitations of the vara model, see §D, above. For instructive discussion of

the complexity of the caste system, see Muthusami v, Masilamani, 1. L. R. 33 Mad. 342

(1909).

350

Identifying the Beneficiaries

courts are no longer hampered by the necessity of giving legal effect toa

picture of Hindu society consisting of the four vamas,?88 they may

appropriately employ the new perspectives to confront the actualities of

Indian society in order to implement the principles of the Constitution.

It is submitted that the empirical approach is also to be preferred

because of its consonance with the ideals and principles of the Constitution, while the formal approach contravenes these and should be discarded. The Constitution sets forth a general program for the recon-

struction of Indian society. In spite of its length, it is surprisingly undetailed in its treatment of the institution of caste and existing group structure in Indian society. But it clearly sets out to secure to individu-

als equality of status and opportunity,?8? to abolish invidious distinctions among groups,” to protect the integrity ofa variety of groupsp— religious, linguistic, and cultural, 9" to give free play to voluntary

associations,292

to confer

the widest

freedom

of association

to the

individual,?93 and generally the widest personal freedom consonant

with the public good.2% Without pursuing all of these in detail, it is

clear that the following general principles are consistently in evidence:

(1) a commitment to the replacement of ascribed status by voluntary affiliations; (2) an emphasis on the integrity and autonomy of groups

within society; (3) a withdrawal of governmental recognition of rank

ordering among groups.

.

In all of these. the empirical approach seems more congenial to the

constitutional design. Its emphasis on the actual conduct of the individual and the actual acceptance by members of the group gives greater play to the voluntary principle. The control this gives the group over

determination of its own membership seems implicit in the recognition

of the integrity of the group. Finally, it avoids the necessity of giving official recognition to the ranking of groups. The formal approach, on the other hand, severely limits the voluntary principle and the autonomy of the group by giving conduct unintended consequences on theoretical grounds and by determining the question of membership without consulting the views of the relevant 288. The Constitution and statute law forbid enforcement of invidious caste distinc-

tions. The enactment

of the various Hindu Code acts in 1955-56 virtually eliminates

varna as an effective legal concept and discredits it as a source of analogy. Although its use as a distinct legal status is still permissible for limited purposes (see nn. 70 and 71

above), the Constitution surely provides a mandate to confine it within the narrowest limits.

289, Preamble. Arts. 1-4. 15. 16, 17. 18, 23, 46,

290. Arts, 14-17, 25-30. 291

292.

ts, 25-30, 347. 350A, 350B.

Arts. 19(1)e, 25, 26, 30,

293. Id.

294. See, generally, Parts HI and IV of the Constitution.

Group Membership

351

groups. It gives, if unwillingly, legal effectiveness to the notions of rank order among groups and mutual exclusiveness among them. The per-

sistence of such an approach should be regarded as an anachronism, and one expects that before long the courts will refine and extend the empirical approach. An empirical, pragmatic approach, sensitive to the egalitarian and welfare objectives of compensatory discrimination, seems particularly appropriate in the group membership area. In handling group mem-

bership problems, the courts draw close to the central paradox of the compensatory discrimination— the employment of traditional cate-

gories of social distinction for the purpose of overcoming these distinctions. It is the task of the courts, in visualizing groups and their boundaries, to employ those parts of the traditional view which are

useful in carrying out the compensatory discrimination policy and which are compatible with the constitutional principles of equality and

voluntarism while discarding those which tend to perpetuate and ex-

tend invidious distinctions.

However, this task itself involves the courts in a dilemma. It propels the courts into subtle distinctions, refined analysis of policy, and a searching examination of facts. But the massive job of certifying crores

of recipients of preferences is going to be done, not by the courts, but by

administrators and clerks. The litigation about group membership so

far has arisen primarily from election disputes, which are atypical in several ways: the courts are directly involved and the stakes are high enough to justify a thorough investigation of the facts and to provide

sufficient incentive to expend resources to render them problematic. But

allotment

of scholarships,

housing sites, or small

loans will be

conducted remote from judicial scrutiny and in circumstances where

litigation will be rare. For the courts to guide policy, the subtleties and refinements will have to be translatable into workable rules that can be

applied routinely and inexpensively at the operating level by administrators and clerks who have neither the training and capacity nor the fact-finding resources of the judges. Inevitably, there will be some slippage between judicial formulations and actual administration. Courts cannot solve or even anticipate the problems of administration, but they can strive to design their tests so as to minimize the inevitable discrepancy.

10

Social Identity and Judicial Action WE BEGAN chapter | with the image of a compartmental

society composed of innumerable mutually exclusive groups, arranged © in interlocking clusters, layers, and strata. Even acknowledging occasional difficulty in locating groups (or individuals), the imagery conveyed a sense of determinate location on an elaborate grid. And it was implied that our knowledge—subject to inevitable blurs and gaps— depicted that complex society as it really was. Our subsequent tour of the issues of group identity leads us to reflect on the limitations of both the compartmental imagery and the ideal ofknowledge as photographic ntation of reality. Social identities are not innate qualities that inhere in or emanate from the group or individual; instead they are asserted to, imposed by, and negotiated with various other components of the society. The definition, location, standing, and boundaries of a group (or an indi-

vidual) may differ as they are viewed from different vantage points. Thus the identity of a caste may differ from the perspective of its members, fellow villagers, the regional marriage network, nearby citydwellers,

religious dignitaries,

politicians, and the educational au-

thorities. Its members may use any of these attributions of identity as counters in trying to renegotiate their identities in any of these other arenas. What the groups are, what their boundaries are, and who is in them depends on the observer, on the frame of reference, and on efforts

at self-presentation. This relativistic vision contrasts sharply with what we might call the

“realist” view of Indian society. Realism and relativism are useful to

mark extremes of ways of looking at Indian society. “Realism” proceeds as if there were bounded social units, corresponding to our intellectual categories, carrying fixed identities, and arranged in relations of mutual

inclusion and exclusion. Thus Hinduism includes varnas, which are in

turn made up of endogamous jati groups; other religions are mutually

incompatible with Hinduism,

vama, and caste; caste is exclusive of

tribe, and so forth. In the realist view all of these elements can be ordered so that each group and individual has a determinate location.

Social Identity and Judicial Action

353

Thus with sufficient information we can mark the line between untouchables and touchables, forward and backward, tribes and castes, Hindus and non-Hindus, etc.

We have moved from this naive realism to a relativist view which posits a pluralistic process of assigning and constructing identities, a process in which elements and fragments of several overlapping but competing symbol systems (e.g., religious purity, politicat dominance, “Western” advancement, etc.) are combined and recombined by indi-

viduals and groups. These elements are not fixed and inherent traits but are counters to persuade others in a continuing argument in which identities are constructed by assertion, negotiation, and imposition. Of course, stable relations of inclusion and exclusion are many and

typical, so that the realist grid is a useful way to organize a multitude of observations. It is tempting to infer that these regularities, multiplied and extended, form a single coherent system, a set of orderly blocks

interlaced with a few deviant and transient instances of conflict, ambiguity, and uncertainty. But, relativism counsels us, conflict, ambiguity,

and uncertainty are not marginal; they are part of the core of the . What we have called the realist view is a tendency to overesti-

mate the fixed, uncontroverted, and regular parts of the process. The

relativist view denies that these exist apart from the movement, ambi-

guity, ambivalence, indeterminacy, background, they are inseparable; those at rest. Realist and relativist approaches ment about the behavior that occurs cognitive and expository strategies

and conflict. Like foreground and the moving parts are as integral as do not necessarily reflect disagreein Indian society. They are different for understanding and discussing

that behavior. Few observers owe exclusive allegiance to one or the

other of these strategies. The same judge, for example, may combine

them or switch from one to another as the context commends. (Simi-

larly, in this book I found the compartmental imagery useful to present an initial sketch, but subsequently veered around toa relativist perspective to do justice to the complexities revealed in the cases.) What isa helpful map for the traveller is not necessarily the most useful picture of the landscape for the surveyor or the engineer. Group identities have a peculiar dual character. They are not natural facts, like geological formations or rainfall, impervious to fiat and uninfluenced by understanding. But neither are they artificial products of policy, like courts, legislatures, or corporations, that can be dismantled or altered by human design. Like language usages, they have an intermediate character, combining natural givenness with some malleability. The relation of government to group and individual identities is

354

Identifying the Beneficiaries

extremely complex, for government cannot promulgate such identities

nor can it simply record them. It cannot mold social reality into the categories it promulgates, because whatever it does is super-added to the existing symbolic array made available by the multiple folk and

learned systems that circulate in local, regional, and all-India networks and are internalized within the actors themselves. Government is an important source of messages about personal and group identity. These

messages may suggest new models and make available and salient symbolic constellations that were previously inaccessible. Government symbols have a special potency because they can be used to advance or defeat claims to which attach concrete rewards and determinate

results— hence the tendency for other identities to be deflected toward

alignment with official categories. Government specifications become

facts that people can use, in combination with other elements, as re-

sources in constructing or changing their identities. Instead of promulgating a definitive ordering, the government introduces new incentives for change and new possibilities for symbolic enterprise.

But it is not open to government to eschew influencing the process in

favor of observing and recording identities as if they were natural

phenomena. The experience of recording caste and caste standing on Censuses is, of course, the classic example of the impossibility of neutral measurement. What is recorded reflects collective and individual

strategies of self-presentation., The official record not only becomes an instrument of official policy, but mints new symbolic currency.

The government (including the courts), though it cannot definitively

portray

the labyrinth of group identities nor regulate it decisively,

cannot escape having policies regarding it. In a curious way government presides over the process and contributes importantly to it. How can government pursue policies that turn on group identity in this relativistic world? In many instances, I have argued that the courts should take a relativist view, recognizing and encouraging ambiguity, indeterminacy, asymmetry, and variation. I submit that courts (and by

implication, government) should incline this way as a general policy. I have argued that relativism (as embodied in what I called the empirical approach to group membership questions and in a flexible “elastic” position on the question of caste units) is more consonant with the constitutional themes of voluntarism, group autonomy, and nonrecognition of rank

ordering. The relativist approach embodies a more

comprehensive understanding of India society that enables us to include ambiguity, indeterminacy, and movementas part of the process of the continual construction and reconstruction of identities, a process that can be understood, if not exhaustively cataloged or mapped in detail. It enables us to apprehend change without concluding that the

Social Identity and Judicial Action

355

old is totally supplanted (for example, the decline of interdependence and the growth of competitiveness among castes; or the effacement among elite groups of the sense of caste hierarchy that was so palpable earlier in the century). The relativist approach alerts us to the presence

of multiple and competing

principles, whose specific and changing

manifestations can only be ascertained by empirical observation. It therefore invites the courts to link up with (and stimulate) current social

scientific scholarship. Relativism is more useful in pursuing the compensatory discrimination policies. It equips the courts with analytic tools to puncture ex-

ploitative assertions of identity (e.g., adoptions or conversions for purposes of securing benefits).1 Visualizing identity as composite and relative enables courts to readily distinguish those aspects of it relevant to compensatory discrimination policy, undistracted by the notion that

it is unitary and inseparable. Similarly, it liberates courts from the

notion that backwardness is a single trait or one that is either present or absent. It enables the courts to examine government action with a

critical eye to insure that government has focussed on aspects of identity that are relevant to compensatory discrimination policy. The relativist perspective can provide government not only with

critical response but with more options for policy. Policy alternatives are narrowed by the realist vision in which untouchability is coextensive with Hinduism, caste units are inseparable from caste standing, caste is lost upon conversion, etc. The relativist perspective

emancipates thought from these fusions, enabling us to disaggregate our observations.

We

can organize them in terms of variation and

continuity rather than in all-or-none terms of inclusion, exclusion, and

implication. Thus government is afforded the option of modulating

measures to correspond with the factual contours of specific situations. Ofcourse, what the courts have to say is only part ofa wider discourse about these policies that takes place in government, among political

elites, in the media, and among wider publics. It is a discourse in which courts are accorded a leading role. And it is one that is characterized by

1, Exploitative assertions of identity (for example, a Brahmin having himself adopted by a Scheduled Caste family in order to gain admission to medical college) often entail a

dramatic inversion of symbols. It might be argued that such attempts should be welcomed as disestablishing traditional notions of hierarchy. But as we noted earlier, one of the paradoxes of compensatory discrimination policy is that it confers benefits on those among the beneficiary groups who are personally least needful. Much of the benefit conferred on those at the bottom is indirect—i.c., through provision of models,

establishment of networks, provision of representation and opportunities for vicarious participation. In whose success and advancement do the worse-off participate in these ways? Group boundaries, as summations of these patterns of participation, are then an important tool of compensatory discrmination policy.

356 Identifying the Beneficiaries considerable cognitive disarray. Consider, for example, the nomenclature used to describe the policies (reservations, special treatment, concessions, privileges) and their beneficiaries (backward classes, Other Backward Classes, Scheduled Castes, harijans, weaker sections, etc.). It

is a landscape of euphemism and equivocation in which terms have multiple, ambiguous, overlapping, and shifting meanings. Precise discussion of policy is rendered difficult: even officials stumble and public discussion is confounded. For example, there is a persistent tendency to fuse all policies and their beneficiaries under such general headings as reservations for backward

classes, so the resentment and dismay engendered by use of

the Other Backward Classes category to stake out massive claims on behalf of peasant-middle groups in some southern states are transferred

to the more modest measures for Scheduled Castes. Atrocity stories about extensive reservations for undeserving beneficiaries, generated

by OBC provisions at the state level, are readily generalized to the discredit of the Scheduled Castes nationally. Another persistent feature of this discourse is that it proceeds without much attention to the actual performance of the programs. Ignorance

about performance is not accidental but is an inherent part of the administration of the policies. The policies are not, for the most part, programmatic—i.e., they do not involve deliberate pursuit of specified goals. They are, as Aggarwal and Ashraf observe, “‘activity centred and not goal centred.” Officials can point with pride to a list of accomplishments. But the relation between program performance and the goals to which the program is ostensibly aimed are left unstated and unexamined. Even if goals were made explicit, the bookkeeping is generally such that it is impossible to separate out the net effects of the program.3 Actually, large numbers of reservations for those highly cherished positions are not filled. Yet overestimation of the effect of reservation is pervasive. There is widespread feeling that enough (or too much) has been done for these people. It is time for them to stand on their own feet. This perception that so much is being done is fostered by the immense variety of programs (in housing, land, legal aid, water supplies, etc.). Many of the programs are very small, and it is onerous and cumber2. Aggarwal and Ashraf 1976: 166.

3. For example, an increase in the number of Class I government servants from the

Scheduled Castes does not tell us the net effect of the reservation. This enables the

government to take credit for all the jobs obtained by the Scheduled Castes, not only those conferred by the reservation. (Credit may be claimed several times over. For some

of these jobs may have been gained through a scholarship program which enabled some Scheduled Castes to gain places on merit.) And it contributes to the overestimation of the occurrence and effects of preferential programs and to their unpopularity.

Social Identity and Judicial Action

357

some to obtain any benefits. Much of their resources are consumed in administration with only a tiny scatter of benefits (e.g., legal aid). Worse, rather than supplementing benefits available through regular government programs, these special programs often offer an inferior substitute. And their existence becomes justification for exclusion from general programs of benefits.4 It is beyond the courts to rescue these policies from systematic itive distortion, for courts cannot control the way that various actors and audiences perceive judicial (and other) pronouncements through the lenses of their own preconceptions and use them in various arenas. Nevertheless, the courts are extremely significant contributors to discourse on these matters. The focussing of attention by conflicting claims dramatizes their pronouncements. Their decisions are given authority by producing tangible outcomes that can (in theory at least)

be reproduced

in similar cases. Although

their interventions are

episodic, the courts are the custodians of a continuing didactic tradition: they have the opportunity to reappraise and rework what they have said before, elaborating, adding emphasis, and pointing out how

they should be read.

The linkage of courts to the beliefs and the rhetorical stock of both elites and wider publics, although subject to noise and distortion, supplies another reason for forthright relativism. By permitting exceptions that blur boundaries, by accrediting variations that dislocate hierarchical markers, and by refusing to promulgate a single authoritative picture of the group contours of Indian society, the courts help to

disestablish the traditional picture of that society, a picture which

provided legitimacy for traditional relations of domination and hierarchic ordering. Of course, ignoring the symbols of domination and hierarchy will not, as some have naively contended, abolish the domination and hierarchy themselves. Indeed, avoidance may conceal the

continuation of old disparities. Courts face the problem of keeping these

categories available for carrying out policies of eliminating old ine-

qualities without invigorating them as sources of symbolic sustenance for hierarchic patterns.

But ifa relativistic approach to q testions of identity generally seems favorable to equalizing policies, it is not sufficient nor is it free of perplexities that make problematic the-role of courts in carrying out

these policies. Courts are passive and reactive institutions. They do not

actively mobilize an agenda of cases. Yet in an indirect and subtle way they do foment and instigate litigation.’ When they indicate the scope 4. See Aggarwal and Ashraf 1976:171.

5. Cf. Mayhew 1975.

358 Identifying the Beneficiaries ofa right or locate standing to vindicate that right, they help to mobilize claimants. Their message is like a seed that requires fertile soil and ample water to nourish it. Some potential litigants are better able to perceive the message and supply the matching resources needed to make it flourish. The results brought about by such judicial messages, then, depend not on the courts themselves but on the distribution of

capabilities among litigants.€

Judicial willingness to go behind conventional categories to assess their empirical “fit” and their policy relevance invites litigants to make arguments and bring forward evidence along these lines. But empirical arguments about complex states of fact—e.g., the contours of group membership or of backwardness—are more difficult and expensive to make than logical arguments about relations of inclusion and exclusion of categories. Rules that entail complex showings of fact and are open to subtle argument about policy accentuate the role of litigant capabilities and amplify the effect of disparities in those capabilities. Compensatory discrimination policies are designed to redistribute resources and opportunities to those who enjoy the fewest advantages. It seems plausible that the intended beneficiaries of these policies enjoy not only less wealth, education, and status but also less capability as users of legal remedies. This supposition seems confirmed by the pattern of litigation in this area.? The beneficiaries of compensatory discrimination are less successful than their adversaries and have been

unable to use litigation affirmatively to improve the implementation of these policies. The interests does not of use legal remedies. tiary demands of the

development of legal doctrine favorable to their itself enlarge the capacities of these beneficiaries to (Indeed, increases in the complexity and evidenlaw may amplify disparities in litigant capacities.)

To deliver fully the benefits of these policies requires an enhancement of 6. The notion of disputant capability includes such attributes as ability to perceive

grievance, use information about the availability of remedies, articulate and manage

claims competently, seck and utilize appropriate help, and bear the psychic and sucial costs of disputing. These may be attributes of individuals as well as of larger social units. Beyond this sort of personal competence are a related set of structural factors: the scale, continuity, and organization of the disputant that permit it to-store information, utilize advance intelligence, develop expertise, establish and maintain credibility as a disputant, adopt long-term strategies, coordinate activities in different forums, etc. This party capability notion is elaborated in Galanter 1974, 1975, 1976.

7. This pattern is analyzed in chap. 14, §C, below.

8. In many cases the putative interests of the beneficiaries are represented by counsel for the government, defending a particular scheme. The statement in the text holds for such cases. They also raise the question of how much the positions asserted by the

government in such cases diverge from those that would be asserted by advocates responsive only to the beneficiaries.

Social Identity and Judicial Action

359

the capability of the beneficiaries to elicit tavorable judicial doctrine and

to utilize it. Such an enlargement of their capacities would require the provision of more continuous and specialized legal services; and it would require better organization of the beneficiaries to coordinate strategy, monitor administration, aggregate claims, support the development of legal expertise, and do other things that would be beyond the capabilities of isolated individual litigants. The implementation of compensatory discrimination policies is en- « trusted to administrators and clerks, not to judges; judicial intervention

is intermittent. The complexities and refinements of judicial doc-

trine have to be translated into: workable rules that can be applied routinely and inexpensively. The individualized weighing of multiple factors that is attractive in the judicial setting is worrisome in the administrative setting. It may lead to errors due to misunderstanding and lack of information; it may confer discretion where that is not wanted; and it may make administration slow and costly. Further

judicial resolution is possible in theory, but in practice informed and

continuing monitoring by the judiciary is possible only if the beneficiaries are capable of bringing to the courts the cases and the data required for such control. The competence of courts to oversee the implementation of compensatory discrimination policy and to resolve the tensions between that policy and other commitments is both created and limited by the kind of institution that courts are.9 Within those

limits there is great scope for achievement, but it is achievement that is

very dependent on having capable parties that can elicit and utilize creative judicial action. 9. On the variation in the roles and structures of courts as institutions, see Galanter (forthcoming); Friedman 1967; Damaska 1975; Shapiro 1975.

°

Part Three COMPENSATORY DISCRIMINATION AND THE JUDICIAL PROCESS

In Part Two the courts and the compensatory discrimination policies formed the frame within which questions of group identity were addressed. In Part Three the courts and compensatory discrimination policy move into the foreground. In roughly a quarter of a century the higher judiciary in India (Supreme Court and High Courts) erected a jurisprudence of compensatory discrimination. This branch of Indian constitu-

tional law was built from scratch; it is very much a domestic product,

produced with little guidance or borrowing from abroad. Although there have been some false starts and curious turnings, the major lines of this body of law can be traced. Chapter 11 discusses the constitutional setting of compensatory discrimination policy and doctrine. It traces the divergent but overlapping

readings of the constitutional design that informed judicial response to specific problems until a remarkable 1975 decision (N. M. Thomas v.

State of Kerala) reopened what appeared to be settled, creating a new set

of doctrinal possibilities and raising the question of how far they would

ramify and how they would relate to earlier doctrine. Chapters 12 and 13 assemble two clusters of problems that arise in connection with compensatory discrimination programs. These chapters examine the judicial response to these problems and relate that response to the design and administration of these preferential

362

The Judicial Process

programs. In these judicial encounters we get some glimmering of the problems of basic institutional architecture inherent in compensatory policies. And we see the problems and possibilities of having courts

monitor these policies.

Chapter 12 takes up a series of problems in which the judiciary is asked to respond to the question of “how much”: how many places can be reserved, for how large a class of beneficiaries, how much variation

from general standards can be tolerated, etc. Chapter

13 takes up a

cluster of problems that center around the

design of redistributive policies: how to distribute benefits so that all’ benefits are not monopolized by the better-off; how to build in flexibility while curtailing discretion that could defeat basic policy; how to articu-

late preferences for several groups of beneficiaries.

The responses of courts to specific claims over a quarter of a century

have erected a body of law which begins to wrestle with these questions.

Neither a stagnant backwater nor between courts and government, it observe Indian lawyers and judges from judicial grappling with specific

the center of intense controversy is an area in which we can readily at work. Chapters 14 and 15 turn problems to the judicial process as

such. Chapter 14 portrays the higher judiciary, the lawyers, and the litig-

ants who have shaped the law of compensatory discrimination. It tries to understand the ways in which the doctrinal product is shaped by the structure of courts, the recruitment and work habits of judges, the organization of legal services, the goals, resources, and capabilities of litigants and the legal culture that all of them share. Chapter 15 turns this around to ask what difference the doctrinal product makes, what impact it has on compensatory discrimination

policies. Even in an area where government is attentive to the courts and there is no pattern of resistance, the actual effects of decisions are always problematic. How can courts optimize the effects of their deci-

sions? What are the options that they have in the light of their changing sense of judicial work and of the re-emergence in the late 1970s’of Scheduled Castes and Other Backward Classes as central policy issues?

1]

The Constitutional Scope of the Compensatory Discrimination Policy In THE British period, arrangements for the distribution of power and benefits according to membership in communal groups were accepted methods for the adjustment of the political balance among communities as well as for the amelioration and protection of the lowly.? The Constitution of independent India drastically restricts the use of such arrangements. They are discarded as a general principle of governmental operation and may be used only for the purpose of promoting equality. Communal preferences and quotas for the purpose

of political accommodation are outlawed.? The protective (or com-

pensatory) discrimination authorized by the Constitution is envisaged as an exceptional and temporary measure to be used only for the purpose of mitigating inequalities; it is designed to disappear along with these inequalities. The Constitution makers did not intend it as a device to consolidate and protect the separate integrity of communal groups.> But while the preferential principle is thus confined in the purposes for

which it may be used, it has been expanded in scope to cover a wide

array of governmental schemes and programs.

1. Quotas and preferences in recruitment for government service were not uncommon. Similar arrangements occurred in education, housing, revenue, and agricultural administration. The Government of India Act, 1919, provided for communal representation for Mohammedans, Sikhs, Anglo-Indians, Indian Christians, Depressed Classes, Aborigines, and other groups. Similar communal representation was provided by the

Government of India Act, 1935.

2. Proposals for reserved seats in legislatures, reserved cabinet posts, and reserved posts in public service for minorities (Muslims, Christians, Sikhs, Anglo-Indians, and Parsis, among others) were eliminated from the final draft of the Constitution. See XII

CAD 229. Cf. the rejection of the proposed elimination of the qualifying word “backward” from Art. 16, which would have allowed provision of communal reservations in

public services (VII CAD 682, 704). For a discussion of the Constituent Assembly's handling of communal representation, see Retzlaff 1963; Austin 1966: 12 ff, 146 ff. 3. There are constitutional provisions for the protection of the identity and integrity of groups: religious denominations have a Fundamental Right to manage their own religious affairs and to maintain and administer institutions (Art. 26); groups with a

364

The Judicial Process

A. PREFERENTIAL TREATMENT IN THE CONTEXT OF FUNDAMENTAL RIGHTS The effort to secure equality by means of preferential treatment must

be seen in its constitutional context. In the Constitution, the com-

pensatory theme appears juxtaposed with the theme of formal equality.

The provisions for compensatory preference appear as exceptions within a framework of enforceable fundamental rights which attempt to curtail the significance of ascriptive groups and to guarantee equal

treatment to the individual.

The Constitution of India confers on all citizens a fundamental right to be free of discrimination by the State on grounds of race,‘ religion, and caste. Governmental discrimination on these grounds is prohibited generally by Article 15(1). More specifically, discrimination is prohibited by Article 16(2) in regard to State employment, by Article 23(2) in regard to compulsory public service, by Article 29(2) in regard to

State-run in regard forbidden descent,’

and State-aided educational institutions, and by Article 325 to electoral rolls. In specific contexts government is further to discriminate on grounds of place of birth,’ residence,® class,® language, and sex.?°

It is envisaged that government will not only refrain from discriminatory Practices in the public sphere, but will actively seek to eliminate them in the private sphere. Untouchability is forbidden, and

citizens are protected from certain discriminatory practices by private

persons and institutions.1! Discrimination by private individuals in regard to use of facilities and accommodations open to the public is

prohibited.12 Discrimination forbidden.

in private educational institutions is

distinct language, script, or culture are guaranteed “the right to conserve the same”, (Art. 29); religious and linguistic minorities have the right to establish and maintain educational institutions (Art. 30 (1)). No similar guarantees are extended to caste groups. (However, at least some caste groups qualify as “‘religious denominations.” See

Galanter 1968.) However, the governmental commitment to the Scheduled Tribes does

include preservation of their distinctiveness. See chap. 5, §C, above. 4. “Race” does not play a prominent part in post-Independence Indian jurisprudence. Its presence signifies an end to earlier colonial discrimination on racial grounds. (See n. 103, chap. 7, above). But it may have some residual conceptual role. Thus one court conceived of preferences for tribal people as preference “on racial ground”

(Mahendra Nath v. State, A. 1. R. 1970 A. & N. 32, at 34). And cf. the special provisions for Indians, discussed at n. 50, chap. 9, above.

5. Arts. 15(2) and 16(2).

6. Art. 16(2). But cf. Art. 16(3). 7, Art. 16(2). 8. Art. 23(2). 9. Art. 29(2). Cf. Art. 30(2) and see Arts. 29(1), 30(1), and 350, 350-A, 350-B, which grant rights regarding language. 10. Arts. 15(2), 16(2), and 325. 11. Art. 17. For a discussion of the working of these provisions, see Galanter 1972. 12. Art. 15(2). 13. Art. 28(3), Art. 29(2).

Constitutional Scope

365

But this attack on discrimination, public and private, is only one facet

of the constitutional scheme to secure equality. The Constitution also directs the government to undertake special measures for the advancement of backward groups. In addition to provisions for reservation in government employment and legislative representation, the Constitu-

tion declares as a “Directive Principle of State Policy:”

The State shall promote with special care the educational and economic interest of the weaker sections of the people, and,

in particular, of the

Scheduled Castes and the Scheduled Tribes, and shall protect them from social injustice and all forms of exploitation.**

The Directive Principles of State Policy are not themselves justiciable,

but the Constitution prescribes it as “‘the duty of the State to apply these

principles in making laws.” 15

The scope of governmental power to make special provision for the backward was soon put to the test. When the Constitution came into

force, the State of Madras apportioned seats in its medical and engineering colleges in accordance with its Communal G.O.,' which provided that for every fourteen seats the selection committee was to admit the following:

Non-Brahmin Hindus ... 6 Backward Hindus ...... 2 Brahmins .2 Harijans 2 Anglo-Indians and Indian Christians ..... Muslims ..............

| 1

A Brahmin student who would have been admitted on purely academic qualifications charged that this was in violation of Article 29(2), which

provided:

No citizen shall be denied admission into any educational institution maintained by the State or receiving aid out of State funds on grounds only of religion, race, caste, language or any of them.

In State of Madras v. Champakam Dorairajan,‘7 the Supreme Court upheld the claim of the protesting student. Finding that the denial to the student of a chance to compete for the other seats was only on the

basis of her caste and violated Article 29(2), the Court advanced two

crucially important propositions. First, it held that “the right to get 14. Art. 46. 15. Art. 37. 16. On the background of the Madras Communal G. O., see Irschick 1969. 17. A. 1. R. 1951S. C. 226, 1951 S.C. J. 313.

366

The Judicial Process

admission

into any educational institution of the kind [covered by

Article 29(2)] is a right which an individual citizen has a citizen and

not as a member of any community or class of citizens.”18 Thus the Court refused to countenance any notion of group rights. Second, the

Court rejected the argument that the Directive Principle of Article 46 by its own

force established a principle of preference which might

lawfully be embodied in legislation. “The Directive Principles of State

Policy have to conform to and run as subsidiary to the Chapter of Fundamental Rights. .. .”!9 The Court had an additional textual argument for this position: noting that Article 16 was explicitly qualified by an exception allowing reservations for the backward in the area of State employment, it suggests that “the omission of such an express provision from Article 29 cannot but be regarded as significant.” The government’s argument about the implications of Article 46 would

render superfluous the exception in Article 16.

It should be pointed out that this was by no means an inevitable or unavoidable reading.?1 An earlier case in the Punjab High Court

upheld reserved seats for Harijans in educational institutions on the

ground that “Article 46 must be taken as an exception to Article 29(2) ” and thus the state could constitutionally promote the educational interests of Scheduled Castes by adopting a system of reservations.22 In later years, the Supreme Court, still maintaining that Directive Principles were subordinate to Fundamental Rights, allowed that “these

directive principles are to be’ taken into account in determining the scope and application of Fundamental Rights.”23 From the vantage point of 1967, the then Chief Justice of the Supreme Court indicated that “‘any reasonable attempt to raise the status of the Backward

Classes could have been upheld on the principle of classification.” One 18. A. I. R. 1951S. C. at 227.

19. A. I. R. 1951 S.C. at 228.

20. A. 1. R. 1951 S.C. at 228.

21, Alexandrowicz 1957: 62 observes that in Champakam the Supreme Court “did not sacrifice the letter of the Constitution to make any spectacularly progressive move in the

spirit of the Constitution.” Cf. the remark of Prime Minister Nehru in the subsequent parliamentary debate that the proposed amendment “might almost be considered redundant” (Parliamentary Debates, Vol. XII-XIIT [Part 11], col. 8820 [16 May

1951). . 22. Om Prakash v. State of Punjab, A. 1. R. 1951 Punj. 93, 97.

23. Cf. with Mhd. Hanif Qureshi v. State of Bihar, A. 1. R. 1958 S. C. 731 at 732, 737, where the “subsidiary” interpretation of Champakam Dorairajan is commended and the Court observes that | although the Directive Principles are fundamental, the state must implement them “in such a way that its laws do not take away or abridge the

Fundamental Rights, for otherwise the protective provision of Chapter III will be ‘a mere

rope of sand.’ ” 24. “To remove the effect of centuries of discriminatory treatment and to raise the downtrodden to an equal status cannot be regarded as discriminatory against any

Constitutional Scope

367

wonders how much the outcome in Doratrajan was affected by the

contingency that the question was presented in the context of a full-

blown scheme of communal quotas rather than as a measure specifi-

cally designed to advance backward classes.

On the same day in Venkataramana v. State of Madras, the Court struck

down part of a similar communal quota in regard to government posts.?5 Article 16(4), it found, permitted reservations only for “‘back-

ward classes.” The Court concluded that of the groups on the Madras

list, only Harijans and “backward Hindus” could be said to be backward classes. The reservations for these groups were allowed, the others struck down. One may imagine a result-in Dorairajan, along the lines of the result in Venkataramana. The Court might have said that Article 46 directs the State to promote the educational and economic interests of the weaker sections—and mentions the Scheduled Castes and Tribes in particular; these groups then are clearly ‘‘weaker sections”; therefore the classification is not on the basis of caste “only”; therefore reserva-

tion for these groups is permissible in spite of Articles 15 and 29. As to

other groups, the State had the burden of showing that they are “weaker sections” and within the exception to Article 15 and 29 implied by the Directive Principle; since there was no such showing, the reservations for these should be struck down. In the event, the Court held that the government had no power to reserve seats for backward

communities

(or, a fortiori, for any com-

munities) in educational institutions. The implication was that all special provisions for backward caste groups (outside the area of government employment) would be open to the same kind of attack. The inference was later drawn by the Bombay High Court, which held that

it was not a legitimate public purpose for the state to provide a housing

colony for Harijans, but unconstitutional caste discrimination forbidden by Article 15.26 Governmental reaction to this doctrine was swift. Along with other Supreme Court decisions that had discomfited the government, it was reversed by constitutional amendment less than two months after it had one. . . . With all due respect [to Champakam Dorairajan] the question of discrimination

hardly arose because in view of [Articles 46, 340, 16(4)]. . . any reasonable attempt to

raise the status of the backward classes could have been upheld on the principle of classification” (Golak Nath v. State of Punjab, A..1. R. 1967 S. C. 1643 at 1706). CE. G. S. Sharma 1965: 183 ff., tracing the greater reliance on the Directive Principles in later

years, who explains Champakam by postulating the necessity of further time for the Court

to realize that “under the special conditions of the Indian society a special treatment to a particular minority for a specific period could be treated as a value comparable to the value of maintaining secular traditions of admissions for educational institutions.”

25. Venkataramana v. State of Madras, A. 1. R. 1951 S. C. 229.

26. Jagwant Kaur v. State of Bombay, A. 1. R. 1952 Bom. 461.

368

The Judicial Process

been delivered. provides:

Among

the amendments

was Article

15(4), which

Nothing in Article 15 or Article 29(2) ... shall prevent the State from

making any special provision for the advancement of any socially and educationally backward classes of citizens or for the Scheduled Castes and

the Scheduled Tribes.27

The Constituent Assembly had rejected the inclusion in the original

Constitution of a provision similar to Article 15(4).28 Now the same

assembly, in its incarnation as the provisional Parliament, passed it

with little opposition,2® on the ground that it was needed to empower

the State to carry out the Directive Principles by insuring that the

Fundamental Rights guarantees of equality did not obstruct substan-

tive equalization. Together with the provisions for special treatment that were in the original text, Article

15(4) forms the constitutional

basis for the subsequent governmental policy of compensatory discrimi-

nation. Before proceeding to an examination of this policy, we shall briefly review the scope and relations of these constitutional provisions. B. THE SCOPE OF THE CONSTITUTIONAL PROVISIONS FOR SPECIAL TREATMENT

The only specific authorizations for preferential treatment in the original Constitution were in the fields of government employment and legislative representation. Discrimination in government employment is outlawed by Article 16, but Article 16(4) permits the State to make

any provision for the reservation of appointments or posts in favor of any backward

class of citizens which, in the opinion of the State, is not ade-

quately represented in the services under the State.

In addition, Article 335 provides: The claims of the members of the Scheduled Castes and the Scheduled 27. Constitution (First Amendment) Act, para. 2 (1951).

28. Professor K. T, Shah’s proposal to amend Art. 15 to allow the state to make

special provision for the “advantage, safeguard, or betterment” of the Scheduled Castes and Scheduled Tribes was rejected by the Assembly (VII CAD 655, 664). Dr. Ambedkar, chairman of the Drafting Committee, opposed the amendment as providing an

opening for segregated facilities (VII CAD 661). It should be noted that the Shah

amendment is narrower in scope than Art. 15(4), which applies to Backward Classes as

well as Scheduled Castes and Scheduled Tribes. 29. This clause was adopted by a majority of 243 to 5, a slightly higher margin than

either of the other major amendments (Parliamentary Debates, Vol. XII-XI1I (part I),

col. 9833). Cf. cols. 9885, 9927 (1 June 1951).

Tribes shall be taken into consideration, consistently with the maintenance of efficiency of administration, in the making of appointments to services

and posts in connection with the affairs of the Union or ofa State. The

relation between

these provisions remains somewhat obscure.

Article 335 is confined to Scheduled Castes and Scheduled Tribes,

while Article 16(4) extends to all “backward classes.” Article 335 seems to include any method of preference, while Article 16(4) is confined to

reservations. Article 335 extends to all appointments “in connection

with the affairs” of the State, which may be broader than the “services under the State” referred to by Article 16(4). Finally, Article 320(4)

provides that public service commissions need not be “consulted as

respects the manner in which any provision referred to in clause (4) of Article 16 may be made or as respects the manner in which effect may

be given to the provisions of Article 335.39

Both Article 15(4) and Article 16(4) permit special provisions for backward classes. Article 15(4) applies to the State in all of its dealings, while 16(4) is confined specifically to the field of government employ-

ment. It has been held that more general Fundamental Rights provi-

sions are displaced in relation to a specific area by provisions specifically pertaining to that area.31 Thus, the area of employment, offices, and appointments under the State is controlled by Article 16 alone, and preferences in this area must be within the scope of 16(4). This includes

judicial office3? as well as administrative posts, but it does not include elective office.33 The authorization

to establish preferences in employment is not

confined to posts directly under the State. Where the State acts as an

employment agency tor State-aided schools, it may make such preference a condition of the aid.>4 The situation in regard to public sector

30. An attempt to include the Other Backward Classes within Art. 335 was proposed by Dr. Deshmukh (see n. 44, chap. 6, above) (IX CAD 598). 31. Dattatraya v. State of Bombay, A. 1. R. 1953 Bom. 311 (Art. 16 displaces Art. 15 in the area of government employment); University of Madras v. Shantha Bai, A. 1. R. 1954

Madras 67 (Art. 15 displaced from area controlled by Art. 29[2]). See also State of Madras

v. Champakam Dorairajan, A. 1. R. 1951 S. C. 226. On the relations between the more general and the increasingly specific guarantees of equality in Arts. 14, 15, and 16, see Rangachari v. General Manager, A. 1. R. 1962S. C. 36; Shyam Behari Tewari v. Union of India, A. I. R. 1963 Ass. 94.

32. Venkataramana v. State of Madras, A. 1. R. 1951 S. C. 229; Kesava Iyengar v. State of

Mysore, A. 1. R. 1956 Mys. 20. 33, Dattatraya v. State of Bombay, A. 1. R. 1953 Bom. 311.

34. The Supreme Court held that Art. 16(4) allowed the state to provide that its Public Service Commission give preference to Scheduled Caste teachers in filling vacancies in state-aided schools run by religious minorities. This provision of the Kerala Education Bill was particularly objected to on the ground that Scheduled Caste

370

The Judicial Process

enterprises is far from clear. Government lawyers advised that the constitutional provisions for reservations were not applicable in the

sense that public sector enterprises were not constitutionally obliged to have a policy of reservations.35 Yet it was open to them to make reservations, and it was open to the government to direct them to do

$0.36

Article 16(4) covers not only preferences in initial recruitment into

government service, but also preferences in promotions within the service. After some hesitation and with considerable reluctance, the

Supreme Court held, in Rangachari v. General Manager (1961), that the

“posts” referred to in Article 16(4) included promotions as well as initial appointments, but indicated that the preference permissible under Article 16(4) would not necessarily extend to other aspects of employment covered by Articles 16(1) and (2)—e.g., salary, increment, pension, retirement age.37 Such matters are “absolutely proteachers might know nothing of the religion of the minority by whom the school was run. Denying that it violated Art. 29(2), the Supreme Court upheld the provision as a permissible condition for receiving grants in aid from the state (In re Kerala Education

Bill, 1957, A. 1. R. 1958S. C. 956, 983). (This provision of the bill was repealed in 1960.)

However, this conclusion is rendered doubtful by Sidhrajbhai Sabbaj v. State of Gujarat, A. 1. R. 1963 S. C. 540, 547, where the Supreme Court referred to the rights of minority institutions as “absolute” and subject only to those reasonable restrictions that are “in

the interests of the institution.” 35. Thus the Law Ministry is reported to have advised that Arts. 15, 16, and 335 were “not applicable to the public sector enterprises” (CWSCST [4th] 2: 9 [1969]). This is

said to be based on the decision of the Supreme Court in Rajasthan State Electricity Board v.

Mohan Lal, A. 1. R. 1967 S. C. 1857. But in that case it was held that the Electricity

Board was clearly an authority to which the provisions of Part III are applicable. The

Court there reached this conclusion by reasoning that duties imposed upon the “State”

by Art. 46 imply that the “State” as defined in Art. 12 comprehends “bodies created for

the purpose of promoting the educational and economic interests of the people” (id., 1862). And surely public sector undertakings are not free to discriminate among employees on caste or religious lines (forbidden the State by Arts. 15(1) and

Shah, J., concurring, pointed question is invested with the the Fundamental Rights that of whether it might act on its

16(2)).

out that there is.a further question of whether the body in sovereign power of the state to impose the restrictions on the state is authorized to do. Thus, there remains a question own to implement 16(4).

36. CWSCST (4th) 2:9 (1969).

37. General Manager v. Rangachani, A. 1. R. 1962 S. C. 36. The question was left open in

the earlier case of All-India Station Masters v. General Manager, A. 1. R. 1960S. C. 384. The High Court in the Rangachari case had concluded that Art. 16(4) is more limited in scope

than Art. 16(2). The latter forbids caste discrimination in respect to “any employment

or office under the state.” Art. 16(4) is an exception to this but only as regards “appointments or posts.” The High Court held that promotion was “employment . . . under the State” but was not “appointment to a service or a post.” Thus, caste discrimination in promotions would be barred by 16(2) and would not be saved by 16(4). The net result would be to permit reservations only at the stage of initial appointment to a service. Presumably, reservation would also be permissible if the

higher post constituted a separate service (A. I. R. 1961 Mad. 35 at 41).

Constitutional Scope

371

tected by the doctrine of equality of opportunity and . . . do not form part of the subject matter of Article 16(4).”38 While the dissenting judges argued that reservation was limited to securing adequacy of quantitative representation of the favored group,3% the majority held that reservation could legitimately be used to secure representation in posts of higher grades.“ Thus, in establishing preference in government service, the State can aim at placing members of privileged groups in positions of authority, responsibility, and prestige. In the 1963 case of

Balaji v. State of Mysore,*1 the Supreme Court took the unusual course of explicitly adding a gloss to its earlier decision, pointing out that the

Rangachari case “posed the bare question about the construction of Article 16(4). The propriety, the reasonableness or the wisdom of the

impugned order was not questioned.”4? In Rangachari, the Court had

pointed to the language of Article 335, requiring the State to take into account the efficiency of the services in making provisions for Scheduled Castes and Tribes, and declared the necessity of striking a reasonable

balance between the claims of these classes and the efficiency of the

services. In Balaji the Court emphasized that the public interest in the efficiency of government services set limits to reservation in promotions, putting outside the scope of Article 16(4) any “unreasonable, excessive, or extravagant reservation,” for that would, by eliminating general competition in a large field and by creating widespread dissatisfaction among the employees, materially affect efficiency.43 Thus the Court gave notice that it would carefully scrutinize the reasonableness of re-

servations in the area of promotions. This came to pass in the Devadasan

case, where the Court invalidated the reservation because of its un-

reasonable extent.** Both Rangachari and Devadasan involved promotion

preferences for Scheduled Castes and Tribes. The authorization for preferences in promotions presumably includes the Other Backward Classes (OBC) as well.45 Since the courts have shown themselves somewhat more stringent in dealing with reservations for OBC, it might well be that such preference in promotions as is allowable for these most

disadvantaged groups would be excessive when done for OBC. 38. A. I. R. 1962S. C. 36 at 42.

39. Wanchoo, J., id. at 48;Rajagopala Ayyangar, J., id. at 49, The former relied on the interpretation of “posts” referred toinn. 54, chap. 12, below. The latter agreed with the High Court, see n. 37, above.

40. The majority couples this with the puzzling observation that it is impermissible to make any reservation at the cost of efficiency of administration (A. I. R. 1962 S.C. at

44). This would seem to limit reservations to posts for which there are a number of

equally qualified candidates. Or perhaps it implies that there is no relation between ordinary selection procedures and efficiency.

41. A. 1. R. 1963 S. C. 649. 42. Id. at 664. 43. Id. 44. Devadasanv. Union of India, A. 1. R. 1964S. C. 179, discussed in chap. 12,§B, below. s. é Triloki Nath Tiku v. State of Jammu and Kashmir, A.1.R. 1967 S.C. 1283, A. I. R. 1969 C1.

372

The Judicial Process

The wording “‘any provision” in Article 16(4) and “any special pro-

vision” in Article 15(4) gives the State great leeway in prescribing the

method of operation of schemes for preference. Governmental agencies have utilized a wide variety of devices for the purpose of conferring advantage on backward groups. Reservations have involved not merely setting aside reserved places, but also such preferential rules of recruitment as waiver of age requirements, of application fees, and of minimum educational qualifications; establishment of a lower minimum of qualifying marks on competitive examinations; special assistance and training in preparation for competitive examinations.*¢ Are such practices covered by the authorization of Article 16(4) that the State can make ‘“‘any provision for the reservation” of government posts? It has been suggested that fixing a lower minimum level of successful marks in a competitive examination is outside the power of

the State, since it is not a “reservation in any sense of the term under Article 16(4).”47 However, such devices would appear to be included

within the constitutional authorization.48 While Article 16(4) confines the State to the method of “reservations,” it is clear that Article 16(4) empowers the State to determine the precise method to be adopted in

effectuating the reservation.49 Unless it included permission for special 46. RCSCST 1961-62: I, 128 ff.

47. Somnath Iyer, J., in Chandra Sekhera v. State of Mysore, A. 1. R. 1963 Mys. 292 at

299-300 (dictum). In Kulkarni v. State of Mysore, A. 1. R. 1963 Mys. 303, petitionet challenged the selection of munsiffs in which maximum age requirements had been

raised and the requirement of minimum experience at the bar lowered for Scheduled Castes and Scheduled Tribes. The petition was dismissed without considering on the merits constitutional objections to these practices. In Devadasan v. Union of India, A. 1.R. 1964S. C, 179, the Court never reached petitioner’s argument of the unconstitutionality of the double standard of minimum qualifying marks. 48. Devadasan v. Union of India, A. 1. R. 1964 S.C. 179 at 187. The position under Art. 16(4) would then be analogous to that under Art. 15(4), where it is “open to the State to . . . fix different minimum marks” so long as the difference is reasonable for purposes of Art. 14. In Sukhdev v. State of Andhra Pradesh, 1966 (1) Andh. W. R. 294 at 312, there wasa

waiver of minimum qualifying marks for medical college admission from 45% to 40%.

Cf. General Manager v. Rangachari, A. 1. R. 1962 S. C. 36 at 42, where the Court imphes

lower standards are permissible. See chap. 12, §C, below.

49. This interpretation of Art. 16(4) is reinforced by the presence of Art. 335, which seems to contemplate preferences in recruitment which lie beyond “reservation” in the widest sense. See General Managerv. Rangachari, A.1.R. 1962 S.C. 36 on the relevance of Art. 335 to the interpretation of Art. 16(4). Art. 335, though, applies only to the Scheduled Castes and Tribes. If “reservation” in Art. 16(4) is narrowly construed, government would still be able to rely on Art. 335 to authorize special recruitment procedures for the groups covered there, but it would be curtailed in using such procedures

for Other

Backward

Classes. Cf. Art. 320(4), which contemplates wide

governmental discretion in respect to the “manner” in which Art. 16(4) and Art. 335 are given effect.

Constitutional Scope

373

rules of recruitment, the provision for reservation would be meaningless, for it would merely mean that those who otherwise qualified would

be entitled to reserved places—but they are entitled to that anyway under Article 16(1).5°

The power conferred in Article 15(4) does not seem confined to mea-

sures which confer benefits directly on members of these groups, but

would seem to encompass indirect provision for their ‘“‘advancement”’

which might otherwise offend Article 15(1). Suppose the State were to pay lawyers, doctors, or agricultural specialists to provide services to 50. The striking down of the “‘carry-forward rule” in the Devadasan case— especially the view that there must be some sort of equal opportunity for others to compete on every occasion of recruitment (A. I. R. 1964 S. C. at 187)—might serve as a basis for attacks on these preferential rules of recruitment. However, it is clear that the vice there

was the extent of the reservation rather than use of devices to broaden Scheduled Castes and Tribes’ eligibility to compete. See discussion at chap. 12, §B, below. 51. Thus,

Art.

15 would

also be displaced from the area covered by Arts. 29(2)

(state-aided education) and 23(2) (compulsory public service).

52. Mahendra Nath Pathak v. State of Assam, A. 1. R. 1970 A. & N. 32. 53. See n. 68, below.

54. Moosa v. State of Kerala, A.A.R. 1960 Ker. 355 (housing for Harijans). Cf. Raju

v. Chief Electoral Officer, A. 1. R. 1976 Guj. 66, 71, where the court rejected a challenge to

the Representation of the People Act’s provision for lower electoral deposits from Scheduled Caste and Scheduled Tribe candidates: lower deposits “cannot be said to be discriminatory

at all, as this concession

has been

granted

in order

to enable

the

members of the Scheduled Castes and Tribes who are economically backward and financially incapable to contest elections.” The court does not mention Art. 15(4), butit is responding to a broadside claim by petitioner that the provision violates Art. 14, a contention that is stated by the court “only for rejecting it.” Although it is not mentioned in the opinion, the A. I. R. headnote writer lodged this aspect of the case under Art. 15(4). 55. Chait Ram o. Sikander, A. 1. R. 1968 Pat. 337.

”

Article 15(4) extends to all areas of governmental activity which are

not controlled by a more specific provision.51 Thus, preferences in housing, education, welfare, government contracts,5? and local political arrangements‘3 are authorized by, and must be within the scope of, Article 15(4). Unlike Article 16(4), this Article does not limit the State to reservations as a method of preference. Reservations may be used to implement Article 15(4), but government may also employ methods such as fee concessions, scholarships, special facilities (housing, medical, etc.) which do not involve reservations at all.54 Special treatment need not take the form of diversion of resources. It may instead involve special protection, as for example, protection against sale of land in execution of a creditor’s judgment. The scope of the “protection” authorized is not restricted by the qualifiers “socially and educationally,” which describe the classes who may be benefitted, but may extend to protection from economic exploitation.55

374

The Judicial Process

backward classes? Or suppose the State were to allot scarce educational opportunities to persons on the basis of an undertaking to provide ser-

vices to legitimate beneficiaries of Article 15(4)? Curiously, only one

court seems to have encountered this kind of argument. In Pradip Tandon v. State of Uttar Pradesh, the state contended that its reservations

for hill and rural areas were “made . . . to provide medical service . . .

to people of those areas.”’56A Full Bench of the Allahabad High Court @ thought this argument belied by a report of a committee on medical

education which suggested an urban bias in medical education and called for various measures to make rural practice more attractive to medical graduates. The Full Bench observed: Medical graduates hailing from rural areas may also be disinclined to return to the villages for medical practice on account of poor facilities. .. The

[state’s] assertion . . . that the reservations . . . was with a view to feed

the dispensaries of these areas appears to be pretentious and cannot be justifiable ground for making reservations. Such grounds of reservation of seats ... will be wholly outside the ambit and scope of Article 15(4). . . .57

The court here appears to assert that reservations have to be based on

the deservingness of the recipients rather than on some beneficial consequences for a backward class. The court here did not believe there was a legitimate backward class involved, nor, apparently, did it accord the argument much weight as a serious statement of government policy.

Hence the issues that lie buried here were not addressed. If the State can make “any special provision” for the “advancement” ofa backward class, may it not do so by conferring benefits on non-members of that

class (e.g., those who undertake to provide services in undersupplied

areas) through whom benefits will be delivered to the backward class?

Since the immediate beneficiaries of such schemes would be selected on grounds of the service to be provided rather than on any of the grounds forbidden by Articles 15(1) or 29(2), it is not clear that Article 15(4) need come into play. But ifit does, it is difficult to see any constitutional warrant for reading it.so narrowly that it could not encompass such

schemes.5®

56. A. 1. R. 1975 All. 1, at 4. This decision was reversed in part by the Supreme Court,

A.1.R. 1975S. C. 563. The Supreme Court did not take up this “distribution of medical

services” justification of the scheme, but it did uphold a reservation for one of the areas on behalf of which the state made this argument. See discussion in chap. 8, §H.

57. A. I. R. 1975 All. 1, at 7. The court dismissed as “irrelevant” to Art. 15(4) a

suggestion that the U.P. government take a bond from medical graduates to serve in

tural areas.

58. Art. 16, on the other hand, specifically refers to reservations for members of the

backward classes. But presumably, selection of non-backward-class employees on grounds of their capacity to benefit backward classes would not be on the basis of any

Constitutional Scope

375

Article 15(4) authorizes not only direct protections, but also provi-

sion for protection by political self-help. The Constitution specifically provides reserved seats in the lower house of Parliament’? and in the

lower houses of the state legislatures for the Scheduled Castes and Scheduled Tribes. These are the only special reservations in legislatures; the Constituent Assembly definitely rejected political safeguards for religious and other minorities.6 All of the other provisions for preferences are merely authorizations empowering the State to make

special provision for the backward. These legislative reservations are the only ones where a reservation is specifically enacted by the Con-

stitution itself.

Seats are reserved in the proportion of the population of Scheduled Castes and Tribes to the total population and are to be filled by joint election. These reservations do not involve “separate electorates” —

i.e., the representation ofa particular group by legislators chosen by an electorate composed solely of members of that‘group.®2 The seats are “reserved” in the sense that candidates who stand for them must belong to the privileged groups, but the entire electorate participates in choosing among candidates so qualified. Separate electorates for Parliament and the state legislatures are specifically outlawed by Article 325, which provides that no person shall be excluded from any electoral roll on

grounds of religion, race, or caste. The Supreme Court has indicated that Article 15 prohibits communal electorates in local bodies.

Contrary to widespread misapprehension, these legislative reservations are the only-ones that are subject to a constitutional time limit. It was originally provided that such reservations should expire ten years

ground forbidden in Art. 16(2) and would thus stand or fall on general classification grounds unless Art. 16(4) were read restrictively as exhausting all governmental power to confer special treatment on backward classes in the area of government employment.

59. Art. 330. 60. Art. 332.

61. XII CAD 299.

62. By separate electorates is meant the representation of religious (or other) minorities by legislators chosen by an electorate composed only of members of that minority. The provision and extent of such representation was an extremely troublesome political issue in India during the forty years preceding Independence. The Govenment of India Act, 1909, gave separate electorates to Muslims; the Government

of India Acts of 1919 and 1935 provided separate electorates for Muslims, Sikhs, Indian Christians, and other groups. Proposals to give separate electorates to the Scheduled Castes under the 1935 Act were withdrawn after adamant resistance by M.K. Gandhi (see Ambedkar 1946). They were given instead reserved seats in the legislatures—i.e., only they could be candidates for these seats, but the whole general electorate chose among the candidates. Separate electorates were a constant focus of dispute between

Hindus and Muslims (see Coupland 1944; Dalal 1940). Concern about separate electorates gave rise to extreme sensitivity to respective population figures of minority groups. 63. Nain Sukh Das v. State of U.P., A. 1. R. 1953 S.C. 384

376

The Judicial Process

after the commencement of the Constitution. In 1959 these provisions

were extended for another ten years,® in 1969 for another ten. and in

1980 for yet another ten.§7 The Constitution contains no provision for representation for Other Backward Classes in legislatures or in other political bodies. However,

such political safeguards may lie within the scope of Article 15(4), which is apparently broad enough to cover the reservation of seats in

elective bodies on a local level.©8 Presumably, Article 15(4) would not authorize additional reservation of seats in Parliament or in the state legislatures, since representation in these bodies is controlled by Parts XV and XVI of the Constitution, which provide reservations only for Scheduled Castes and Tribes and only in the respective lower houses. But Article 15(4) allows reservation for Scheduled Castes and Tribes (and presumably for Other Backward Classes as well) in local bodies such as district and municipal boards, panchayats, etc.6 Ordinarily, the use of communal criteria in conducting such elections would run afoul of Article 15(1),7° but Article 15(4) acts as an exception in proper cases. It is not clear whether special representation sanctioned by Article 15(4) must follow the legislative model of reserved seats from

joint electorates or whether Article 15(4) authorizes the use of separate

electorates in filling these reservations.’ Article 325, the ban on communal electorates, ,is concerned not with local bodies but only with Parliament and the state legislatures. 64. Art. 334.

65. Constitution (Eighth Amendment) Act, 1959.

66. Constitution (Twenty-third Amendment) Act, 1969.

67. Constitution (Forty-fifth Amendment)

Act, 1980.

68. A. R. V. Achar v. State of Madras, Writ Petition No. 568, High Court at Madras, August 25, 1952, aff'd on other grounds A. 1. R. 1954 Mad. 563, upholding reserved seats on municipal council for Scheduled Castes and women. Cf. Dattatraya v. State of Bombay.

A.1.R. 1953 Bom. 311, holding that Art. 15(3) authorizes such reservations for women. 69. See Dulichand v. Union of India, 1974 M. P. L. J. 200, 203, where the court,

ignoring Art. 15(4), holds that Art. 14 does not prevent the government from reserving places on a Cantonment Board for Scheduled Castes and Scheduled Tribes because they “form a class by themselves” and “a provision or reservation of seats for the backward classes in matters relating to election . . . makes a reasonable classification which hasa necessary nexus with the object sought to be achieved.” The use of the term “backward classes” here may be only as a synonym for Scheduled Castes and Tribes; clearly it is only provision for the latter groups that was before the court. The classification grounds that the court invokes might be applied to Other Backward Classes as well.

70. Bhopalsingh v. State, A. 1. R. 1958 Raj. 41; Nain Sukh Das v. State of U.P., 1953 S.C. 384.

71. A. R. V. Achar v. State of Madras, n. 58. If separate electorates are permissible to insure representation for backward groups, it would presumably be equally permissible to accomplish the same result by delineation of wards on caste lines as is reported to be

done in Orissa (RCSCST

1960-61: I, 20).

Constitutional Scope

377

C. THE EXCEPTIONAL CHARACTER OF COMPENSATORY DISCRIMINATION We have seen that in the text of the Constitution, the general principle of compensatory discrimination is established as a Directive Princple, but the specific provisions authorizing it are framed as exceptions to general Fundamental Rights. This arrangement expresses the basic tension between the broad purposes to be achieved and the commitment to confine the device and make it comport with other constitutional commitments, especially that to formal equality. The exceptional character of compensatory discrimination has frequently been noted by the courts. That Article 15(4) “has to be read as a proviso or an exception to Articles 15(1) and 29(2)” was evident from

its history, according:to the Supreme Court in Balajt v. State of Mysore.72

This has been the received characterization of Article 15(4). Thus in

State of Andhra Pradesh v. Sagar, the Supreme Court emphasizes that as an

exception, Article 15(4) cannot be extended so as in effect to destroy the

guarantee of equality in Article 15(1).73 In Devadasan v. Union of India, the Supreme Court emphasized that Article 16(4) “tis by way ofa proviso or an exception” to Article 16(1) and “cannot be so interpreted to nullify or destroy the main provision.” Thus its “over-riding effect” is only to permit a “reasonable number of reservations ...in certain circumstances. That is all.”7* The characterization as an exception was challenged by Subba Rao, J., dissenting in Devadasan. In his view, Article 16(4) “has not really carved out an exception, but has preserved a power untrammeled by the other provisions of the Article.”75 An extension of this argument was put

forward in C. A. Rajendran v. Union of India, where it was urged that Article 16(4) was not merely an exception engrafted on Article 16, but was itselfa Fundamental Right granted to the Scheduled Castes and Tribes and untrammeled by any other provision of the Constitution.76 The Supreme Court’s response was that Article 16(4) imposed no duty on the government to make reservations for these classes, but that “Article 16(4) is an enabling provision and confers a discretionary

power on the State to make a reservation.”77

72. A.1. R. 1963 S. C. 649, 657. 73. A. 1. R. 1968S. C. 1379, 1382. The characterization of Art. 15(4) as an exception

reappears in State of A.P. v. Balaram, A. 1. R. 1972 S. C. 1375, 1394, and Janki Prasad vr.”

State of J. @ K., A.1, R. 1973 S. C. 930, 937. 74. A.1.R. 1964S. GC. 179, 187.

75. Id., at 190.

76. A. 1. R. 1968 S. C. 507. The argument was based upon Subba Rao, J.’s dissent in Devadasan v. Union of India, A. 1. R. 1964 S.C. 179. 77. A. 1. R. 1968S, C. at 513. More than a decade later, when the issue of reservations had become fiercely controversial, a Working Group appointed by the Union Home

378

The Judicial Process

But if the particular means are discretionary, the object is not: the Constitution explicitly declares it “the duty of the State” to promote the

interests of the “weaker sections” and to protect them.78 And if these provisions are exceptions, they are exceptions of a peculiar sort. They do not merely carve out an area in which the general principle of equality is inapplicable. Rather, they are specifically designed to imple-

ment and fulfill the general principle.”?

Articles 15(4) and 16(4) are undoubtedly exceptions to the constitu-

tional prohibition of State employment of the otherwise forbidden

criteria of caste, religion, and so forth. But it does not follow that they are exceptions to the policy of equal treatment mandated by Articles 14, 15, and 16. In respect to the general policy of equality they represent an

empowerment of the State to pursue substantive equality in regard to the disparities between backward classes and others. It might be ar-

gued that in a State generally committed to formal equality this com-

mitment to reduce these disparities is an exception. Thus the State

could pursue a policy of overcoming the inequalities of backward groups that would lie outside its mandate in dealing with the poor, the

handicapped, veterans, or victims of disasters. But it seems clear that these disparities of circumstance and fortune may be addressed by

policies utilizing reasonable classifications. So, what these clauses do is

to insure State power to pursue substantive equality vis a vis historic formations in Indian society. It was the realization that mere provision of formal equality not suffice to bring about the desired “EQUALITY of status opportunity” that led to the adoption of these provisions. As Bench of the Kerala High Court observed:

certain would and of a Full

It has however been realized that in a country like India where large sections of the people are backward socially, economically, educationally

and politically, these declarations and guarantees [of equality] would be meaningless unless provision is also made for the uplift of such backward

classes who are in no position to compete with the more advanced classes. Thus to give meaning and content to the equality guaranteed by Articles 14, 15, 16, and 29, provision has been made in Articles 15(4) and 16(4) enabling preferential treatment in favour of the “weaker sections.””®° Ministry suggested that Art. 16(4) be replaced by a mandatory provision for reservation

of posts in state and public sector employment and also in those areas of private employment where reservations might be enforced (Overseas Hindustan Times, 21 June

1979). 78. Arts. 37, 46. 79. Cf. Gupta 1969:83, who notes that these provisions “are only to be viewed as

defining the principle of equality as contained in Art. 14. They do not derogate from the

principle of equality in any manner.” 80. Hariharan Pillai v. State of Kerala, A. 1. R. 1968 Ker. 42 at 47-8.

Constetutional Scope

379

Indeed, as the Supreme Court has observed, guarantees of equality

might by themselves aggravate existing inequalities. If taken literally,

instead of giving equality of opportunity to all citizens, it will lead to glaring

inequalities .. . [I]n order to give a real opportunity to [the backward to]

compete with the better placed people . . . [Article 16(4) is included in the

Constitution). The predominant concept underlying [Article 16] is equality of opportunity in the matter of employment; and, without detriment to said concept, the State is enabled to make reservations in favour of backward

classes to give a practical content to the concept of equality.®*

The tension between these commitments to non-discrimination and to substantive equalization was poignantly expressed by Prime Minister Nehru when he remarked in the course of the First Amendment debate that “we arrive at a peculiar tangle. We cannot have equality because in trying to attain equality we come up against some principles of equality.””®? The textual juxtaposition of guarantees of equality and

authorization of compensatory discrimination reflects a deeper conflict

between different views of equality and divergent notions of the goal and scope of protective discrimination. While in practice these views tend to merge and overlap, it may be helpful here to isolate in pure form what we may conveniently label the horizontal and the vertical perspec-

tives on equality and compensatory discrimination. In the horizontal view, the relevant time is the present. Equality is visualized as identical opportunities to compete for existing values

among those differently endowed, regardless of structural determinants

of the chances of success or of the consequences for the distribution of values. One of the dissenting judges in the Thomas case sums it up neatly when he cautions that Article 16(1) “‘s; of equality of opportunity,

not opportunity to achieve equality.”® In this view, preferential treatment is accepted as a marginal adjustment to be made where results of complete equality are unacceptable. Compensatory discrimination detracts from equality: it amounts to a kind of social handicapping to

insure fair present distribution among relevant units. Thus in Devadasan v. Union of India, the Supreme Court emphasized that in combining

these provisions it might strike a fair balance between the claims of the

backward and the claims of other communities. The relation of equality and compensatory discrimination is viewed 81. Triloki Nath Tiku v. State of Jammu and Kashmir, A. 1. R. 1967 S. C. 1283 at 1285.

82. Parliamentary Debates, Vol. XII-XIII (part II), col. 9617 (29 May 1951). 83. Gupta, J., dissenting in State of Kerala v. N. M. Thomas, A. I. R. 1976 S.C. 490 at

543. 84. Devadasan v. Union of India, A. 1. R. 1964 S. C. 179; cf. C. A. Rajendran v. Union of India, A. 1. R. 1968 S. C. 507.

380 The Judicial Process

very differently in what we may call the vertical. perspective. In this

view the present is seen as a transition from.a past of inequality toa.

desired future of substantive equality; the purpose of compensatory discrimination is-to-promote equalization by offsetting historically accumulated inequalities. Thus, compensatory. discrimination does not detract from equality in the interest of present fairness; rather, it is'seen as requisite to the fulfillment of the nation’s long-range goal of substantial redistribution and equalization. Not only present claims but histori-

cal déprivations and national aspirations are relevant. Such a view was given its most clear judiciat expression in Viswanath v. Government of Mysore, where Hegde, J., repulsed the argument that reservations should

be confined with the observation that counsel

did not appear to be very much alive to the fact that theré can be neither stability rior real progress if predominant sections of an awakened Nation

live ‘in primitive conditions, confined to unremunerative occupations and having no share in the good things of life, while power and wealth are

confined in the hands of only a few and the same is used for the benefit of the

sections of the community to which they belong. . . .[(U]naided many sec-

tions of the people, who constitute the majority in this State cannot compete with the advanced sections of the people, who today have a monopoly of education and consequently have predominant representation in the Gov-

ernment services as well as in other important walks of life. It is cynica} to

suggest that the interest of the Nation is best served if the barber’s son continues to be a barber and a shepherd’s son continues to be a shepherd. . . . We have pledged ourselves to establish a welfare State. Social justice is an important ingredient of that concept. That goal cannot be reached if we overemphasize the “merit theory.” Advantages secured due to historical reasons cannot be considered as fundamental rights guaranteed by the Constitution. The nation’s interest will be best served—taking a long-range view—if the backward classes are

helped to march forward and take their places in a line with the advanced sections of the people.®5

From this “long-range” perspective, Justice Hegde later elaborated,

the “immediate advantages of the Nation [in the effective utilization of

talent] have to be harmonized with its long-range interests.”"8¢

Rarely do these contrasting views of equality and compensatory discrimination appear with such purity and clarity; more common is 85. A. T. R. 1964 Mys. 132 at 136. A more summary and less vivid version of these

observations was delivered by Justice Hegde after his ascension to the Supreme Court in Periakaruppan v. State of Tamil Nadu. 86. Periakaruppan v. State of Tamil Nadu, A. 1. R. 1971S. C. 2303 at 2309.

Constitutional Scope the attempt

to harmonize

them. Thus

Gajendragadkar, J., observes:

381

in Balaji v. State of Mysore,

It is obvious that unless the educational and economic interests off the

weaker sections of the people are promoted quickly and liberally, the ideal of establishing social and economic equality will not be attained. . . .87

Surely the State is authorized to take “adequate steps” toward that

objective, but these special, exceptional provisions do not override the Fundamental. Rights of others. Furthermore, there are other crucial national interests which have to be taken into account:

The interests of weaker sections of society which are a first charge on the

States and the Centre have to be adjusted with the interests of the community as a whole.®*

The inevitable weighing and balancing is rendered particularly difficult because the constitutional provisions set'up another tension as to the relevant units whose interests are to be balanced or who are ultimately. to be equalized, as the case may be. The Constitution confers Fundamental Rights on individual citizens in their personal capacity, not as members of communal groups. All citizens have a fundamental

right that another, excepting a member ofa backward class, shall not be preferred by the State on the basis of his membership in a. particular

group. However,

the government is obliged to advance the weaker

sections or backward classes. Therefore the scope of compensatory discrimination involves tension between. individuals or groups as. ab-

jects of State policy.

The: constitutional embrace of the antagonistic principles of equal treatment and compensatory discrimination, individual rights and group rights, confronts both government and courts with the problem of reconciling them in specific settings. The sweeping language of Articles 15(4) and 16(4) indicates that their framers relied primarily on the discretion of the politicians and administrators of the future, rather than on the courts, to effect such a reconciliation. But while these

provisions give the executive and the legislatures broad discretion in their application, judicial intervention is not entirely excluded. Leverage for judicial oversight is supplied by the placement of these provi-

sions as exceptions to the judicially enforceable Fundamental Rights.

These rights can only be vindicated to the extent that the courts scrutinize the government’s designation of backward classes to see that

only the backward are included, that the extent or method of operaiton does not prejudice others unduly, that the schemes are designed and 87. A. I. R. 1963. S. C. 649 at 661.

88. Id., at 663.

382

The Judicial Process

administered to work in favor their detriment. Such review vidual rights but to effectuate venting unwarranted dilution inclusion

of beneficiaries,

of the intended beneficiaries and not to is necessary not only to vindicate indithe policy of these provisions—by preof benefits (for the more unrestrained the

the less the assistance

the intended

ben-

eficiaries will receive) and abuses that undermine public support for

these measures.

As we shall see, the courts have indeed played a major role in shaping

policy in this area by defining the constitutional boundaries of preferential treatment. Recent developments

in the constitutional doctrine of

equality indirectly raise the quesiton of whether courts might play an even more central role in the design and implementation of compensat-

ory policies.

D. ANEW CONSTITUTIONAL VISTA? THOMAS AND AFTER For the first quarter-century of constitutional development, the ap-

proaches discussed in the previous section bounded the discourse about the way in which competing commitments to formal equality and com-

pensatory discrimination might be combined. But much of the earlier

understanding of constitutional policies of compensatory discrimina-

tion is cast into doubt by a remarkable 1975 decision of the Supreme

Court in State of Kerala v. N. M. Thomas.®? Since there will be many occasions to refer to this decision, I will set out the story here. As will

be evident, it is far from clear what it means or what its reception will be.

Employees of the Registration Department of the State of Kerala were divided into Lower Division Clerks and Upper Division Clerks. The former could be promoted to the higher position on a ‘“‘Senioritycum-merit” basis. To qualify for promotion it was necessary to pass some tests—‘‘Account Test ...Lower Kerala Registration Test and

. Test in the manual of office procedures” —this was the merit

89. A. 1. R. 1976S. C. 490.

90. The version of the facts here is a composite drawn from the accounts provided at

‘various places in the seven opinions. I noted no major discrepancies in these accounts. See A. I. R. 1976 S.C. at 493-95, 496-97 to 501, 502-04, 520, 523. One factual question puzzled me: what would happen to these clerks at the expiration of the extension period if they had not passed the tests? My initial reading was that during the period that they were incumbents in the higher posts, they were eligible for promotion on seniority, and since they were exempt from passing the tests, they would be confirmed in the higher posts on grounds of seniority. But the judges are unanimous that this is not the case, that the extension is only temporary, and that the tests must be passed before they can be confirmed in the higher posts (Ray, C.J., at 502, Mathew, J., at 520, Beg,J. at 523, Fazl

Ali,J. at 544).

Constitutional Scope

383

requirement. Among those who satisfied this prerequisite, the promotions went to the most senior Lower Division Clerks. The rules allowed for temporary appointments to the higher posts for a two-year period during which the clerk would have to pass the required tests; the two years was extended to four years in the case of Scheduled Caste and Scheduled Tribe clerks. Nevertheless, a number of SC clerks had not

satisfied the test qualifications within the extended period and were facing reversion to the lower posts. In 1972 the state government promulgated a new rule, 13AA:

Notwithstanding anything contained in these rules, the Government may,

by order, exempt for a specified period, any member or members, belonging to a Scheduled Caste or a Scheduled Tribe, and already in service, from

passing the tests.

On the same day the state promulgated an order granting Scheduled Castes and Scheduled Tribes already in service “temporary exemption

. . from passing all tests . . . for a period of two years.” In 1974 this was extended for a further period to insure each employee two chances

to appear for the required tests. This time the government ordered that “these categories of employees will not be given any further extension of

time to acquire the test qualifications.”

Because of earlier difficulties with the test barrier, there was a heavy

concentration of SC Lower Division Clerks with high seniority. When the test barrier was removed temporarily, many of them were promoted

to the higher posts. Thus in 1972, of 51 vacancies in the category of

Upper Division Clerks, 34 were filled by SCs who had not passed the

tests, and only 17 were filled by persons who had

petition was filed by N. M. Thomas, a Lower

passed the tests. A writ

Division Clerk who did

pass the test and would have been promoted but for the extensions authorized by Rule 13AA. A Division Bench of the Kerala High Court concluded: “What has been done is not to reserve ...posts....[RJeservations had already been made... .What has been attempted by Rule 13 AA is to exempt persons from possessing the necessary qualifications.” Such exemption lies beyond the scope of Article 16(4)’s authorization of reservations, and on the scale it is done here, directly violates Article 335’s directive that claims of Scheduled Castes and Tribes may be taken into account

in government employment “consistently with the maintenance of efficiency of administration.”91 On appeal to the Supreme Court, counsel

for the state took an innovative tack and argued that the extension 91. Thomas v. State of Kerala, 1. L. R. 1974 (1) Ker. 549 at 556-57. The quantum of

preference aspects of the case are discussed in chap. 12, §B, below.

384

The Judicial Process

need not be subsumed under Article 16(4)’s provision for reservations, but

could

be justified

as a reasonable

classification

under

Article

16(1).92 A seven-judge bench decided five to two to reverse, issuing

seven separate opinions.93 One of the majority judges thought that Article 16(4) rightly interpreted would authorize the state’s provision;

the other four accepted some version of the broad classification argument adyanced by the state.

Chief Justice Ray’s opinion for the majority sketches the outline of the

classification argument: providing equal opportunity in government

employment is a legitimate objective; Article 46 directs the State to

promote the economic interests of Scheduled Castes and Tribes with special care; Article 335 directs the State to take into their consideration

their claims regarding service under the State. Thus the classification of employees belonging to these groups to afford them an extended period to pass tests for promotion is “‘a just and reasonable classification having Tational nexus to the object of promoting equal opportunity . . . relat92. In the High Court, counsel for the state had argued unsuceessfully that the state’s

action fell within the meaning of “reservation” in Art. 16(4). I. L. R. Ker. 1974(1) at 561. Counsel had also advanced an ingenious textual argument that if the reservations were covered by Art. 16(4), then preference in government employment apart from reservations was authorized by Art. 15(4), but the court held that Art. 16(4) covered the employment area and its provisions exhausted the goverment’s power to deviate from the guarantees of Arts. 16(1) and 16(2) (jd.,at 557). There is no indication that counsel urged that Arts, 16(1) or 16(2) themselyes conferred power to make such arrangements. 93. State of Kerala v. N. M. Thomas, A. 1. R. 1976 S. C. 490. The appeal was argued just after the onset of Mrs. Gandhi’s Emergency Rule, and the decision was announced on

September 19, 1975, at the end of three months of emergency rule and at the height of optimism that it heralded an “egalitarian breakthrough.” 94. Although Thomas marks a sharp departure from earlier doctrinal analysis of the special provisions for backward classes, it was not without precursors and anticipations. us classification standards were used to permit reservations for various deserving

groups in medical admissions. See Chanchala v. State of Mysore, A. 1. R. 1971 S. C. 1762

and other cases discussed in chap. 12, §D, below. For example, Sardool Singh v. Medical

College, A. 1. R. 1970 J. & K. 45, applied general Art. 14 classification standards to

it reservations in medical admissions for children of defence personnel, noting that Art. 14 permits such classification quite apart from the specific provisions for reserva-

tions in Arts. 15(4) and 16(4). Accord: Kushma Joshi v. Pro-Vice Chancellor, A. 1. R. 1969 J. & K. 136. Reservations of government posts for ex-servicemen were upheld on a

similar ground in Daya Ram v. State of Haryana, A. 1. R. 1974 P. & H. 279. The

classification argument might be made where the beneficiaries would come under Art.

15(4) or 16(4), as in Raju v. Chief Electoral Officer, A. 1. R. 1976 Guj. 66, decided nine

months before Thomas. The High Court had no difficulty in justifying a lower electoral

deposit for Scheduled Castes and Tribes on Art. 14 grounds. See discussion at n. 54,

above. The extent to which classification principles were perceived as available to

justify compensatory preference is suggested by the remarks of Chief Justice Hidayatullah in Golak Nath v. State of Punjab, A. 1. R. 1967 S. C. 1643 at 1706, discussed in n. 24, above.

Constitutional Scope

385

ing to [public] employment.”95 The difference in condition of these

groups justifies differential. treatment; just.as rational classification is permissible under the general equal protection provision of Article 14, so it is permissible to treat unequals unequally under Article 16.9% The extent of the doctrinal innovation here can be appreciated by.

considering the opinion of Justice Beg, the only member of the majority. wha does not participate in the reconceptualization of Article. 16. He helpfully restates a conventional understanding of the constitutional provisions. In this view. “the guarantee contained in Article 16(1) isnot by itselfaimed at removal of social backwardness due to socio-economic

and educational disparities produced by past history of social oppres-

sion, exploitation, or degradation of a class of persons.”’97 Instead, “it was in fact intended to protect the claims of merit and efficiency. against incursions of extraneous considerations.”%* And efficiency tests, in turn, “‘bring out and measure ... existing inequalities in competency and capacity or potentialities so as to provide a fair and rational basis for justifiable discrimination between candidates.”

Thus provisions for equality of opportunity are meant to insure “fair

competition” in.securing government jobs; they are not directed to “removal of causes for unequal performances.” But such provi-

sions do not stand alone: they are juxtaposed with Articles 46 and 335, which imply “‘preferential treatment for the backward classes” to miti-

gate the rigor of equality in the same sense of strict application. of uniform tests of competence. Article 16(4) was designed to reconcile the conflicting pulls representing... justice conceived of as equality (in the - competition) and of Articles 46 and 335, embodying the duties . promote the interests of the economically, educationally and

of Art. 16(1), conditions of of the State to socially back-

ward, so as to release them from the clutches of social injustice.1°1

Thus Article 16 (4) may be thought to “exhaust all exceptions made

in favour of backward classes.”’102 Yet the effect of the Kerala promo-

tion rules here is ‘“‘a kind of reservation,” for it is a temporary promotion that would be confirmed only if the appointee satisfied specified tests within a given time. These rules may be viewed as “implementation ofa

policy of qualified or partial or conditional reservations” which could

“be justified under Article 16(4).”23 Dismissing the interpretation of Article 16(4) as an exception to Article 16(1), Justice Mathew articulates the view of equality that 95. A. I. R. 1976 S.C. at 500. 96. Id., at 502. 97. dd., at 522. 98. Id., at 522. 99. Id., at 522. 100. Id., at 522. 101. fd., at 522. 102. Hd., at 522. 103. Id., at 524.

386

The Judicial Process

implies the doctrinal shift. The equality of opportunity guaranteed by the Constitution is not only formal equality with fair competition, but “equality of result.”2% In order to assure the disadvantaged “their due share of representation in public services” the constitutional equality of opportunity was fashioned “wide enough to include . . . compensatory measures.”1°5 Thus the guarantee of equality “implies differen-

tial treatment of persons who are unequal.” Article 16(1) is “only a part of a comprehensive scheme to ensure equality in all spheres.”1°7 It implies “affirmative action” by government to achieve equality— that is, “compensatory state action to make people who are really un-

equal in their wealth, education or social environment. . equal.”

If equality of opportunity guaranteed under Article 16(1) means effective

material equality, then Article 16(4) is not an exception to Article 16(1). Itis only an emphatic way of putting the extent to which equality of opportunity could be carried viz, even up to the point of making reservation. 1

Thus

[t]he state can adopt any measure which would ensure the adequate rep-

resentation in public service of the members of the Scheduled Castes and Scheduled Tribes and justify it as a compensatory measure to ensure

equality of opportunity provided the measure does not dispense with the

acquisition of the minimum basic qualification necessary for the efficiency of administration.12°

Justice Krishna Iyer propounds a complex vision of the constitutional commitment to equality. Interpreting the Constitution by “a spacious, social science approach, not by pedantic, traditional

legalism,”*"! he proposes to erect a “general doctrine of backward classification” to pursue “real, not formal, equality.”"!? According to the doctrine of backward classification, the State may, for purposes of securing genuine equality of opportunity, treat unequals equally. Thus Article 16(4) serves not as an exception [to the strictures of Article 16(1) and (2)} but as

an emphatic statement, one mode of reconciling the claims of backward people and the opportunity for free competition the forward sections are ordinarily entitled to . . ..Closely examined it is an illustration ofa constitutionally sanctified classification.13 104. Id., at 518.

105. Id., at 518.

106. Id., at 516.

109. fd., at 519.

110. dd., at 519.

111. dd., at 525.

107. fd., at 519.

108. Jd., at 516. The idea of affirmative action 1s attributed to the United States Supreme Court, citing cases involving the unconstitutionality of the poll tax:and the right of criminal defendants to a free transcript and counsel on appeal. 112. Id., at 529.

113. Hd., at 535.

Constitutional Scope 387 So, in addition to reservations provided by Article 16(4), the State may also confer “lesser order[s] of advantage” on the principle of classification under Article 16(1).14*

At this point in the argument there is a crucial divergence between the views of Justice Krishna Iyer and those of Justice Mathew. For the

latter, the compensatory measures authorized by Article 16(1) might be extended to “‘all members of the backward classes,”115 not only to the Scheduled Castes and Scheduled Tribes. But Justice Krishna Iyer’s more complex vision of Article 16 contains a second layer: the power of classification outside the boundaries of Article 16(4) for purposes of overcoming inequality may be used only on behalf of Scheduled Castes and Tribes. Article 16(4) covers all backward classes, but to earn the benefit of grouping under Article 16(1) based on Articles 46 and 335

.. . the twin considera-

tions of terrible backwardness of the type harijans have to endure and maintenance of administrative efficiency must be satisfied.116

“Not all caste backwardness is recognized” as a basis for differential treatment under Article 16(1).127 The differentia . . . is the dismal social milieu of harijans. . . . The social

disparity must be so grim and substantial as to serve as a foundation for benign discrimination. If we search . . . we cannot find any large segment other than the Scheduled Castes and Scheduled Tribes . . . .[N]o class other

than harijans can jump the gauntlet of “equal opportunity” guarantee.

Their only hope is in Article 16(4).118

This is perplexing, for it appears that the stronger measures of reserva-

tion may be taken on behalf ofall the backward, yet those who suffer the most terrible backwardness are the only ones entitled to measures of “a

lesser order of advantage.” Later he suggests that to allow the Other Backward Classes to parti-

cipate in these benefits may be detrimental to those who are most

deserving:

[N]o caste, however seemingly backward .. .can be allowed to breach the dykes of equality of opportunity guaranteed to all citizens. To them the answer is that . . . equality is equality ....The heady upper berth occup-

ants from backward classes do double injury. They beguile the broad

114. Id., at 536. 115. Id., at 519. 116. dd., at 536. Although here and elsewhere in his opinion, Justice Krishna Iyer uses the term “harijans,” his explanations suggest that he is employing this as a shorthand

term for both Scheduled Castes and Scheduled Tribes. 117. Id., at 537. 118. Id., at 537.

388

The Judicial Process

community into believing that backwardness is being banished. They rob

the need based bulk of the backward of the ... advantages the nation proffers.119

This distinction is justified because “the Constitution itself makes a super-classification between harijans and others, grounded on the fun-

damental disparity in our society and the imperative social urgency of

raising the former’s sunken status.’’'?° From the provision of Articles 330, 332, 335, 338, and others, we may deduce that “the Constitution

itself demarcates harijans from others....This is based on the stark backwardness of this bottom layer of the community.”!21 This constitutional differentiation of Harijans is specifically extended to the area of government employmentas part of the State’s obligation to “promote the economic interests of harijans and like backward classes.” Articles 14 to 16 are, according to Justice Krishna Iyer, “‘the tool kit” to carry

out the “testament” of Articles 46 and 335.122

An attenuated form of this “super-classification” argument is found in each of the four majority judgments that embrace the classification

argument. That argument posits a general authorization flowing from Article 16(1) to adopt reasonable classifications for purposes of securing equality of opportunity. But it is conceded that this does not include a power to employ those classifications specifically forbidden in Article 16(2). Therefore each of these four “classification” opinions argues that the Scheduled Castes and Tribes do not constitute a classification on

the basis of “caste.”

Thus Justice Krishna Iyer argues that the ban of 16(2) doesn’t arise

in connection with measures for the Scheduled Castes and Scheduled

Tribes, for they are “no[t] castes in the Hindu fold but an amalgam of castes,

races,

groups,

tribes,

communities

or

parts

thereof found...

to be the lowliest and in need of massive State aid and notified as such by the President.’’!23 Article 16(4) allows use of these forbidden grounds to identify the backward classes that may be recipients of reservation. The four judgments imply that the other kinds of compensatory treatment justified directly by Article 16(1) are available only to classes which avoid the classification forbidden by Article 16(2), includ-

ing Scheduled Castes and Tribes, since they are not castes. But if the

idea is to confine compensatory classification under Article 16(1) to

SC and ST, this argument proves too much. For there are innumerable

categories—e.g., on the basis of income, occupation, physical hand119. 121. 123. at 549,

Id., at 539. 120. fd., at 532-33. Id., at 533. 122. fd., at 533. Ct: Chief Justice Ray, id., at 501 ; Justice Mathew, id., at 519; Justice Faz! Ali, id., 552. But cf. Justice Gupta, id., at 542.

Constitutional Scope 9389 icap, etc.— that are not based on the classifications forbidden in Article

16(2). The argument for SC and ST as a “‘super-classification” cannot

be sustained on the basis of the structure of Article 16, but only on

the basis of their special recognition in the Constitution. But if that is the argument

for their distinctiveness,

the fact that they are not

“castes” within Article 16(2) is a distracting irrelevance. In this and other matters, the Thomas opinions leave open many perplexing questions. Are equalizing measures permissible even if they employ the categories forbidden by Article 16(2)? Or, conversely, is the power to adopt equalizing classifications under Article 16(1) to be used exclusively to address disparities along the dimensions listed in Article 16(2)? How about differences in class? income? Is the doctrine of classification merely permissive, so that government may make such compensatory classifications, but need not? Suppose it fails to do so? Do classifications that do not take account of inequalities violate the Article 16(1) guarantee of equality of opportunity? Yet Thomas is welcome because it makes reflection on compensatory treatment unavoidable. By revealing the constitutional indeterminacy and doctrinal disarray in this field, it poses new challenges for jurisprudence and policy. Where courts could once rely without much thought on quotations from Balaji or Sagar, they must now articulate their choices. Actors, governmental and private, can no longer assume that

the categories of compensatory policy are immutable. In terms of doctrinal housekeeping, there was little need Thomas reconceptualization of Article 16. “Reservation” in 16(4) could readily have been construed to accommodate the scheme (as it was by Justice Beg). Such a construction would

for the Article Kerala usefully

have clarified the status of the age-waivers, fee concessions, travel allowances, coaching schemes, lowering of minimum marks, and other

provisions that typically accompany reservations and often exist apart from reservations per se. If the State enjoys the plenary power to reserve places for a beneficiary group, does this not imply a power to take lesser measures to help the members of that group obtain places? In order to join issue with the large questions, the Thomas majority (the four) had to accede to a very narrow definition of “reservation” so that all of these

practices are now exposed to challenge and possibly to the necessity of securing justification on broad classification principles, a course which may be strewn with conceptual land mines. Thus the radical reconceptualization of Article 16 arrives on the agenda at the cost of an unimaginatively narrow reading of what common sense would regard

as the most relevant constitutional provision, Article 16(4).

A generous construction of Article 16(4) not only would have addressed the problem in Thomas, but would have been readily comprehensi-

390 The Judicial Process

ble to all the different groups concerned with reservations in government posts— officials in charge of appointments and promotions, government servants, and aspirants to government posts. Thomas reaches the same immediate outcome by elaboration of ambiguous and recondite doctrine that is not readily accessible to officials who must design and manage programs of preference, to the beneficiaries of such programs, to disappointed non-beneficiaries—and in large measure is not readily accessible even to lawyers and judges. Perha , though, the accomplishment of Thomas lies in its providing

government with more ample means to pursue compensatory policies. The doctrinal obstacles that might have impeded governmental policies in this area stood for the most part on weak ground, unlikely to with-

stand sustained conceptual attack.12 Openings for favorable reinterpretation were plentiful. To bestow on government more ample authority to do something does not automatically mean that more of it will get done. Was the critical shortage one of doctrine favorable to such policies? or the will, energy, and competence to implement them?

Consider the situation of the Kerala registration clerks in Thomas. The state’s scheme here seems less a carefully calculated modification of job requirements than a desperate improvisation amounting to a confession of failure ofits earlier policies. Extensions hadn’t worked before, so another one is tried. There is no indication that any thought was

given to (1) some way of helping the Scheduled Caste Lower Division Clerks prepare to pass these tests; nor to (2) modifying the tests to eliminate cultural biases or extraneous matters and measure qualities

genuinely needed for the job; or (3) modifying the job to make it suitable to candidates

with the qualifications of these applicants.

The im-

mediate thrust of the decision isi to enlarge the State’s authority to confer preferential treatment. But the failure of the earlier measures was

due not mainly—perhaps not at all— to lack of State authority, but to

lack of will or capability to make its schemes work. Enlarging State

authority will not necessarily supply the lack. Indeed, it may allow the

State to substitute easy shortcuts (like the one here) for thinking about how to do the job effectively. Since it may now in effect decree the result of keeping these clerks in the higher posts, it may have less incentive to

devise ways of motivating or enabling them to grasp these opportunities and improve their performance, thus prolonging their lack of qualification and reinforcing their dependence. And, as the State’s authority is more broadly defined, there will be fewer occasions for the courts to

124. E.g., the quantitative restrictions imposed by Devadasan, discussed in chap. 12,

§B, below, or the restrictions on the use of layers of preferential treatment, discussed in chap. 13, §C, below.

Constitutional Scope

391

observe state programs and to monitor and energize government performance. Of course, it is not only a question of the capability of the State, but also of the recipients of preferential treatment. The argument in favor of the extension of the Scheduled Caste clerks in the upper-division posts is that it will somehow motivate or equip them to pass the test eventually. And unless they are motivated and enabled to take the initiative and improve their own capacity to perform, the extension (and any compensatory measure for that matter) is only a temporary palliative. There is no indication that the State of Kerala in any way enhanced its

ability to elicit a satisfactory performance from these clerks. 125

Thus Thomas does not offer any hope of breaking out of the pattern of patronage and dependence. Both majority and minority judgments visualize the Scheduled Castes as the passive recipients of governmental largesse, rather than as active participants in their own improvement. And for all its apparent radicalism, Thomas enlarges State power in a way that may jeopardize the future of compensatory preference for the backward classes. In two senses this enlargement is a false victory for the Scheduled Castes. First, the Court passed up the chance to make the modest contribution of usefully expanding the meaning of reservations to align doctrine with existing practice about age waivers, coaching, etc. In taking the more radical course, the Court attempts to make legal doctrine yield up benefits that it cannot yield. Doctrine can give only authority, not

power. And the crucial shortage was not of State authority, but of the will and capacity of the State to deliver benefits (and of the recipients te ualive those opportunities to enlarge their capabilities). The result is symbolic

breakthrough

in which

Scheduled

Castes,

their well-

wishers, and wider publics are beguiled into thinking that much (or too much) is being done forScheduled CCastes. Second, the new reading of equality may detract from the attention and priority accorded to the backward classes. The new equality doctrine is so ample that it sweeps in its path all that confines the

commitment to compensatory treatment to specific historic groups. If

Articles 14—16 proclaim a regime of substantive equality and if the State may employ classification to remedy any falling short of the equality thus mandated, the government’s responsibility to confer compensatory preference is vastly larger than it has hitherto been understood. It is a responsibility that runs not only to Scheduled Castes and Tribes and 125. I am indirectly informed that by June 1978, most of the Scheduled Caste clerks

whose promotion sparked the Thomas case had been provisionally certified as Upper Division Clerks upon passing the departmental test.

392

The Judicial Process

to backward

classes more widely conceived, either in terms of social

groups or the poor; it also includes those who sufer difficulties as a result of personal misfortune (disaster victims), accidents of personal history (the physically handicapped) or as a result of meritorious service to the nation (ex-servicemen or dislocated children of diplomats). In a setting of chronic shortage, an enlarged commitment to remedy all undeserved difficulties betokens a commendable generosity of spirit.

But it also raises the question of priorities and of allocation of scarce

resources, including attention. Government’s authorization to pursue substantive equality is vastly greater than the resources that will conceivably be available to it. Among the claimants on its compensatory powers will be many who are better placed to press their claims on the attention and sympathies of government. Will not the commitment to the lowest social groups— especially where these are perceived to receive massive benefits—be overwhelmed by governmental response to better-placed claimants on its compensatory attentions? Of course, the State could always take account of these difficulties

under Article 14 in terms of reasonable classification. But the notion

that the State has a general obligation to produce substantive equality means that the kinds of disadvantage that afflict the better-off—diplomats, central government employees, retired army officers, physically

handicapped

children

in well-to-do,

educated

families—are

now

elevated, as far as compensatory responsibilities go, to parity with the government’s commitment to overcome disparities associated with the traditional social hierarchy. The sense of a regime of formal equality qualified by a singular exception to alleviate disparities derived from position in the traditional social hierarchy is liquidated or dissolved into a general and unfulfillable commitment to substantive equality. The distributive potential of this new dispensation is dramatically realized in Jagdish Rai v. State of Haryana, where the state reserved a substantial portion of government posts for ex-servicemen on the

ground that they were handicapped because “over the years [they] have

lost opportunities for entering government service and have also lost contact with ordinary civilian life.”!26 The Full Bench sweeps aside the notion that reservations have to be justified by Article 16(4) as “a relic of the old way of thinking . . . . The old idea has now given way to the idea that [Article 15(4) and Article 16(4)] . . . are themselves aimed

126. A. I. R. 1977 P. & H. 56, 61. Since India has no conscription, these were volunteers. Two separate reservations figured in this case: the Division Bench addressed a 50% reservation for “Ex-Emergency Commissioned Officers” as demonstrators in

a Dental

College

(id., at 57). On

reference to a Full Bench,

attention shifted to a

reservation of 28% of posts of sub-inspectors in the Food and Supplies Department for “released Army Personnel” (id., at 58).

Constitutional Scope

393

at achieving the very equality proclaimed and guaranteed by Article 14

and other clauses of Articles 15 and 16.127 After extensive citation from Thomas, which it commends for having “‘got rid of the old sterility” and “introduced a new dynamism and a new dimension into the con-

cept of . . . equality of opportunity,” the court observes that “[i]t is no longer necessary to ‘apologetically’ explain laws aimed at achieving equality as permissible exceptions. It can now be boldly claimed that such laws are necessary incidents of equality.” 128 Reservation of posts

for ex-servicemen is justified, for they suffer difficulties in competing with civilians for civilian jobs, and the state has an obligation to provide

them employment. Thus the state is justified in classifying them separately as a source of recruitment.!29 To secure a just proportion of

posts to those who suffer a peculiar handicap in competition is “an extension of the principle of Article 16(4) to those that do not fall under Article 16(4).”3°

In effect, other deserving groups are now entitled to reservations along with the backward classes specified by Article 16(4), and in effect all the kinds of preferential treatment which seemed to be allowed exclusively to the backward classes may now be bestowed on other groups regarded as deserving by the State. It is a fitting symbol that the losing petitioner in Jagdish Rai was herself a Scheduled Caste. Thus Thomas opens Pandora’s box: compensatory classification is available— perhaps incumbent— to succor all the disadvantaged. The earlier notion that “the interests of the weaker sections are a first charge”’131 on government is dissolved into a diffuse, and in the nature of things, largely symbolic, generalized egalitarianism. Justice Krishna Tyer’s doctrine of super-classification may be understood as an attempt to close the lid again, confining compensatory classification to Scheduled Castes and Tribes. Were this doctrine to gain acceptance, State authorization to confer benefits on these groups might end up not very far from where liberal reading of Articles 15(4) and 16(4) would have left it. Super-classification would re-establish the priority of Scheduled Caste claims by emphasizing a picture of Indian society as riven by an unbridgeable dichotomy between Scheduled Castes and Tribes and the rest of the population. This dichotomous picture, by drawing a rigid line between Scheduled Castes and Other Backward Classes, could impede program administration and obstruct the even127. Id., at 58. 128. Id., at 60. 129, Id., at 61. The “source” language is borrowed from Chanchala v. State of Mysore, ATR. 1971S. C. 1762 (discussed in chap. 12, §D, below). This borrowing confirms the suspicion that the distinction between making reservations and specifying sources is a

thin one indeed.

130. Id., at 61.

131. Balajiv. State of Mysore, A. 1. R. 1963 S. C. 649 at 651.

394 The Judicial Process tual dismantling of preferential treatment by preventing merger of these categories. More immediately, it would further stigmatize Scheduled Castes and Tribes by portraying them as uniquely hapless and helpless specimens, a potential demonstrated in the overtones of

condescension found in some of the Thomas opinions. But it would be shortsighted to judge Thomas only in terms of what it does to governmental power in this area. There is another side to the coin and one that is potentially of even greater significance. If Articles 14, 15, and 16 are read as mandating the pursuit of substantive equality, then to what extent is substantive equality—or at least governmental

efforts to promote it—an enforceable Fundamental Right? Since those

denied Fundamental Rights may resort to the courts to enforce them, would the potential beneficiaries of such equalizing measures have a

right to resort to the courts to secure governmental compliance with this right of equalization? It would not be surprising if the courts would shrink from affirmative

enforcement of these reconceptualized rights to equality. But imagine for a moment that they were willing to do so. Scheduled Castes and others would not have to wait for government preference in their behalf,

but could take the initiative in the courts to secure the “enforcement” or implementation of their rights to substantively equalizing measures. Scheduled Caste movements could mobilize around these issues, gene-

rating the kind of political movement that would make government responsive to judicial proddings. We shall in a moment mention the doubts that surround this scenario. But to emphasize it once more, in this reading the ultimate

signiticance of Thomas is not the enlargement of State authority to confer preferential treatment, but the acknowledgment of a Fundamental

Right to substantive equality and the possibilities for affirmative litigation by disadvantaged groups to force the State to fulfill its responsibilities. This would amount to an ironic reversal of the meaning of Thomas. We started with the obvious view that Thomas loosened judicially

imposed restraints on government, allowing it to patronize the least advantaged; we end by viewing Thomas as suggesting the imposition on government of a new and onerous accountability to these disadvan-

taged, an accountability mediated through the courts. This scenario is subject to a number of contingencies. First, do the

litigants and their lawyers have the legal imagination to devise these

claims? Second, do they have the capacity to organize to sustain such litigation and press for the implementation of favourable rulings? (Of course, the capacity for organization might be enhanced by creative deployment of legal services.) Third, are the judges likely to be respon-

sive to such claims, claims that ask them to depart from deeply held

Constitutional Scope

395

notions about the judicial role? They would not only have to innovate original standards of what is appropriate movement toward equality, they would have to undertake a sustained activist monitoring of government that they may find both ideologically uncongenial and institutionally discomfiting.132 (Were they to shrink from this, while maintaining an interpretation of equality rights as substantive, Thomas would, in the

guise of exalting the Directive Principles, have helped to demote these Fundamental Rights to the non-enforceable status that led to the dis-

dain of the Directive Principles.) Fourth, even if claimants come for-

ward and courts are responsive, do the realities of litigation in India, the

delay and maneuver and cost, make it possible for groups to improve

their position through litigation against powerful adversaries? This is related toa fifth contingency: are the real barriers to improvement of the

conditions of these groups reachable by even most well-disposed and

capable judiciary? To what extent is the notion of remedying these problems through litigation yet a further tempting illusion? We may be

better able to address these questions after we have followed the course

of

the

first

discrimination.

quarter-century

of

litigation

about

compensatory

It is not clear yet in what ways and how deeply Thomas will transform

earlier doctrine and practice. It was pronounced by a divided court; it

expresses the heated symbolic egalitarianism that was both institutionalized and discredited during Emergency Rule. Although it pro-

vides no satisfying answers, its presence permits and requires some fresh thought about the principles underlying compensatory discrimination policies

132. On the accomplishments, problems, and discomforts of the American counterpart of this kind of activist ‘public law” judging, see Chayes 1976; Horowitz 1977 and Galanter, Palen, and Thomas 1979. On its divergence from the Indian judicial style, see chap. 14, §A, below.

The Scope and Extent of Preferences: Constitutional Limits and

Judicial Control A. HOW

MUCH

RESERVATION? THE MINIMUM

ALTHOUGH there is a general constitutional duty to protect and advance the interests of the ‘weaker sections,” the Constitution does not explicitly obligate the government to employ any particular method of doing this. There is no provision in the Constitution that the State need reserve any minimum number of posts in government service seats in educational institutions; nor divert any minimum part of its resources to

benefits for backward groups. Except for reserved seats in the legisla-

tures, the quantum of preferential treatment is unspecified. The courts have been unresponsive to claims attempting to establish some minimum quantum of compensatory discrimination. We saw earlier that courts were unwilling to admit as justiciable claims for inclusion by groups which claimed to be rightfully included within the Backward Classes or Scheduled Castes.1 Similarly, the courts have been unwilling to vindicate the claims of admittedly proper beneficiaries who

complain that the provisions for preference made by the government

are insufficient in quantity and scope. In C. A. Rajendran v. Union of India,? the central government had confined reservations in promotions to Class III and Class IV posts. A Scheduled Castes employee in Class

III claimed that the limitation to inferior posts violated the guarantee

afforded by Article 16(4). He argued that Article 16(4) was not merely an exception engrafted on Article 16, but was itselfa fundamental right,

granted to Scheduled Castes and Tribes, which places on the government the duty of making reservations for these groups. The Supreme

Court repelled this contention with the observation that

Article 16(4) does not confer any right on the petitioner and there is no constitutional duty imposed on the Government to make a reservation for

1. See chap. 8, §1.

2. A.1.R. 1968S. C. 507.

Scope and Extent of Preferences

397

Scheduled Castes and Scheduled Tribes, either at the initial stage of recruit-

ment or at the stage of promotion. In other words, Article 16(4) is an enabling provision and confers a discretionary power on the State to make a reservation. .. .3

Other courts have concurred that Articles 15(4) and 16(4) do not confer on backward groups any Fundamental Right to such arrange-

ments.* These provisions represent an exception to the rights that

others would enjoy in their absence to complain of such arrangements as violative of the Fundamental Rights granted in Articles 15, 16, and 29; they thereby empower the government to confer preferential treat-

ment, but they do not require it todo so. Government may constitutionally omit to make any particular scheme of preferences.5

The courts seem to infer that since there is no duty to confer any particular sort or amount of preferential treatment, there is no duty to confer any at all. In effect, they hold that a discretion sufficiently broad to allow a zero response to any individual claim is taken to imply a

discretion to make a zero response to every claim. But this is somewhat paradoxical in view of the clear and explicit constitutional duty to make some special provision to advance the interests of the weaker sections. Article 46 directs:

The State shall promote with special care the educational and economic interests of the weaker sections of the people, and, in particular, of the Scheduled Castes and Scheduled Tribes, and shall protect them from social

injustice and all forms of exploitation.

While this Directive Principle is “not enforceable in any court

. . . it

shall be the duty of the State to apply [it] . . . in making laws.”’® The

3. C. A. Rajendran v. Union of India, A. 1. R. 1968 S. C. 507 at 513. This is a holding in regard to promotions dictum in regard to initial appointments. In regard to initial recruitment, this point was explicitly decided in Dasa Rayudu v. Andhra Pradesh Public

Service Commission, A. 1. R. 1967 A.P. 353, 363, where the court rejected a petition seeking to require the Andhra

classes.

4. Arts.

15(4) and

Pradesh

Government

to restore reservations for backward

16(4) are enabling provisions which “do

... not impose an

obligation, but merely leave . . . it to the discretion of the appropriate Government to take suitable action, ifnecessary” (Balajiv. State of Mysore, A. 1. R. 1963S. C. 649 at 664). 5. “[I]t must be remembered that Art. 15(4) and Art. 16(4) are not mandatory, but

are merely enabling provisions. The State Government therefore may or may not make any rule or issue any executive direction for the purposes of reserving some posts for extending other concessions to backward classes . . . when the State Government can in its discretion make such rule or issue such administrative instructions, it can also cancel or modify it. It was not incumbent upon the State Government to prepare any list of backward classes, nor is it incumbent upon it now to prepare any list”( Dasa Rayudu v. Andhra Pradesh Public Service Commission, A. 1. R. 1967 A.P. 353 at 363).

6. Art. 37.

398

The Judicial Process

commitment to preferential treatment is even more specific in the employment area. Article 335 states: The claims of the members of the Scheduled Castes and of the Scheduled

Tribes shall be taken into consideration, consistently with the maintenance

of efficiency of administration, in the making of appointments to services and posts in connection with the affairs of the Union or a State.

The statement in C. A. Rajendran that “there is no constitutional duty” seems to involve an unspoken equation between “constitutional duties” and duties ‘enforceable by a court” and a finding that the latter is not present. It is not clear whether the Court is suggesting that the government’s duty is unenforceable in courts under any circumstances or whether the government’s provision here is so comfortably above the constitutional floor that it is clearly under no enforceable duty to do more. This would leave open the situation where government res-

ponse was absent or patently inadequate. Thus, for the time being at least, there is a complete absence of judicial limits of any kind over minimum preferential treatment. When government does confer preferences, the courts have been dis-

inclined to exercise scrutiny over their scope or coverage. The govern-

ment need not include among the beneficiaries every group which might conceivably qualify as backward. It may make reservations in some services and not others, in some grades and not in others. Thus, in

C. A. Rajendran, the discrepancies in the treatment of one stream of

promotion and another, of higher and lower grades, of one service differently from others, were found to be justifiable classifications which did

not offend equal protection. The Court did not require from the govern-

ment any showing of the reasonableness of drawing the lines where it had, finding it ‘‘obvious” that an order limiting reservations to Class

III and IV posts was a reasonable classfication in view of the higher

requirement of efficiency and greater responsibility of upper echelons.” This unwillingness to oversee the coverage of preferences excludes not only beneficiaries complaining of their insufficiency, but also nonbackward-class employees who claim that preferences are unfairly con-

centrated to their disadvantage. In Devadasan v. Union of India, the Supreme Court held that employees who complained of reservation in promotions in the Central Secretariat whereas other agencies had no

such arrangement had no complaint “merely because reservation is not made in every kind of service under the State.””® 7. A.1.R. 1968 S. C.507. 8. Devedasanv. Union of India, A. 1. R. 1964'S. C. 179 at 185.

Scope and Extent of Preferences

399

B. HOW MUCH RESERVATION? THE MAXIMUM The Constitution does not explicitly provide any maximum or limita-

tion on the quantum or extent of preferences. Article 15(4) authorizes “any special provision for the advancement” of the backward; Article 16(4) empowers the State to make “any provision for the reservation of appointments or posts” in favour of any under-represented backward class.9 In spite of this broad and unrestricted language, a doctrine of

constitutional limitation has emerged in dealing with those preferences

which take the form of percentage reservations of government posts and

places in educational institutions.

In Venkataramana v. State of Madras, where the question of extent was

not in issue, the Supreme Court upheld reservations for Harijans and

Backward

Hindus of 19.7% of posts filled by competitive examina-

tion.1° In Kesava Iyengar v. State of Mysore, the first court to consider explicitly the question of extent upheld a reservation of seven out of ten posts.11 The Mysore High Court there conceded that “reservation” in

Article 16(4) signified a “small fraction” or “small portion of the main” (presumably less than one-half). However, on the ground that “each backward class is an independent class whose claim for appointment can be sustained under Article 16(4),”!2 the court suggested that the

state could reserve such a portion for each backward community and could, on this basis, give all the available posts to backward groups. This expansive view, which permits total reservations to exceed 50% and would allow reservation of all posts, is now discredited. In Triloki Nath Tiku v. State of Jammu and Kashmir [11], the Supreme Court rejected the state’s policy of giving 50% of government posts to Kashmir Muslims, 40% to Jammu Hindus, and 10% to Kashmir Hindus,

on the ground that “[i]n effect the State policy ...was a policy not of reservation of some appointments or posts; it was a scheme of distribu-

tion of all the posts communitywise.” Such distribution was contrary

to Article 16(1) and (2) and was not saved by Article 16(4). The chief draftsman of the Constitution, defending Article 16(4) 9. Cf. the limitation in Art. 335 to claims consistent “with the maintenance of

efficiency of administration.” But Art. 335, unlike Art. 16(4), is not limited to “reserva-

tion” as a means of preference. 10. A. I. R. 1951 S.C. 229, The State had specified that of 83 vacancies as munsifis, 12 be filled by members of the Subordinate Judicial Service and 71 from the bar, assistant public prosecutors, etc. on a basis of communal quotas. The quotas originally set for

Harijans and Backward Hindus amounted to 40.8% of appointments from the bar

(36.1% of all posts). The appointments actually made included 14 members of these groups among the 71 appointed from the bar (19.7% of this group, 16.8% of the total).

11. A. LR. 1956 Mys. 20. 13. A. I. R. 1969S. C. 1.

12. Id., at 24. 14, Id.,at 4.

400 The Judicial Process before the Constituent Assembly, indicated that the reservation authorized was “‘a minority ofseats” and gave the example of an aggregate reservation of 70% of posts as falling outside the power bestowed by the clause.15 Although the draftsman visualized judicial review of the extent of reservation, the text of Article 16(4) itself supplies no apparent

warrant for it, unless the word “reservation” is construed necessarily to

imply a minority (less than 50%) of posts—not merely a minority for any single backward group, as the court in the Kesava Iyengar case construed it, but for the aggregate of all backward groups.

In Balaji v. State of Mysore, the Supreme Court has attempted to

supply a constitutional limit to the extent of preference, not on narrow technical

construction

of “reservation,” which was not available to

it—-since the reservation was under Article 15(4) rather than Article

16(4)—but on broader grounds of policy.16 The Mysore scheme reserved a total of 68% of places in engineering and medical colleges: 15% for Scheduled Castes (SC); 3% for Scheduled Tribes (ST); and

50% for Backward Classes. Proceeding from the notion that Articles 15(4) and 16(4) are provisions of a special and exceptional character,

rather than principles that can be given general operation, the Court finds that these provisions should not be read as completely excluding or ignoring the fundamental rights of all citizens. It characterized the purpose of these special provisions not merely as conferring special privileges on the backward, but as serving the interests of the whole society by promoting the advancement of its weakest elements. Thus, 15. Dr. B. R. Ambedkar, VII CAD 701. 16. A. I. R. 1963 S. C. 649. So long as the argument for the percentage limitation

rested on the word “reservation” in Art.

16(4), it presumably would not apply to

benefits under Art. 15 (4)—including “reservations” in educational institutions. Such a distinction was apparently inferred by the chairman of the Backward Classes Commission. Contrast his recommendation of maximum 49% reservation of government posts

with this recommendation ofa 70% reservation in existing educational institutions (I BCC ix, xii). Cf. Ramakrishna Stagh v. State of Mysore, A. 1. R. 1960 Mys. 338, where the High ‘Court struck down a scheme of reservation in profession! colleges by which Scheduled Castes and Tribes enjoyed a 20% reservation and another 45% was distributed among various groups designated as Backward Classes. While subjecting this scheme to withcring criticism. the court did not see fit reservations even though the total reservation comprised Unlike the Kesava Iyengar and Venkataramana cases, above, this case involved a scheme of benefits under Art. 15(4), government posts under Art. 16(4). Were such a distinction

to object to the extent of well over half of the seats. nn. 11 and 10 respectively, rather than reservation of recognized the state would

have been able to reserve only 50% of the government posts for members of a preferred group but able to reserve for them a larger percentage of the educational opportunities

necessary to qualify for these posts. A percentage limitation applicable to reservation of

seats in educational institutions and on local political bodies had to rest on some other ground than interpretation of “reservation.”

Scope and Extent of Preferences

401

any scheme for preference can be weighed in terms of the interests of the whole society. In considering reservations in institutions of higher

education, it is necessary, then, to balance the policy of advancing the

backward against the national interest in the full utilization of talent. Lowering of the quality of graduates, says the Court, is the “inevitable

uence of reservation.”17 But the need for technical, scientific,

and academic personnel “‘s so great that it would cause grave prejudice to national interest if considerations of merit are completely excluded

by wholesale reservation of seats in all technical, medical, or engineer-

ing colleges or institutions of that kind.”18 Thus, special provisions for the backward must be within “reasonable limits”: “the interests of the weaker sections of society which are a first charge on the States and the

Centre have to be adjusted with the interests of the community as a

whole.”9 Although recognizing the difficulty of adjusting these competing claims, the Court experiences no difficulty in finding that reservation of “practically all the seats available in all the colleges . . . clearly would be subverting the object of Article 15(4).”2° While reluctant to set

definite limits, the Court indicates that “speaking generally and in a broad way, a special provision should be less than 50%; how much less than 50% would depend upon the relevant prevailing circumstances in each case.”’22 The reservation of 68% here it finds plainly inconsistent with Article 15(4). This clear inconsistency with the concept of the special provision, combined with sole reliance on caste as a criterion for selecting these classes, was such a covert or latent transgression of the

constitutional power as to amount to a “fraud on the Constitution.”22 How much less than 50% need the reservation be? The Court mentions in passing the recommendation of the central government, in-

corporating the recommendations of the All-India Council for Techni-

cal Education, that the total reservation (for SC and ST as well as Other Backward Classes) may be up to 25%, with marginal adjustments not exceeding 10% in some exceptional cases. The central government had specifically suggested to the Mysore Government that in all nongovernmental institutions the reservations should not in any case exceed 35%.33 In discussing reasonableness, the Court mentions that the University Education Commission observed that reservations should not exceed a third of the total number of seats.?4 Reservations of 25% or 35% are identified as acceptable, and reservations over 50% as clearly unreasonable, leaving an area of doubt between 35% and 50%. Of course, even a reservation in the acceptable range would not auto17. AT. R. 1963 S.C. at 662. 21. dd. 22. Id., at 664.

18. Id. 19. Id., at 663. 20. Id. 23. Id., at 656. 24. Id., at 662.

402

The Judicial Process

matically be “reasonable”; it might still be attacked as unreasonable under the circumstances. The limits suggested by the Court include the

share reserved for SC and ST. Since it seems to be generally accepted that these groups may be given reservations in the range of their ratio to the total population (about 15% and 6% respectively on an all-India basis), this would seem to leave a very limited scope for reservations in

favor of Other Backward Classes. Although

Balaji directs attention

to the lower percentages

that

might, under the circumstances, be reasonable, the state governments

and the high courts have ignored the qualifications and have responded by taking the 50% that Balaji mentioned as the farthest permissible limit as a flat maximum. The first High Court to apply the Balaji views

on the permissible extent of reservations upheld a reservation of 30% of places in engineering and pre-medical colleges for Backward Classes.?5

This reservation was in addition to reservations of 15% for SC and 3%

for ST. The Mysore High Court there found that the reservation of 30% for Backward Classes was not “excessive.” Backward Classes were

defined in terms of income and occupation; the state contended that, as

so defined,

they are “very much

more

than 50%

of the [state’s]

population.”26 The High Court accepted the 50% limit mentioned in Balaji as an absolute maximum and took no notice of the lower estimates of reasonableness mentioned in Balaji.27 Seven years later the Supreme Court found Mysore’s reservation of 48% of medical school places for backward classes (including SC and ST) “apparently . . .on

the high side” but not shown to be “unreasonably excessive.”8

Similarly, the Kerala High Court later upheld a scheme reserving

50% of munsif’s posts, 10% for SC and ST and 40% for various Backward Classes. The court noted that 50% was the upper limit but that within this limit “‘it is for the State Government

to decide.”29 Thus,

while Balaji has brought about the end of those expansive schemes of

reservation which transfromed the whole competition into a set of com-

munal quotas, it has not had as restrictive an effect on the extent of reservation as might have been expected from an examination of the 25. D. G. Vishwanath v. Government of Mysore, A. 1. R. 1964 Mys. 132. 26. Id., at 137. 27. The Supreme Court, upholding the scheme against challenge on other grounds,

did not comment on the quantum of reservations (Chitralekha v. State of Mysore, A. 1. R. 1964 S. C. 1823). Mudholkar, J., dissenting, felt that the extent of the reservations was

‘one of several factors which reduced the consideration of merit to an impermissible level.

28. Chanchala v. State of Mysore, A. 1. R. 1971 S.C. 1762 at 1770.

29. Hariharan Pillai v. State of Kerala, A. I. R. 1968 Ker. 42 at 51.

Scope and Extent of Preferences

403

decision. The qualifications have fallen away, leaving a flat 50% rule.3° In the Balaji case, the extent of the reservation was only one of several

objectionable features, and the Court indicated that its remarks about a

50% limit were not to be taken as a precise formula. However, less than a year later another bench of the Supreme Court?! carried the limita-

tion on reservations much further and made it more rigid. In Devadasan v. Union of India, the central government provided reservations of 12%2% for SC and 5% for ST for promotions from Grade IV to Grade III posts in the Central Secretariat Service to be filled by competitive examination. The government applied a “carry forward rule” under which unfilled reservations from the previous two years were added to the reservation for the current year. A total of 45 appointments were made, of which 29 (64.4%) were from the SC and ST.3? Finding that the carry-forward rule permitted reservations of over 50% (in the third year), the Supreme Court, one judge dissenting, found the scheme unconstitutional. The majority found that the holding of the Balaji case was that any reservation greater than 50% was unconstitutional.33 But their application of this “holding” to a very different set of facts greatly curtailed the government’s power to make reservations and its discre-

tion in implementing them. Article 16(4), says the Devadasan majority, is ‘a proviso or an excep-

tion” to Article 16(1) and cannot be interpreted to nullify or destroy the

main provision. “To hold that unlimited reservation could be made

under cl. (4) would in effect efface the guarantee contained in cl. (1) or at best make it illusory.”34 Article 16(4), then, must be read to avoid

obliterating the rights granted in the rest of the article. Thus, “the

overriding effect of cl. (4) on cls. (1) and (2) could only extend to the 30. As one judge baldly states it: ‘It is settled law that 50 percent posts have to be thrown open to the general public” (Daya Ram v. State of Haryana, A. I. R. 1974 P. & H. 279). However, this requires a qualification, for distribution restrictions may reduce the open competition portion further. See below at §D of this chapter. 31. It should be noted that although Devadasan was decided only eleven months after

Balaji, the five judges sitting in the former included none of the five who had sat in the latter. Sixteen months later, in B. N. Tewariv. Union of India, A. 1. R. 1965S. C. 1430, the

Supreme Court indicated that the Devadasan decision did not affect the current reservations for Scheduled Tribes or Scheduled Castes, nor the validity of the provision for “carry-over” of reserved seats from one to another of these groups. The bench in the latter case included three members of the Balaji court and a single member of the

Devadasan court.

32. Apparently, the three years’ reservations of 1742% of current appointments could

add up to a reservation of 64.4%, since there were fewer mew vacancies to be filled during the third year. In fact, the government had reserved a total of 32 out of 48 vacancies (66.6%), but it filled only 29 of the 32 (A. 1. R. 1964S. C. 179 at 180).

33. Id., at 186.

34. A. IR. 1964S.C. 179 at 187.

404

The Judicial Process

making ofa reasonable number of reservation [sic] of appointments and

posts in certain circumstances.””35

A “reasonable” number is one which strikes a “reasonable balance”

between the claims of the backward classes and those of other citizens. The notion of a “reasonable balance” between conflicting interests

echos the Balaji case, were the Court contemplated that “steps for the advancement of the weaker elements” must be balanced with “the requirements of the community at large.”36 However, the notion of reasonableness in Devadasan is both novel and exceedingly restrictive. To the Devadasan majority the purpose of reservations under Article 16(4) is to insure that the backward “are not unduly handicapped in securing government employment.”’37 Where a rule makes reasonable allowance for the backwardness of members of a class ... what the state would in fact be doing would provide the members of the backward classes with an opportunity equal to that of the members of the more advanced classes in the matter of appointment to public services. If the reservation is so excessive that it practically denies a reasonable opportunity for employment to members of other communities .. it would be open ... for a member ofa more advanced class to complain that he had been denied equality by the State.3®

Thus reservations are legitimate to the extent that they provide back-

ward classes with an “opportunity equal to that of members of the more advanced classes”; apparently they are excessive if they do more than provide equal opportunity. But such a formula raises some puzzling questions. What does equal opportunity mean in such a setting? Let us assume that in a service of 100 employees there are 20 members of

Backward Group X. Let us further assume that the X population con-

tains the same distribution of native talent, effort, and incentive as the population as a whole, but that it differs in background, education, and

training to an extent that in a test of objective performance the Xs will score substantially lower than the rest of the population. Let us further assume that if there are 10 promotions available in this service and “equal opportunity” to compete for them, that none of the 20 Xs would qualify in open ‘‘merit” competition. If a scheme of preference or “compensatory discrimination” is introduced into such a situation, it cannot be with the purpose of giving the Xs equal opportunity to compete, for this they enjoy already, but for the 35. Id.

36. A. I. R. 1963 S. C. at 663. Cf. 662. 1 am unable to find anywhere in the text of the Balaji decision any reference to ‘‘claims of other employees” such as is adverted to by the

Devadasan majority at A. I. R. 1964S. C. 187. 37. A. 1. R. 1964S. C. at 183.

38. Id., at 185.

Scope and Extent of Preferences

405

purpose of giving them greater opportunity to achieve success in the competition. The scheme is justified precisely on the ground that the

deficiencies of the Xs in background and education make them unable

to compete on terms of equality. If they have merely equal opportunity to compete for all posts, they will have less than an equal probability of succeeding. A scheme of reservations, then, is designed to increase their probability of success and it does so by curtailing the opportunity of others to succeed— by denying these others the opportunity to compete for some portion of the available places. What the majority in Devadasan seems to suggest is that the maximum

permissible reservation for the Xs is that under which both they and the

others would enjoy “equal opportunity.”’ But equal here cannot mean equal opportunity to enter the competition, for no reservation is needed to accomplish that and no reservation is compatible with that. Does it mean equal probability of success? Let us assume that all of our 20 Xs and 80 others compete for the promotions. If5 of the 10 higher posts are

reserved for the Xs, it is clear that the statistical probability ofa given X

attaining one of the higher posts is greater (25%) than the chances of one of the others (6.25%). Does the fact that the chanes of success are, at least theoretically, far higher for the Xs make their opportunity unequal? Must these opportunities be equalized by confining the reservation to one in which the successful candidates are likely to stand in the same proportion to those who compete? This would limit the reservation to the percentage of Xs already eligible to compete— in this case 20%, or 2 of the posts— which would give both the Xs and the others a

10% statistical chance of success. But since the probability of succeed-

ing also depends on ability to obtain the minimum qualifying marks, and this does not necessarily bear any relation to the extent of reservation, such equalizing of statistical probabilities would be a barren exercise with no relation to the actual chances of the various competitors.3? But even if it were possible to do so, it is clear that preferences are 39. ‘The discrepancy between reserved percentages and posts obtained nearly always is to the disadvantage of the backward groups. For example, in the year ending 31 Mar. 1962 appointments to the Grade III of the Central Secretariat Service (Permanent Strength) were as follows:-

By direct

recruitment By Promotion Total

Scheduled Castes.

Scheduled Tribes

= Others

%of Scheduled Castes & Tribes

0 10

0 1

18 496

0 2.17

10

1

Sit

2.09

These are the same posts involved in the Devadasan case. Figures from RCSCST 1961-62: 135.

406

The Judicial Process

not intended to “equalize” opportunities in this fashion. The arrange-

ment of reservations as fixed quotas, by making them in effect separate

competitions, has been definitively rejected by the courts.4° Another meaning of “equal opportunity” might be the notion that the reservation should be of such a size that equal “effort” should be rewarded by equal success, in spite of lower performance due to deficiencies of background and training. This kind of intuitive fairness is probably closest to the notion of equal opportunity. But although such fairness may be one factor taken into account in deciding on the extent of preferences, it is patently impossible for either the government or the courts to administer any rule that involves the measurement of effort or

other subjective factors which make some as deserving as others. The Court adds further complexity to the meaning of equality by noting that “the reservation for backward communities should not be so excessive as to create a monopoly or unduly disturb the legitimate claims of other communities.”41 Other communities, of course, have no claims as communities, nor do their members have any constitu-

tional claims by virtue of their membership in such communities. Even the backward classes have no claims as communities. Except in the case of political safeguards, members of backward classes are entitled to special treatment as individuals; their membership in certain communities is only used to identify them as deserving beneficiaries. The

right to equality and freedom from discrimination is an individual one.*2 The only “claim” that members of other communities have is a right to compete as individuals in open merit competition; they have no “claim” for “reasonable” representation for members of their respec-

tive communities.44 The backward classes enjoy the same right to compete in open merit competiton and in addition a right— where it is

provided by the State—to a selection biased in their favor. Neither of these competing “claims” vests in any communal group; the rights of individual members of backward and advanced groups cannot be affected by the number of others who succeed or fail. 40. Dippala Sura Dora v. V. V. Giri, A. 1. R. 1958 A.P. 724; Raghuramulu v. State of Andhra

Pradesh, A. 1. R. 1958 A.P. 129;S. A. Partha v. State of Mysore, A. 1. R. 1961 Mys. 220 at 235; Shimoga Durga Bovi v. Returning Officer, 1962 Doabia’s Election Cases 31 (Mysore

High Court, 1960); Digambar Rao Bindu v. Dev Rao Kamble, 15 E.L.R. 187 (Bombay High

Court, 1958). See discussion in chap. 13, §B.

41. A.1.R. 1964S. C. at 187.

42. Venkataramana v. State of Madras, A. 1. R. 1951 S. C. 229; State of Madras v. Champakam Dorairajan, A. 1. R. 1951 S.C. 226. 43. Cf. V. V. Giri v. D. Sura Dora, A. 1. R. 1959 S. C. 1318, where a combination of

reservation and “merit” gave both seats in a double-member constituency to Scheduled Tribe candidates. The Supreme Court upheld this result by construing the reservation

of one seat as a guaranteed minimum, not a compartment. Obviously, a concern for the “legitimate claims of other communities” did not enter into the Court’s calculations.

Scope and Extent of Preferences

407

How then are the right of all to open merit competition and the right of some to a biased or weighted selection to be reconciled? There is no mechanical rule that can achieve such a balance. Presumably, what the Court is saying is that in striking the balance there must be enough unreserved places to make the opportunity of non-backward individuals to gain a place on merit a substantial one. But it is clear that how substantial this right is for individuals bears no necessary relation to the percentage of places reserved; it depends at least as much on the number of places available, the absolute number of competitors, and the absolute number of places reserved. And, of course, it differs from

one individual to another depending on his own ability, training, and preparation and that of his rivals. There is no mechanical formula by which it can be determined that this opportunity is “equal” between different individuals. Thus in reconciling these conflicting commitments more factors than mere “equality” must be taken into consideration. These might include judgments about the morale and incentive of employees, the full utilization of talent and training, the efficiency of the service, the intuitive fairness of having equal-effort rewarded by equal

success, the intangible effects of reservation on the backward classes in

terms of their economic well-being, incentive to education, feelings of

self-esteem, and political effectiveness, and the effect of their presence

on government dealings with the disadvantaged. Striking a balance among these competing and elusive interests of the backward classes, the other employees and the public at large is a judgment of policy. The Constitution has not specified any formula for striking it. What impels the Court to deem this a case requiring judicial overturning of the balance struck by the government when it applied the “carry-forward

rule”?

Once a percentage limitation on reservations is inferred from the

Constitution, some peculiar problems arise when a time dimension is added. Presumably, such a limit applies to the total number of posts or seats reserved. But, typically, the percentage of backward classes among present incumbents of these posts is much lower than the reserved percentage. Does the limit also apply to the permissible reservation among appointments currently being made? For example, if a state reserved 30% of Grade I posts for members of group X, may 60%

of 100% or current appointments be reserved for the Xs until the 30% is reached? Temporary reservations of up to 100% have been employed by some states.“ Again, can the state place its 30% reservation for the Xs on a long-term basis, so that a paucity of X candidates in earlier competitions can be redressed in later ones? This is, in the abstract, the

problem of the ‘“‘carry-forward rule.” 44, See RCSCST

1953: 134; RCSCST

1954: 171; RCSCST 1961-62: 139:

408

The Judicial Process

Justice ‘Subba Rao, dissenting in the: Devadasdn case, suggests that the reasonableness of a reservation is to be measured by its relation to.total cadre. strength: a ‘reservation ‘is constitutional so long. as‘ the “privileged” groups do not make upa disproportionate part of the total complement of employees. For -bim the words “any provision” in Article 16(4) are. brodd enough to cover schemes of reservation which operate over several years and.in'relation to the eritire strength of the

cadre;

reasonableness

percentage of seats use any method, prescribed level is: equality in Article

is not to be determined. on the basis: of :the

resérved on one particular occasion. The state may including reservation of 100% of posts until the reached.45 But for the majority -‘the guarantee of 16(1) :

means that on every occaison for recruitment the State should see that all

citizens are treated equally. .. . [Elvery citizen who is seeking employment , . . under the State is entitled to. . . an opportunity for seeking such employment

. .

whenever it is intended to be filled. Ln order to effectuate

this guarantee each year of recruitment will have to be considered by itself and the reservation for backward communities should not be so excessive as to disturb the legitimate claims of other communities.46 _

Itis “each would cies*7

not clear here whether the majority means ‘“‘every occasion” or year.” They are obviously not the same. The. former standard rule out such devices as rosters of rotation for recurrent vacanand special examinations open only to:the members of the

backward ¢lass,4* even if designed only to fill the reservation ‘for the

given year. Subsequent courts seem to have fastened on the “every year” standard and to have ignored the “every occasion” language. But whichever it is, the decision seems to make mandatory a temporal 45 A.1.R. 1964S..C. at 192,

46. Id., at 187.

47. Prescription of rotational rosters (in which sequential vacancies are arranged i in cycles of, e.g., 20 and certain ones in the cycle are specified as reserved for certain groups) is a common dévice for implementing reservations in hiring. For a particularly elaborate example, sée Government of Kerala 1960: 66. Rotational systems (apart from backward classes matters) were held not to violate Art. 16(1) (Kelkar v, Chief Controller of Import and Exports, A. 1. R. 1967 S. C, 839). Recently the Supreme Court has described

the “rotational or roster system [as] a common adopted and easily understood method”

of administering entry quotas, but held that it was not the only atlowable method

(Chauhan v. State of Gujarat, AVR. 19778. C. 251, 261). 48. Cf. the special examination in the Central Secretariat Stenographer’s Service in

Dec., 1960 (RCSCST 1961-62: 135). Devadasan involved promotions. If the same rule

were to apply to original recruitment, presumably special examinations to fill vacancies for Scheduled Castes and Scheduled Tribes could not be held—unless government servants have interests of greater weight than those attempting to enter service. 49. Thus in C. A. Rajendran v, Union of India, A. 1. 8. 19688. C. 507 at 513, the Supreme Court reads Devadasan as saying that each year must be considered by itself.

:

Scope and Extent of Preferences 499 segmentation of reservation. Each year’s reservation is to be judged reasonable or unreasonable by itself, regardless of. the extent to which

reservations have gone unfilled in the past. The majority does not

specify that “carry-forward” provisions are invalid only when they raise a current reservation over 50%. Presumably, such a rule would be impermissible whenever ‘its operation deprived other employees of opportunities equal to those of the backward classes.51 It appears, too, that this equality must prevail in every separate category of competition within the year’s recruitment.S? For in this case, posts to be filled by competitive examination constituted only

30% of the appointments for the year. Another 40% were to be filled by

direct recruitment from those who had obtained lower ranks on the LAS.,

etc., examinations, and 30%

were to be filled by promotions

from Grade IV on the basis of “seniority-cum-fitness.” Even if we

assume that these other methods of selection were accompanied by the

reservation applicable to “direct recruitment” (16 2/3% for SC and 5%

for ST), the actual percentage of posts reserved during the year may have been only 37.56% (assuming there were no carry-forward provisions applied to those categories of recruitment). Actually, the Court does not consider any of the recruitment done under the other headings 50. But cf. General Manager v. Rangachan, A.1.R. 1962 S.C. 36, at 39/45. The order there providing reservations in promotions was issued in 1959 to be effective from Jan. 4, 1957.The Court there held that the reservation could operate retrospectively as well as

prospectively. Presumably, then, it is permissible for the State to cumulate reservations in retrospect but not to do it prospectively by a “‘carry-forward rule.”

51. The majority does not indicate whether it is still open to the government to leave

unfilled the reserved places which are not filled by backward classes in a given year. There is nothing in the majority opinion to indicate that other employees have any constitutional right to compete for the reserved places which cannot presently be filled with qualified members of the backward classes. By making no appointments at all for these places, government could increase the number of vacant appointments—and the number of reserved places—at the next competition when there might be more

candidates who could qualify for the reserved places. Since there would seem to be no

constitutional objection to such a practice, government would be able to achieve at least

part of the object of the “carry-forward” rule. However it could only do so by sacrificing

the value of a full roster of employees during the interim. What the majority decision does is to remove from government the option of both filling its roster now from the

available candidates and preserving the reservation until qualified members of the

backward classes are available. 52. Thus government may not club together otherwise independent cadres of posts solely for purposes of administering a reservation where that unduly impinges on the chances of non-backward employees. When grouping promotions of Railway Mail

Service Inspectors and Postal Inspectors led to two of three vacancies in the former going to Scheduled Castes, it was “excessive reservation” which “unreasonably denied

this opportunity” toa high ranking non-backward employce (Director General of Posts and

Telegraphs v. Natarajan, 1971 [2] M. L. J. 79, 80, aff'g A. I. R. 1970 Mad. 458).

410

The Judicial Process

during the same year, so we have no indication of the precise extent of

the effective reservation of posts during that year.53 In order to appreciate the extent to which the Devadasan majority

departs | from the Balaji case and extends judicial policy-making in this

area, it is helpful to consider some of the salient differences between the

two cases. First, the factual situation in colleges, where there is a

complete turnover of students every 2 or 3 years, is quite different from that of a government service where the incumbents stay from 20 to 30 years. In schools a smaller reservation may suffice to bring the proportion of the backward in the school population up to the reserved strength in a few years; in the service such a process would take a long time. For example, under a reservation of 172% for a service where the average

tenure of employment is 30 years, it would take 30 years before the reserved percentage was reached (assuming that the full number of backward class members succeeded every year).54

Second, the higher percentage of reservation created by the operation

of the carry-forward rule is quite different from a straight reservation ofa similar percentage. For the carry-forward rule depends upon the fact that others had increased success in the past; the chances of success for

the backward this year are increased commensurately with their lack of success in previous years. Again, the carry-forward provision is selfliquidating and non-recurring; to the extent that the higher reservation is effective, the reservation returns to its original lower level. Thus arguments based on the public interest in utilizing talent and insuring

the efficiency of the services have less application to the carry-forward

rule than they do toa situation like that in Balaji, for over the long run the reservation is not increased above an acceptable level.

The whole policy of preferences introduces a factor of historical

balance or compensation;55 the backward are to be given a more than

equal chance this year in order to overcome the deprivations and inequalities of the past. But according to the majority, the government,

53. Calculation based on data at A. I. R. 1964S. C. at 182 fF. 54. The inordinate length of time required for such reservation to become effective

inspired Wanchoo, J., dissenting in General Manager v. Rangachani, A.1.R. 1962S. C. 36 at

46-48, to propose that Art. 16(4) allowed two kinds of reservation: reservation of a

percentage of current appointments and also reservation of a certain portion of “posts”

in the whole service. Under the second type all appointments could be reserved to the backward

classes

until

the

reservation

was

filled, so long as candidates

met

the

minimum qualifications. This was based on a textual argument about the meaning of the term “‘posts” in Art. 16(4), which was rejected by the majority there. The Devadasan majority reads Balaji as forbidding even a temporary reservation about 50% (A. I. R.

1964 S. C. at 188).

55. Cf. Gajendragadkar, J., for the majority in General Manager v. Rangachari, A. 1. R.

1962 S. C. 36 at 42.

Scope and Extent of Preferences

411

which is empowered to depart from equality with a view to this long

historical time-perspective, is not authorized to devise its measures with

an eye to results in the two previous years.

Third, the Devadasan situation concerns a scheme of reservation by

the central government, rather than that of a state government as in

Balaji. In Balaji the Court had national policy available to it as a guide

in reconciling competing interests in interpreting Article 15(4).5¢ But no such guide is available to the Court when it undertakes to sit in judgment over the policy formulated by an agency directly responsive

to the highest levels of national policy-making. Yet the majority confines the discretion of the central authorities more strictly than Balaji

did the state’s discretion.

Fourth, connected with this is the fact that the beneficiaries in the

Devadasan case were Scheduled Castes and Tribes, who are specified under special constitutional authority and to whom there is a firm

constitutional commitment of preferential treatment. In the Balaji case,

on the other hand, the beneficiaries were backward classes created ad hoc

by the state.5” Fifth, in considering the various interests which must be balanced in determining the reasonableness of, the reservation, the Devadasan case takes a narrower view of the interests involved. On the one hand are the

public interest in efficient operation of the services and the “legitimate

claims” of the other employees; on the other, the right of the backward to adequate representation in the services. But as the Court in Balaji pointed out, special treatment is not solely for the purpose of benefitting

the backward in the short run; there is a long run interest of the society

at large in promoting the advancement of the backward elements in the population.5* The claims of the backward classes are not necessarily in derogation of the interests of the wider public; they may be the instru-

ment for forwarding the public interest in an egalitarian society.

Finally, the Devadasan majority takes a much more assertive view of the Court’s role in sitting in review of the government’s determination

of policy. In General Manager v. Rangachari, the Supreme Court had 56. See A. I. R. 1963 S. C. at 662.

57. It should be noted that Devadasan was the first case in which any court has struck down as unconstitutional a central scheme for preferential treatment. It was also the first time since the addition of Art. 15(4) to the Constitution in 1951 that any court has held unconstitutional any arrangement, central or state, for the benefit of the Scheduled

Castes or Tribes. All the previous cases in which arrangements had been held unconstitutional had dealt with state schemes for the other backward classes. 58. A. I. R. 1963 S. C. at 662. For an eloquent rejection of the facile equation of the “merit” claims of the advanced with the national interest, see Vishwanath v. Government of Mysore, A. 1. R. 1964 Mys. 132 at 136. Cf. General Manager v. Rangachari, A. 1. R. 1962

S.C. 36 at 42.

412

The Judicial Process

declined to undertake any supervision of the government’s policy in making the reservation retroactive. Conceding that it might be wise not to give it retrospective effect, the Court noted that “with the propriety or the wisdom of the policy underlying the circulars we are not directly concerned”’.59 In exercising its power the State “‘must attempt . . . to

strike a reasonable balance between the claims of backward classes and

the claims of other employees as well as the important consideration of the efficiency of administration.” The striking of the balance is left to the government with the Court merely admonishing it that such factors must be considered in every case.

In Balaji v. State of Mysore, the Court speaks of “adjusting” the

““nterests of the weaker sections of society” to “the interests of the

community as a whole,” and declares that “a formula must be evolved

which would strike a reasonable balance between the several relevant

considerations.”61 But it is clear that it is the government which must do the required balancing and adjusting. All the Court is asserting is its right to step in when the government’s final ‘‘adjustment” is “plainly

inconsistent” with the constitutional authorization for preference.® It seems clear that both the Balaji and Rangachari cases concede that the balance between. contlicting interests is to be struck by the State in the

first instance. There is a recognition that there is a wide range of possible arrangements that might be considered reasonable adjustments of the complex factors that must be taken into account. The work of the courts is clearly confined to setting constitutional boundaries by insuring that government has in fact taken the relevant interests into account, but stops far short of substituting their judgment of how they

should be reconciled. But in Devadasan, the Court seems to go much further in addressing

itself to the task of striking a reasonable balance of conflicting interests. It not only sets constitutional boundaries of reasonableness but goes on to specify in some detail what reason requires. The difference is particularly striking because in Balaji the Court concerned itself not with detailed balancing of Article 15(4) and 15(1) but with broad consideration of interests to determine whether the state’s action was justified under Article 15(4).6 In determining this, it had an abundance of concrete data about the state’s selection procedure for backward classes, and the 68% reservation was of such magnitude and obvious impact that no delicate consideration of marginal utilities was required. However, Devadasan involves the government’s decision for arranging over time an admittedly modest and proper reservation: The Court had

before it no concrete data about the basis upon which government 59. Id., at 433. 60. Id. 61.4.1. R. 1963 8. C. at 663.

62. Id.

63. Cf. id., at 657, 659.

Scope and Extent of Preferences

413

decided on this scheme or how it had worked in practice. In effect,

then, Devadasan represents 2 shift from a “rule of reasonableness”: under

which state action is tested:to see whéther the balancing of interests by the state falls within broad constitutional limits, ‘to a “balance. of

interests” rule by which the Court undertakes directly to determine the constitutionality of the government’s action by weighing and balancing

competing claims.

The readiness of the Devadasan majority to step in to oversee the government’s balancing of interests in such a marginal arid ambiguous matter raises the question of the role of the courts in reviewing government action under Articles 15(4) and 16(4). The dissenting judge in

Devadasan adoptsa line of severe judicial restraint. To him, Article 16(4) “has not really carved out an exception, but has preserved a power

untrammeled

by the other provisions . . .of Article. [16].”6° What

provisions are.to be made in exercising this power is, for him, entirely

within the discretion of the State, so long as they fall under the mandate

of that clause. Judicial review would be confined to seeing whether government action properly fell within:that mandate—i.e., whether the arrangement-was in fact “a reservation of appointments or posts” arid whether infact the beneficiaries were a backward class of citizens.

. There is some support in ‘the text of the Constitution for-such a

restrained view of the judicial function in these matters.® It is dear that the-Constitution puts the primary responsibility with- government for Setting policy in this field. AH the cases before Devadasan qualify’ this view only to indicate that they will consider broad lines of policy as laid down in the Constitution in: determining whether the State’s action is.a valid exercise of these powers: The courts find themselves in the difficult position of needing sufficient power to check abuses and see that consti-

tutional principles are protected while leaving to the government suffi-

cient flexibility to implement effectively the policy of: preferences. The Devadasan opinion. was the first exercise‘of judicial review to-involve serious curtailment of the government’s discretion. |:

64: Again, the Balaji court could find some constitutional guidelines in the fact that the state had utilized constitutionally forbidden criteria. Even if the presumption of reasonableness which ordinarily accompanies governmental classifications was not entirely dissipated by the flagrant use of criteria forbidden by Art. 15(1), it added a concrete element of constitutional disreputability which was absent in Devadasan.

65: A. I. R. 1964S. C. at 190.

66..Cf. Art. 320(4), which exempts implementation of Arts. 16(4) and 335 from even the necessity of consultation with Public Service Commissions. While the absence of this ordinary guarantee might be thought to impose on the courts a special responsibility to check abuses, it is difficult to imagine the Constituion makers excluding the Public Service Commissions and simultaneously anticipating a detailed power of review in the courts.

414

The Judicial Process

The paradoxes of expansive judicial intervention combined with a

narrow view of the interests involved are dramatically displayed in

Abdul Latif v. State of Bihar.6? In 1952 the Bihar Government had decided to give preference to members of Scheduled Castes and Scheduled Tribes in the settlement of new excise shops. In 1958, finding that the implementation of this policy was “not .. . very effective,” the government decided to take further measures. It gave special publicity

about proposed settlements to ST and SC. It stipulated that if there was

a single suitable candidate from one of these groups, he should have the shop. If more than one suitable candidate from these groups applied,

the shop was to be given to one of them chosen by lot.

In 1963 an excise shop was to be settled in Ranchi and the 39 applicants included 7 from the SC and ST. The Superintendent of Excise settled the shop by drawing lots among the 7. Abdul Latif, who had petitioned to be included in the drawing, obtained a writ from the

High Court quashing the Superintendent’s order. The Patna High

Court found that an arrangement which “completely excludes the rest of the society . . . is clearly outside the scope of Article 15(4).”® Citing

Balaji on the need for reasonable limits on reservations and for adjusting

the interests of the weaker sections with those of the whole society, the court found that the order violated Article 15(1). The vice of the arrangement was that no others were allowed to compete. But of course

this is precisely what every reservation does— it sets aside some portion of the available prizes and says that members of the non-preferred

groups may not compete for them. The question, then, is whether the

portion set aside is within the limits permitted by the Constitution. In order to answer this question, we must know what is the relevant unit or group of prizes of which we are measuring a portion. If we assume, as did the Patna High Court, that each single bhang shop is a relevant unit, then the portion reserved is 100%, which is clearly out-

side the limit suggested by Balaji’s case. If, on the other hand, we are willing to take an overall or long-run view of excise shops as a whole, then this appears as a smaller reservation. Balaji tells us that reserva-

tions must be reasonable. It is clear that wherever, as in Abdul Latif, selection is in small increments, an acceptable reservation can be devised only if we take the long-run view of the whole class of separate competitions. But if Devadasan is correct in saying that the reservation in each separate competition must be reasonable, then it is impossible to have a reservation here or in anything that is distributed in small

67. A.LR. 1964 Pat. 393. (Decided 27 March, 1964. Although Devadasan, which was decided 29 August 1963, was already reported, the Patna Court took no notice oft.) 68. Id., at 395.

Scope and Extent of Preferences

415

increments—e.g., it would be unconstitutional to provide a roster for reservation of every tenth future vacancy, since that vacancy would be 100% reserved when it came up. Government presumably has a mandate to provide that some reasonable number of its licensees are from the Scheduled Castes and Sche-

duled Tribes. Since selection of licensees is done one at a time, a

percentage reservation is of no utility. Must government reorganize its system of licensing so that shops are settled in groups, which would allow for a percentage reservation? Or is there another way to achieve the desired reservation? Might not the Abdul Latif situation be thought of as involving two stages of reservation— (1) a reservation ofa certain percentage of liquor shop licenses and (2) a further reservation of opportunities to enter into the competition to acquire them? If we think of opportunities to enter the competition as analogous to government posts or places in colleges, then government might require that a certain number of places in the competition or drawing for liquor shops be reserved for Scheduled Castes and Tribes. Balaji by itself would seem to require only that the reservation of chances to enter the competition be a reasonable one. But what does reasonable mean in this context? DevadaSan suggests that it means substantially equal opportunity for the non-backward in each selection—i.e., that a number of non-backward

Ys be allowed in the competition so that their chances are substantially equal to that of the preferred Xs. The fallacy here is that the chances of each individual competitor in a drawing are equal; his chance is affected only by the absolute number of others in the competition and not at‘all by the relative number of Xs and Ys. As this example points out, the equal-opportunity test suggested by Devadasan does not apply to individuals at all, but applies only to the collective chances of the Ys. But

while the Constitution empowers the government to be concerned with

the percentage of Xs who enjoy such benefits, it does not give to individual Ys any justiciable rights in the collective chances of the Ys. Abdul Latif presents in miniature the crucial problems of policy involved in the tension between “‘compensatory discrimination” on the one hand and “equal opportunity” on the other hand. The first suggests reservation, which in the nature of the situation must be 100%; the

second points to the desirability of avoiding 100% reservations, even temporarily, because of their upsetting and discouraging effects. The choice is unavoidable. But is it to be made by the courts or by the government?® In Balaji the Supreme Court indicated that courts might 69. The Constitution itself provides no final formula judiciary is proper. The sweeping phrases “‘Nothing in sion’’ suggest an absolute derogation from the coverage 15(4) speaks of “special” provision only, and special can

for when intervention by the this Article” and “any proviof Art. 15. But, per contra, Art. be said to militate against the

416

©The Judicial Process

enquire whether what the government did was within the scope of the

purpose of Article 15(4)—and this scope was to be ascertained by

consideration of a variety of publi¢ interests. In regard to admission to medical colleges Balaji found that excessive reservation neglected and

undermined important national interests in the full utilization of talent.

In regard to government posts, Devadasan found that excessive reservation threatened the public interest in morale and incentive in national services. But Abdul Latifis notable for the absence of any consideration of public policy. It is difficult to see what the analogous public interest is in this case. Can any national interest in motivation for talent ormorale in services be said to be present in the case of a bhang shop?’ May-not

government decide that the encouragement of enterprise in this field is

outweighed by the benefit to society from channelling these opportunities to disadvantaged groups?7! Does the public have any real interest in assuring opportunities to

invest in bhang shops to members of non-backward communities? If the

courts are to undertake a balancing of interests, surely. the public interest is not to be equated with the interests of the sections of the population that do not erijoy reservations. As many courts have noted, special treatment for backward groups is authorized precisely because there is a strong national interest in the advancement of the weaker sections of society.?? In the case of women, no one doubts that special provisions for the advancement of women are justifiable, even when

men are thereby put at a temporary disadvantage.”3 It is presumed that what helps women to assume a productive place in the society is good for the society. The interests of the society cannot be equated with the interests of the men who are temporarily discomfited by the special treatment of women. Similarly, there is no justification for equating the

interests of the advanced

classes with those of the nation.74 The

absoluteness of the derogation. Plainly, the power of review here flows, as Upendra Baxi

has noted, not from the text of the articles but from the structure of Part HI of the

Constitution—from Arts. 12, 13 and 32. 70. It is particularly difficult to find an element of efficiency or incentive where, as here, the selection is by lot. Concern for enterprise in this area is especially ironic when we note that the prohibition policy commits the government to eventually taking away all rights to operate such shops. 71. May government not decide that extensive reservation might be desirable in promoting a particular development scheme?

72. Balajiv. State of Mysore, A. 1. R. 1963S. C. 649 at 662. Cf. Vishwanathv. Government of

Mysore,A.1. R. 1964 Mys. 132 at 136; Hariharan Pillai v. State of Kerala, A. 1. R. 1968 Ker. 42, 73. For a few examples, see Dattatraya v. State of Bombay, A. 1. R. 1953 Bom. 311;

Ramchandra Mahton v. State of Bihar, A. 1. R. 1966 Pat. 214. 74. See text at nn. 36, 58, above.

‘Scope and Extent of Preferences

417

Constitution makers had the foresight to realize that the overcoming of

vast social and economic disparities was a requisite for having a modern nation. While they entrusted this task primarily to the government, there is some room for judicial review, to see that the power is used only for the public interest. But it must be recalled that one of the public interests in the balance is the nation’s long-run interest in the spread of opportunities and resources into the lower strata of the population. This interest may sometimes be outweighed by other pressing interests. But it is difficult to see how it failed to prevail against an interest as indefinite, diffuse, and questionable as that involved in Abdul Latif.

In the delicate task of balancing the merit principle with other interests, a flat percentage limitation on the extent of reservations is of less use than it might appear. It has the virtue of blocking schemes of reservation so expansive as to transform the competition into a set of communal quotas. But the results of Devadasan and Abdul Latif provide eloquent testimony of the possible dangers of the mechanical application of such a standard. The percentage of reservation may be a poor indication of its impact: where reservations go largely unfilled. Beyond

this, the flat, percentage limit contains other dangers of deception. Even

where a reservation is filled, the size of a reservation is not necessarily

directly related to the amount of distortion of the merit principle in a

given competition. For example, a reservation of 50% of seats for a group of backward-class candidates who would obtain 40% of the seats on merit is actually less of a distortion of merit than a 25% reservation

fora group of backward classes who would obtain only 5% of the seats

on merit. In the first case, only 10% of the successful candidates would not have prevailed in open merit competition; in the second case, 30% would not have been selected on merit. The effective impact of the reservation in the second case is twice as great as in the first case,

although the reservation in the second case is only halfas large. The real

impact of a scheme on the chances of others and on the merit principle cannot be known from the percentage of places reserved. The effective impact of a reservation, then, depends not orily on its size, but on a varicty of other factors: whether it operates as a guaranteed minimum or on an “over and above merit” basis; whether there is a minimum floor of qualification that must be met; how many of

the group for which reservation is provided would succeed on merit; how the reservation is divided into compartments; how the unfilled reserved places are disposed of. We shall be taking up several of these matters below. In the meantime, it should be clear that the impact ofa

reservation on the chances of others or on the merit principle cannot be determined from the nominal size of the reservation. What is needed isa

way of estimating the net effect of resrvations. One simple wav of doing

418

The Judicial Process

this would be to ask how many those selected by a particular scheme of selection would have been selected on a pure merit basis.75 It would

then be possible to compare the effective extent of reservation, rather than its nominal extent.

Since Devadasan the Supreme Court has never returned to full-dress

consideration of how much preferential treatment is permissible, but it

has obliquely undermined its earlier learning on the subject. When the Court encountered the quantity-of-reservation question a decade later

in State of Kerala v. N.M. Thomas,’6 it was entangled with much broader

issues about the scope of the government’s power to pursue policies of compensatory discrimination. The Kerala service rules specified a period during which government servants temporarily appointed to

higher posts might pass the tests required for permanent appointment in those posts. A series of government orders granting Scheduled Castes

and Tribes employees a temporary waiver from required examinations enabled a number of Scheduled Castes Lower Division Clerks to be promoted to Upper Division Clerks on the basis of seniority in their

higher posts. During 1972, out of 51 vacancies in the category of Upper

Division Clerks, 34 were filled by Scheduled Castes members who had not passed the tests; only 17 were given to persons who had passed the test.77 Thus the beneficiaries of preference constituted two-thirds of that year’s appointments; this preponderance was one of the grounds upon which the scheme was struck down by the High Court.78 A seven-judge bench of the Supreme Court reversed, five to two.

Since it overruled the High Court on this, the Devadasan rule of 50% in any year seems seriously undermined. However, there are suggestions

in the five majority opinions that the scheme in question was not a reservation regulated by Article 16(4) but a preferential device em-

powered directly by Article 16(1) or (2). If so, it may be that the Devadasan rule is still applicable to reservations even though it does not 75. Data on the extent to which backward classes would have obtained seats on merit

are scarce. The Kumara Pillai Commission in Kerala (1965: 134) presented some data

of this kind. For example, in medical school admissions in 1962-63, 40% of the seats

were reserved, but in fact 71% of the successful candidates would have been selected on

the basis of merit. The effective reservation, then, was 29%. Less systematic data come to light in Sagar v. State of Andhra Pradesh, A. 1. R. 1968 A.P. 165 at 189. In a medical

college selection, 123 of the 189 backward classes candidates admitted would have secured places on merit. To the extent that reservations are not effective, they are

unnecessary and should be abolished. But the point here is that they may be ineffective and at the same time deceptive as to the amount of special treatment actually being

conferred.

76. State of Kerala v. N. M. Thomas, A. 1. R. 1976 S. C. 490, The broader issues in this case are discussed at chap. 11, §D, above.

77. Id., at 495.

78. I. L. R. 1974 (1) Ker. 549.

,

Scope and Extent of Preferences

419

apply to other preferential measures regarding government employ-

ment. Of the five majority judges, three address the quantity of preference question in their opinions. Chief Justice Ray in the majority judgment finds the High Court mistaken in condemning the 34/51 as excessive. The promotions made in the services as a whole are nowhere near 50 per cent of the total number of posts. The Scheduled Castes and Scheduled

Tribes constitute 10 per cent of the State’s population. Their share in the gazetted services of the State is‘said to be 2 per cent, namely 184 out of 8,780. Their share in the non-gazetted appointments is only 7 per cent, namely 11,437 out of 1,62,784 . . . therefore . . . the orders are meant to implement

notonly . .

Apparently, posts in the seems to be point as the their portion

. Article 335 but also the Directive Principle under Article 46.7%

the 50% limit is to be measured by the total number of services rather than the number of promotions. Again it suggested that higher than 50% is permissible until such protected groups have places in service proportionate to in the population.

Fazl Ali, J., concurring, frankly repudiates Devadasan and suggests

that such a rule may defeat provisions for preference in public employment.® The 50% limit “no doubt laid down” by earlier decisions of the Supreme Court is to be taken as “a rule of caution” which “does not

exhaust all categories.”81 Justice Krishna Iyer expresses agreement

with Faz] Ali’s reading of Article 16(4).®? This is somewhat confusing,

since these judges found that the provision in question is justified directly under Article 16(1) and does not involve Article 16(4). Krishna

Iyer, J., too would not agree with an arithmetical limit of 50% in any

one year. “Overall strength in a department does not depend on recruitment in a particular year, but the total strength of a cadre.”’83

By overturning

the High

Court’s application of Devadasan the

Supreme Court reopens the question of the quantity of preference that is constitutionally permissible. Unfortunately, the question is not well

framed by the Thomas facts. Kerala’s award of two-thirds of a year’s

promotions to Scheduled Castes appears less a deliberate application of

policy than a temporary (and possibly inadvertent) result of a series of

earlier unsuccessful measures designed to facilitate SC promotions. The Court’s response plunges the whole question into obscurity. It upholds at least temporary provision for more than 50% of current appoint-

ments.® It is unclear, though, whether it is abandoning the Devadasan 79. A. 1. R. 1976S. C. at 501.

81. Id. at 554-55.

80. Id., at 555.

82. Id., at 537.

83. Id., at 537.

84. In response to the Thomas case the Government of India in 1977 abandoned its

earlier instructions (embodying the Devadasen 50% limit) and ordered that even 100%

420 The Judicial Process 50% rule as to all reservations under Article 16(4)—as a later bench implies.85 And, beyond that, it is unclear whether the entire Balaji

limitation on the extent of preferences and the extent of preferred groups is unsettled. Justice Fazl Ali, the only member of the majority who is explicit about this, makes it clear that he would reopen the whole

question of the scope of preference policy. Suppose, he asks, a state had a population of which 80% were backward classes and the government

reserved 80% of the jobs in its services for them. And he answers: it

could not be said that such a provision “violates the permissible limits”’ of Article 16(4).8° C. THE INTENSITY OF PREFERENCE Connected with the amount of preferential treatment is a set of problems that may be thought of as questions about its intensity: just how

far may the State deviate from the standards it would otherwise employ?

May it waive minimum standards altogether? Just how much may an individual beneficiary be helped and how much may those around him

be discomfited? May the State deviate as much in distributing outcomes (services to citizens or desirable posts) as in distributing opportunities (admission, etc.) or improving access to them (e.g., by coaching or waiver of application fees, etc.)? It is accepted that the State may lower barriers to opportunity

exclusively for backward classes: provisions for waiver or reduction of

filing fees, application

fees, and so forth are widespread

and have

attracted few challenges.87 The Representation of the People Act’s provision for lower election deposits for Scheduled Castes and Tribes

was held not to be discriminatory, for it enabled members of these groups to contest elections.88 Clearly, preferences need not take the form of reservations, but may be to enable members of these groups to enter the competition by waivers of fees, age limits, and so forth. In line with

the general notion that special measures to enable them to compete at of available posts could be reserved where reservations had gone unfilled in previous years (Department of Personnel and A.R.O.M. No. 16.3.1973 ESD [SCT] dated 27.12.1977). 85. Thirteen months after Thomas, in Chauhan v. State of Gujarat, A.1.R. 1977S. C. 251 at 263, the Court (per Krishna Iyer, J.), after distinguishing the Devadasan carryforward situation on the facts, remarks that “it is doubtful whether Devadasan’s-case

survives State of Kerala v. N. M. Thomas. . . .”” 86. A. I. R. 1976 S. C. 490 at 555. This inference may gain persuasiveness in the setting provided by the new primarily tribal states in northeastern India. InNagaland

there are indeed 80% reservations for Scheduled Tribes.

87. See cases cited in thap. 11, n. 47.

88. Rajuv. State, A. 1. R. 1976 Guj. 66.

Scope and Extent of Preferences

421

intake are unobjectionable, special arrangements. have been made to prepare members of these groups for examinations for government

service.89

It is established that the State may, to insure the inclusion of these groups, employ some lower standard of selection at the intake point. For

example, SC and ST applicants for Class III posts in central services are

judged by “‘relaxed standards.””®° It is generally accepted that reservations may entail the setting of a lower floor of minimum qualifying

marks for members of the preferred groups. The common practice*! of most medical colleges.to relax the minimum qualifying marks for SC and ST by'5% has been upheld; it is ‘‘open to the State to... fix different minimum marks” so long as.the difference is reasonable for

purposes of Article 14.92

But courts are divided on how far this can go. In Dilip Kumar »v.

Government of U.P.,.a Division Bench of the Allahabad High Court held

that it was allowable to reduce the minimum qualifying marks “‘to.any reasonable figure below 30 per cent,” but that to specify “that there shall be no minimum qualifying marks for the candidates from Uttar-

khand Division” was “unreasonable and arbitrary” because it defeats

“the object-of securing suitable candidates.” In a different case, the

Supeme Court eventually upheld the state’s designation of residents of

Uttarkhand as backward, without taking up the “no minimum marks” question. However,

in

Gupta

v. Union

of India,95 another

bench

of the Al-

89. See the account of coaching centers for preparing members of Scheduled Castes and Tribes for high level examinations discussed in chap. 4, §B, above. 90. In non-technical and quasi-technical Class III posts, if the relaxed standards fail to admit enough candidates to fill the reservation the best candidates who fill the minimum educational qualifications are to be appointed and given in-service training.

In practice, appointing authorities often deviate from these instructions (RCSCST

1975-77: 1, 50). 91. RCSCST 1975-77, I] 36-37. U2. Sukhdev v, State of Andhra Pradesh, 1966(1) Andh..W. R. 294at 312. That such lowet

Standards are also permissibile in the government employment area is implied in General Manager v. Rangachari, A. 1. R. 1962 S. C. 36 at 42; Devadasan v. Union of India, A.1. R.

1964 S.C. 179 at 187.

93. Dilip Kumar v. Government of U.P., A. 1. R. 1975 All. 592 at 596. The no-minimum-

marks feature produced a very high variance between reserved and general candidates.

In Pradip Tandon v. State subsequent U.P. medical obtained a test score of candidates) was 266. The marks.

of U.P., A. I. R. 1975 All. 1 at 2, it is reported that in a admissions competition the last admitted general candidate 271 marks and the lowest score on the waiting list (of 100 lowest successful candidate in the reserved category had 128

4. State of U.P. v. Pradip Tandon, A. 1. R. 1975S. C. 563. 95. A. T. R. 1974 All. 288.

422

The Judicial Process

lahabad Court not long after dealt with a case of admission to the elite Indian Institute of Technology. The normal minimum qualifying marks were 40%, and it was alleged that persons receiving as little as

10% had been admitted to reserved seats. Once a reservation has been made, says the court, “‘the method of filling . . . reserved seats may be different. . . . The qualifying marks for this class may also be different

or there may be no qualifying marks at all.” The court asserts that such a waiver cannot lead to the dire result of lowering the quality of engineers, prophesied by the petitioner, for “it is only at the stage of admission that concession is granted. ... No concession is being granted them in passing the final examination: unless they qualify themselves through the education given them at the Institute they will not pass and the country will not get technologists and engineers who are not fully qualified.”97 Thus the court visualizes a combination of maximum flexibility at the intake stage with inflexible standards at the

output. But it is not clear that this neat pattern is constitutionally mandated or readily realizable.%*

The question has never been addressed directly. Where it was ob-

liquely present in several recent cases, the judicial response is so dispa-

rate that it is hard to predict what would happen if it were encountered directly. In Janki Prasad v. State of Jammu and Kashmir, the Supreme

Court struck down the use of the selection committee in the promotion of teachers to headmasters, because the procedure was one in which relevant information was disregarded and the standards were set so low as to reflect abandonment of concern with quality. But the setting of absurdly low minimal scores made it a “travesty of selection.”100

Although reservations for Backward Classes were involved in the selec-

tion, it was not a waiver of quality for them that was objected to, but the setting of a general floor for all candidates. Also, this selection was contaminated by other constitutional shortcomings: there was no use of essential information about performance on the job, and the Backward Classes reservations were defective in various ways. Again, the estab-

lishment of these inadequate selection standards was seen as part of a prolonged effort to effectuate a scheme of unconstitutional communal 96. Id., at 289.

97. Id.

98. The system approved in Gupta proved unsatisfactory, for many of those recruited

“even after special coaching . . . could not withstand the strain of the five year course.” A new arrangement admitted only those Scheduled Caste and Scheduled Tribe students who obtained two-thirds of the marks obtained by the lowest ranking successful general candidate, with a special system of recruitment to fill the remaining reserved places (RCSCST 1975-77: I, 100). For a sensitive account of the working and effects of the preferences at 1.T.T. Bombay, see Kirpal 1978. 99. Janki Prasad v. State of J. @ K., A. 1. R. 1973 S. C. 930. 100. Id., at 936.

Scope and Extent of Preferences

423

quotas, an effort marked by shifting stratagems and defiant evasion of

judicial decrees.1°' If Janki Prasad indicates that the courts will scrutinize the setting of minimal standards, it tells us little about how much a state may vary minimal scores in a selection where relevant information is utilized, and where there is a reservation for legitimate beneficiaries that is legitimate in scope.

The intensity of preference that is allowable what kind of opportunities are at stake. Is distribution of individual benefits, or is the directly affected? Is the matter in question the

may vary, depending on it primarily a matter of quality of public service consumption of benefits

or the production of them for the public? Thus in Janki Prasad, the Court

observes that

where appointments and promotions to responsible public offices are made, greater circumspection would be required in making reservations for the benefit of any Backward Class because efficiency and public interest must always remain paramount. It is implicit in the idea of reservation that a less meritorious person is to be preferred to another who is more meritorious. 1°2

If reservation implies inferior service to the public and the public’s interest is paramount, the scope for reservation might be severely restricted—at least in regard to “responsible public offices” like

headmasters.

The “intensity” issue surfaces briefly in the Thomas case. Scheduled

Caste clerks were kept in the posts of Upper Division Clerks for many

years by a series of orders exempting them temporarily from the re-

quirement

of passing various departmental

examinations. Justice

Khanna, dissenting, holds that in regard to government posts “it is not

permissible to waive the requirement of minimal educational qualifica-

tion and other standards essential for the maintenance of efficiency of service.”1° The majority judges concede that Article 335 inserts into any scheme for preference in government employment a requirement that it be consistent with the efficiency of administration. Yet none of them appear troubled by the possibility that efficiency might be com101. The headmaster posts at issue in Janki Prasad are the very same ones that were

involved in Triloki Nath Tikku v. State of J. & K., A. 1. R. 1967 S. C. 1283 and A. I. R. 1969

S. C. 1. Janki Prasad was the fourth appearanceof this dispute in the Supreme Court. In Makhan Lal v. State ofJ. & K., AAR.

1971 S. C. 2206, 2209, 2210, the state was

admonished for responding to Triloki Nath with “absolute non-compliance” with its own rules and “complete defiance of the law laid down by this Court.”

102. A. I. R. 1973 S.C. at 939.

103. A. I. R. 1976 S. C. 490 at 506.

104. Even compensatory measures to ensure equality of opportunity justified by a broad classification argument may “not dispense with the acquisition of the minimum basic qualification necessary for the efficiency of administration” (Justice Mathew, id.,

at 519),

424

The Judicial Process

promised by the waiver of these by other evidence of competence. duled Caste clerks are “senior ence.””1 Justice Beg notes that

examinations. Some Justice Fazl Ali notes ... and have gained these employees must

are reassured that the Schemore experibe “otherwise

qualified and suitable for appointment” and were “‘presumably quite

satisfactorily discharging their duties in the higher grade.”’1°7 Others indicate some skepticism about whether the test really measures qualification for the job. Thus, Justice Mathew finds the existence of lengthy

exemptions for all employees an indication that “passing of the test is not absolutely essential for holding the post.” And Justice Krishna Iyer introduces an elaborate “‘trichotomy” to distinguish “basic qualifi-

cations for eligibility to a post,” “secondary qualifications . . . useful to discharge the duties of the post,” and “virtues which will make the employee ultra efficient.”1 The tests in Thomas are “demanded for better performance, not basic proficiency.” Since “their absence [is] not - fatal to efficiency,” relaxation of standards is allowable “in suitable . cases.”110 Suitability in this case is established by the Olympian

observation that

After all, here he is a pen-pushing clerk, not a Magistrate, accounts officer, forest officer, sub-registrar, space scientist, or top administrator or one on whose initiative the wheels of a department speed up or slow down. 111

Summing up, Justice Krishna Iyer suggests that “to relax on basic ualification is to compromise with minimum administrative effi-

ciency” and is presumably barred by Article 335. But “to relent fora time

on additional test qualifications is to take a calculated but controlled risk, assured of a basic standard of performance.”12 Presumably, then, such additional qualifications can only be waived in “suitable cases,”

“for a time,” only where there is some assurance of a “basic standard of 105. Curiously, the majority judges in Thomas erect a temporal counterpart to the

notion that there cannot be a waiver of minimum marks. Thus they emphasize that the state could not dispense with the tests forever, for to promote a Scheduled Caste clerk “without passing any test at all . . . [would] destroy the concept of equality” (Faz! Ali, J., A. I. R. 1976 S. C. 490, at 551). In spite of their emphatic embrace of substantive

rather than formal equality, a close reading of the Thomas majority opinions suggests that the substantive equality pursued by compensatory classification is not the only

equality to be found in Art. 16. 106. Id., at 552.

107. Id., at 523.

108. [d., at 520. At one point Justice Krishna Iyer seems about to take up the question

of the way in which

ficiency is defined: “efficiency means. . . not marks in examina-

tions only, but responsible and responsive service to the people.” But he veers away without developing the point (id., at 534). 109. Id., at 527. 110. Zd., at 527. 111. Id., at 527. See his further observations on “alert quilldriving,” id., at 536. 112. Id., at 527.

Scope and Extent of Preferences

425

ce.” Thus this acceptance of broad waiver in Thomas carries

the suggestion that lowering of standards for compensatory purposes is limited and hints at principles for judicial definition of those limits. Thomas suggests that the source of those limits is the competing commitment

to efficient administration, not the frustration of other

employees. Reservations that are otherwise valid are not invalidated by the fact that they endow beneficiaries with very extensive benefits, to

the discomfiture of those surrounding them. Thus, reservations in promotion that entail the promotion of junior officers over their seniors are allowable, notwithstanding the vexation of the latter, because of long-run benefits from erasing group disparities. 113 Although the State may make such generous waivers, it is not required to do so. When the State of Orissa conducted a combined examination for several of its services, 133 candidates were successful—i.e.,

had test scores high enough to support appointment. Although 18% of

the places in each of the two services were reserved for Scheduled Castes, there was just one successful SC candidate, who had scored

105th on the examination. Of the successful candidates, 33 were ap-

pointed to the elite “A class” posts and 53 to “‘B class” posts. The sole

SC appointee challenged the Public Service Commission’s recommen-

dation of him to a “B class” post and sought an order that the reservation -be implemented by appointing him to an “A class” post. The High

Court declined to intervene, for such an appointment depended on whether the PSC thought him suitable, “having regard to the maintaining of the efficiency of the administration.”""* Thus the government’s preference policy is mediated through judgments involving competing goals of policy. This is most clear in the government service area, where Article 335 requires an assurance that efficiency will be maintained. But similar considerations would seem to be present wherever service to the public rather than distribution of purely individual benefits is involved. D. THE WIDER USE OF THE RESERVATION DEVICE The reservation device has been increasingly employed to assure

places to other categories of persons, in addition to the backward

classes, who seem especially desirable or deserving. Government jobs

have occasionally been reserved for political sufferers and exservicemen. But professional and especially medical education is the place where reservations have proliferated. From time to time, states have provided reservations in the professional colleges for such dispa-

Tate categories of persons as overseas Indians; refugees; nominees of 113. State of Punjab o. Hira Lal, A. 1.R. 1971 S.C. 1777. 114. Sudharkarv. State,A. 1. R. 1974 Or. 224.

426

The Judicial Process

neighboring states or of the central government; students who have

distinguished themselves in sports, extra-curricular activities, National

Cadet Corps, or President’s Scouts and Guides; children of medical practitioners, armed services personnel, university employees, displaced goldsmiths, and “political sufferers.” In addition, there are

sizable reservations for women. State governments have attempted to distribute the available seats by residence in districts or regions and to allocate

seats

examinations.

among

those

who

qualified

by

several

alternative

None of these categories are “backward classes”: the provisions of Article 15(4) are not applicable. But the allotment of places under these headings has affected provisions for backward classes in several ways. Before exploring the connections, it is necessary to sketch the judicial response to these other categories of reservation. Reservations for women are upheld simply on the ground that Article

15(3) authorizes the State to make special provisions for women.115

The

other reservations and distributions, when

challenged,

have to

meet the general requirements of reasonable classification of Article 14:116 they must be based upon an intelligible differentia and bear a rational relation to the object being pursued. The results have been

varied.

District and regional residence distributions were upheld in several states!!7 before the Supreme Court held that “district-wise”™®8 and [regional] “unit-wise” 19 selections are contrary to Article 14. Distribution of places among candidates qualifying by different examinations was struck down,??° but distribution among the graduates of different universities has been upheld.121 115. Sagarv. State of Andhra Pradesh, A.1.R. 1968 A.P. 165; Ramchandra Vishnu v. State of Madhya Pradesh, A. 1. R. 1961 M.P. 247. Since sex is not one of the criteria forbidden by

Art. 29(2), which deals specifically with admissions to educational institutions, such

reservations might be valid even without consideration of Art. 15(3); cf. University of adios 2 Shantha Bai, A. 1. R. 1954 Mad. 67; Anjali o. State of West Bengal, A. 1. R. 1952 al,

116. P. Rajendran v. State of Madras, A. 1.R. 1968S. C. 1012; Subhashini v, State of Mysore,

‘ t R. 1966 Mys. 40. Contra: Ramchandra Vishnu v. State of Madhya Pradesh, A.1.R. 1961

117. Sagar v. State of Andhra Pradesh, A.1. R. 1968 A.P. 165; Lalita Shuri Tikku v. Stateof

Jammu and Kashmir, A.1.R. 1966J. & K. 101.

118. P. Rajendran v. State of Madras, A. 1. R. 1968 S. C. 1012. This result was anticipated in State of Kerala v. Jacob Mathew, 1. L. R. 1964(2) Ker. 53.

119. Periakaruppan v. State of Tamil Nadu, A. 1.R. 1971S. C. 2303.

120. State of Andhra Pradesh v. Balaram, A. 1. R. 1972 S.C. 1765. An earlier version was

upheld in Sagar v. State of Andhra Pradesh, A. 1. R. 1968 A.P. 165; an even earlier one was

struck down on the ground that the shares were disproportionate, Sukhdev v. Government of Andhra Pradesh, A. 1. R. 1966(1) Andh. W. R. 94 (1963); and a yet earlier one upheld in Gullapalli Nageswara Rao v. Principal, A.1.R. 1962 A.P. 212. 121. Chanchala v. State of Mysore, A. 1. R. 1971 S. C. 1762. Universities can favor

Scope and Extent of Preferences

427

Courts have struck down reservations for President’s Scouts and

Guides, !2? and for children of medical practitioners, !23 university employees,124 legislators,125 and displaced goldsmiths,126 on the ground

that they have no relation to the object of medical education.

On the other hand, judges have uniformly endorsed reservations for outstanding sportsmen on the grounds that they exhibited qualities which were desirable in medical practitioners.127 Reservations for

children of defense personnel and ex-servicemen were thought to be

clearly related to defense policy.128 Similarly, children of political sufferers may be disadvantaged by the unsettled life and economic sac-

rifices of parents; a careful and detailed provision which makes it possible to identify them has been held to have

‘“‘a reasonable nexus

with the object of the rules which can be nothing else than a fair and just

distribution of seats.”’129 Looser provisions have been struck down or have provoked judicial condemnation.11 Reservations for childen of applicants who qualify by their own examinations—i.e., are their own students. Since student bodies are largely local, this has some of the same effect as the invalid districtwise distribution. See Katzenstein 1979, chap. 9.

122. Sagar v. State of Andhra Pradesh, A. 1. R. 1968 A.P. 165.

123. Jacob Mathew v. State of Kerala, A. 1. R. 1964 Ker. 39 at 68 (Aff. on this point by

1. L. R. 1964 [2] Ker. 53); Surendrakumarv. State of Rajasthan, A. 1. R. 1969 Raj. 182. 124. Umesh Chandra Sinha v. V. N. Singh, A. 1. R. 1968 Pat. 8. 125. Surendrakumar v. State of Rajasthan, A. 1. R. 1969 Raj. 182. 126. Sagar v. State of Andhra Pradesh, A. 1. R. 1968 A.P. 165. 127. State of Kerala v. Jacob Mathew, 1. L. R. 1964(2) Ker. 53 (rev’g on this point A. I. R. 1964 Ker. 39); Subhashini v. State of Mysore, A. I. R. 1966 Mys. 40 at 46 (“ta good

sportsman cannot afford to be a book worm . . . he is likely to be a better doctor or engineer than his competitor who knows only books but not men and matters”); Sukhdev

v. Government of Andhra Pradesh, 1966(1) Andh. W.R. 294 (1963). 128. Subhashini v. State of Mysore, A. 1. R. 1966 Mys. 40; Surendrakumar v. State of Rajasthan, A. 1. R. 1969 Raj. 182; Sardool Singh v. Medical College, A. 1. R. 1970J. & K. 45; Kushma Joshi v. Pro-Vice Chancellor, A. 1. R. 1969J. & K. 136. Similarly, reservation of posts for ex-servicemen has been held permissible to encourage young men to join the forces and expose themselves not only to the dangers of that profession but to the additional hazard of compulsory early retirement (Daya Ram v. State of Haryana, A.1.R.

1974 P. & H. 279). 129. Chanchala v. State of Mysore, A. 1. R. 1971S. C. 1762 at 1776. 130. Surendrakumar v. State of Rajasthan, A. 1. R. 1969 Raj. 182. 131. In Ramchandra Vishnu v. State of Madhya Pradesh, A.1.R. 1961 M.P. 247; Vinor Sagar Sood v. State of Madhya Pradesh, A. 1. R. 1967 M.P. 182. The Madhya Pradesh Court,

while condemning these reservations as unrelated to the object of promoting efficiency in the medical profession, said that if it were a statutory provision it would not hesitate to strike it down as violative of Art. 14. But since it was merely an executive instruction and not a law, it did not attract Art. 15. The hope expressed in the first case that government would respond favorably to judicial animadversions was proved fruitless in

the latter. In the reductio ad absurdum of such reservations, the State of Himachal Pradesh

reserved medical school places for wards of “‘political sufferers and persons of Himachal

428

The Judicial Process

Indian diplomats and for various categories of out-of-state students

(overseas Indians, Colombo Plan scholars, refugees from Burma, Asian and African students, Goans, residents of Union Territories and of

other states) were held to be reasonably related to the State’s policies of

breaking down barriers and creating an integrated society.132

Whatever plausibility one accords the supposed link between athletic

ability and qualifications for medical practice, the claim on behalf of children of political sufferers or of diplomats is of a different kind. It is

not a claim of unrecognized qualification, but a claim for fairness to

offset disadvantage. Government might, in addition, assert an interest

in the composition and morale of the diplomatic corps or in integrating

the society by including residents of other states and so forth. What are

the objects that government may legitimately pursue when it distri-

butes, for example, medical admissions? Presumably it aims to select

those best qualified to practice medicine, and presumably, it is not

confined to prior academic performance in judging such qualifications.

Indeed, if the goal of the exercise is to enhance the provision of health

services to the population, there may be an obligation to enlarge the

notion of qualification accordingly (to include, for example, inclination to provide service in under-serviced areas and perhaps disinclination to emigrate).133 But, may government take into account other goals, unrelated to provision of health services? In one area, the answer is clearly yes— the State is empowered to promote inclusion of the backward classes while distributing places.134 May it pursue other goals that

are extraneous to health services?

Pradesh with outstanding social background.” A Division Bench struck this down as violative of Art. 14 (Sukhvinder Kaur v. State of Himachal Pradesh, A. 1. R. 1974 H.P. 35). A few months later, when the son of an I.N.A. veteran sought to be admitted under this heading, another Division Bench (including one of the same judges) went to heroic lengths to produce an acceptable construction, finding that’ [t] he phrase ‘outstanding social background’ necessarily refers to outstanding performance ... in the political

field, coupled with . . . backwardness in the matter of education” (Registrar v. Shashi Pal

Singh, A.1.R. 1974 H.P. 41 at 45). 132. Chitra Ghosh v. Union of India, A. 1.R. 1970S. C. 35; Subhashiniv. State of Mysore, A.

1. R. 1966 Mys. 40; Sagar v. State of Andhra Pradesh , A.1.R. 1968 A.P. 165.

133. A substantial proportion of India’s medical graduates do emigrate. One report estimates that from 1961 to 1964, 18% of medical graduates emigrated annually, leading to a permanent loss each year of about 7% of all medical doctors (World Bank 1975: 83). Indeed, M. R. Balaji, petitioner in the leading case, was reported in 1979 to be practising medicine in the United States (interview with S. K. Venkata Ranga lyengar by Robert Hayden in Bangalore, Oct., 1979). Likelihood of emigration might be thought relevant if measurable (and if the administration of such measures were feasible). But cf.-the judicial response to a post hoc claim of selection on the basis of providing service to rural areas (discussed above in chap. 11, at nn. 56 ff.). 134. Inclusion is used here to comprehend the whole range of goals (acceptance,

integration, fairness, representation, redistribution, incubation) which may be present in measures favoring backward classes. For a sketch of these, see chap. 3, §D, below.

Scope and Extent of Preferences

429

This question has not been addressed so clearly as by a divided 1955

Full Bench of the Patna High Court in Sukhnandan Thakur v. State of

Bihar.35 A retrenched employee complained that in appointing and

retaining supply inspectors, the state had favored political sufferers and

displaced persons over other employees with greater seniority. Chief Justice Das, dissenting, asks: What...are

the

objects

of

employment...under

the

State.

Un-

doubtedly, one object is to get the work of the office done efficiently. I am . .. unable to agree that this is the sole object of state employment. . . . State employment can be for other purposes as well . . . here the directive principles in Part IV of the Constitution . . . are relevant.137

Reading Article 39’s directive about adequate means of livelihood and

Article 41’s about the right to work, he concludes that

the object or purpose of state employment may [also] . . . be to give public

assistance in cases of unemployment or any other cases of undeserved want.38

Displaced persons and political sufferers are instances of undeserved want, and the preference for them is justified. Where a large number of persons possess qualification for a job, a selection must be made “‘based on some principle which is related, not so much to the required or minimum qualification for the office, but to something . . . which is

peculiar to the candidate himself. . . .””439

If, under conditions of oversupply, the selection process could be used to pursue other policies without significant sacrifice of “merit,” it did not strike the other judges as constitutionally permissible. It was not open to government to set qualifications unrelated to “‘the demands and needs of the office.” The Government’s test was not rational, for it is

“based on extraneous or collateral considerations” *1 By addressing these irrelevant considerations, the appointing authority has violated

Article 16(1). “[T]he essential point is that the qualities or qualifications required of the candidate must have some relevance to the efficient

performance of the duties and obligations or the particular office.” #42 Although it was never again addressed so explicitly, there has been a clear shift away from this view. Chitra Ghosh v. Union of India**3 concerned 135. A. I. R. 1957 Pat. 617. . 136. Of the 169 Supply Inspectors appointed, 52 were political sufferers and 10 were displaced persons (id., at 628). 137. Id., at 624. 138. Id., at 624. 139. Id., at 626. 140. Ahmed, J., id., at 621. The judge notes that Art. 16(4) and Arts. 16(3) and 16(5) represent an exception to this and that their presence argues that “no other exception was intended” (id., at 622).

142. Ramaswanmii, J., Id., at 632.

141. Ramaswami, J., id., at 629.

143. A. I. R. 1970S. C. 35.

430

The Judicial Process

admissions to a medical college in Delhi. In addition to reservation of

15% for SC, 5% for ST, and 25% for women, there was provision for 23 of the 125 seats to be reserved for the following categories and filled by nominees of the central government:

(c) Sons/daughters of residents of [specified] Union Territories . . .

(d) Sons/daughters of Central Government servants posted in In(e) (f) (g) (h)

dian Missions abroad Cultural Scholars Colombo Plan Scholars Thailand Scholars Jammu and Kashmir State Scholars*44

Applying the tests of permissible classification under Article 14, the Court notes that since it is not possible to throw open admission to students from all over the country, “the governnient cannot be denied the right to decide from what sources the admission will be made.”245 This is a matter of policy depending on “an overall assessment . . . of

the requirements of residents of particular territories and other categories of persons for whom it is essential to provide facilities for medical education.”’46 Have the sources been properly classified? The Union Territories and Jammu

and Kashmir categories are responsive to special shortages

of medical education facilities. The Cultural, Colombo Plan, and Thai-

land scholars are justified by “reciprocal agreements of an educational 47 As to the children of diplomats: “‘it is . . . well and cultural nature.” known that due to exigencies of their service these people are faced with

a lot of difficulties in the matter of education.”"*8 With the sources

classified in this manner, “it is difficult to see how that classification has no rational nexus with the object of imparting medical education and also of selection for the purpose.’’!49 This is a revealing locution. It is evident that some of these sources are s¢lected on grounds of fairness,

avoidance of hardship, indirect contribution to cultural enrichment, or the morale of the overseas employees and are not directly related to the specified object of securing “the best possible material” for medical training.45°

The

single-minded

“relevant

qualifications”

notion

of

Sukhnandan is at least enlarged in some unexplained way and possibly displaced entirely. This suggests large possibilities for the use of the 144. Id., at 37. Since the reservations for women

are to be made after the Government

of India nominations are filled, the maximum reservation could run as high as 58%. But

in fact it was less. The Government of India nominated only 9 allowed for categories (c)-(h). And very likely the net effect women was much less than the 25% specified. 145, dd., at 39. 146. Id., at 39. 147, Id., 148. Id., at 38. 149. Id., at 39. 150. Id.,

persons against the 23 of the reservation for

at 38. at 38.

Scope and Extent of Preferences

431

reservation device to pursue all kinds of governmental policies. We shall return to the question of possible conceptual limits after considering whether there are fixed quantitative limits on its use. A new sort of “how much” problem arises when reservations under these other headings are combined with the reservation for the backward classes under Article 15(4) and the total is above 50%. Does the Balaji-Devadasan 50% rule apply to all reservations—i.e., does it concern the extent of the opportunities for merit—or is it merely a limita-

tion on what the State can do for the backward classes? In Subhashini v. State of Mysore,15) 16.1% of the 750 seats-were reserved for miscellaneous categories (out-of-state, foreign students, sportsmen, etc.), and then 48% of the remainder were reserved for the Backward Classes, Scheduled Castes, and Scheduled Tribes. Thus a total of 56.3% of the seats

were reserved and only 43.6% remained open to merit competition. To the argument that Balaji ruled out reservations which reached higher than 50%, the court replied that that case-was concerned only with the permissible extent of reservations under Article 15(4) and did not include other reservations.152 Since each of the reservations here was reasonable, the Constitution was not offended by the fact that they

totalled more than 50% .153

The Punjab Court has taken the opposite view. In Ramesh Chander Garg v. State of Punjab,“ medical college seats were subject to the following reservations: Nominees Nominces State of Scheduled

of Government of India of Government of Jammu and Kashmir Castes and Scheduled Tribes

Backward classes Sportsmen

6% 20%

2% 2%

Students from recognized backward areas of Punjab unable to get,admission on merit Women

10% 20%

Total Reservations

60%

151. A. I. R. 1966 Mys. 40.

152. Where the state had reserved 90% of the places in a condensed M.B.B.S. course

for licentiates in state service, the Mysore Court held that this was not a reservation but

an allotment and that Art. 15 standards did not apply (Narayana Swami v. State of Mysore,

A. 1. R. 1968 Mys. 189). 153. Sagar v. State of Andhra Pradeh, A.1.R. 1968 A.P. 165, similarly read Balaji as

dealing only with reservations under Art. 15(4) and as standing for the proposition that Art. 15(4) reservations should not exceed 50%. Reservations for women were justified separately under Art. 15(3), and the miscellaneous ones (sportsmen, etc.) did not offend

either Art. 15(1) or 29(2).

154. A. I. R. 1966 Punj. 476. The scheme became even more elaborate in succeeding

432

The Judicial Process

The Punjab High Court took the view that a reservation of 60% was “clearly hit by the test laid down by [Balaji].” 155 “[S]pecial provision is

allowed to be made for some kinds of minorities and this permission should not be used by State authorities . . . so as to reduce the seats available to candidates competing on an open merit basis to less than

even 50 per cent.’’156 In view of the interests of the meritorious candi-

dates and the national interest in utilizing talent, the Court held “that

making reservation for the weaker sections of the society under any

label whatsoever should not in the aggregate exceed 50 per cent of the

total available seats... .”157 The 60% reservation is “invalid and violative of the rule of equal protection of laws and amount{s] toa fraud

on Article 15(4) of the Constitution.” 158 Neither of these views is entirely satisfactory. The Subhashini view ignores the concern expressed in Balaji and other cases to give scope to

talent. The constriction of merit operates to the detriment of the public

and the general applicant as much where the reservation is for sportsmen as where it is for backward classes. Of course, it could be argued

that these are admissions on “merit,” with the ordinary measure of

merit broadened to include athletic ability, etc. But if these are desir-

able qualities to be encouraged among the student body, some way might be found of measuring it in all students, not just in a few, and measuring it in increments rather than on an all-or-none basis. On the other hand, the Punjab position is not convincing insofar as the categories are not truly additive but are cross-cutting. It was the reservation for women that sent the total over the 50% mark. It is not clear why this should be counted. Special provisions for women occupy a unique constitutional position. If the state had two medical colleges of equal size, one only for women and one only for men, would it be fore-

closed from making any reservation for Scheduled Castes in these colleges? Why should the result be different if the two medical colleges

are co-educational and places in them are reserved for women? Unlike

the other reservations, those for women cut across the other categories

and are not additive. This was in fact the situation in Garg. More than

20% of the successful merit candidates were women, and the reservation was not utilized at all. In fact, 251 of 400 candidates there were

admitted on the basis of merit; the effective reservation was less than years, but the reserved total dropped to-50%. Several categories (candidates from

Border areas, children of defense personnel, etc.) were added, but the reservation for

women was reduced from 20% to 1%. See Gurinder Pal Singh v. State of Punjab, A.1. R. 1974 P. & H. 125 at 126. 155. Id., at 478.

156. Id., at 478. This position was recently taken (without discussion) by the court in

Chhotey Lal v. State of U.P., A. 1. R. 1979 All. 135, 152 [NIC]. 157, Id., at 478. 158. /d., at 478.

Scope and Extent of Preferences

433

38%. So the court’s objection was highly formalistic and completely

unrelated to the extent to which merit considerations were actually

diminished in the selection. Both positions reveal the inadequacy of applying a strict percentage rule without consideration of underlying policy and the factual implications of the arrangements. The size of the reservation does not necessarily reflect its impact on the merit principle. The broader the categories for which reservation is made, the more likely they are to pick up

students who would be selected on merit. Thus large categories like “women” either may have no effect on the selection or may have a marginal effect rather than seriously detracting from the merit principle. The more the categories cut across one another, the less will a simple addition of percentages reveal the effective extent of reservations. The problem of composite reservations emphasizes the need to go beyond nominal reservations toa measurement of effective reservation. The conflict among High Courts pointed to a set of questions which

remain unanswered after the Supreme Court’s purported resolution of the matter. In Chanchala v. State of Mysore,*59 the Court encountered a

version of the Mysore admissions policy that had achieved a Byzantine complexity. The four medical colleges run by the state in connection

with its three universities had places for 765 entering students. (In addition, the state controlled 59 places in the four private medical colleges located in the state.)16° Of the total of 824 places, 60 were set aside for the following categories: students from Union territories and States where there are not medical colleges, students from relatively less developed Commonwealth countries,

cultural scholars and students under T.C.S. of the Colombo Plan and Special Commonwealth Assistance Plan, students-from Nepal, repatriates from Burma, Ceylon, Mozambique, children of Defence Personnel and Ex-Defence Personnel, students who have passed L.A.M.S. and L.U.W.S.,

lady students sufferers, and

taking family planning programme,

. . . students from Goa.161

children of political

Of the remaining places, 16? 80% were open to those who passed the Pre-University Course (“the common qualification for eligibility” to 159. Chanchala v. State of Mysore, A. 1. R. 1971 S.C. 1762. 160. At A. I. R. 1971 S. C. 1764 the Court refers to government control of 10% of 480 seats, but at pp. 1765 and 1767 the figure mentioned is 59 seats. It is not clear to what extent the distribution or reservation provisions applied to these 59 seats. 161. /d., at 1765. 162. Not 705 but 720, since the Court reports that “any one of the lady candidates can take up” one of the 15 seats set aside for the Family Planning Programme (A. I. R. 1971 S.C. at 1770).

434

The Judicial Process

medical courses) or an equivalent examination; 20%

were open to

graduates who had taken the required subjects. Forty-eight percent of these same places were reserved for backward classes (15% for SC; 3%

for ST; and 30% for Other Backward Classes). Finally, each of the three universities was supposed to allot at least 80% of its seats to students passing from its own colleges. By my calculation less than 46% of the seats remained open to “general” applicants, assuming that the backward classes and miscellaneous reservations did not overlap very sub-

stantially in practice. The Court distinguished the “university-wise”’ distribution from the district and area distributions previously struck down, finding that it related to legitimate educational objectives. It is open to the govern-

ment “to decide the sources from which admission would be made, provided of course such classification is not arbitrary and has a rational basis and a reasonable connection with the object of the rules,” When

it turns to the 60 seats for the various categories, the Court

again invokes the notion that this is not a reservation but a specification

of sources.

[T]he Government is entitled to lay down sources from which selection for

admission would be made. A provision laying down such sources is . . . not a reservation as understood by Article 15 against which objection can be taken on the ground that it is excessive.16

Apparently, the 50%

limit is specifically related to reservations for

Backward Classes under Article 15(4).

Setting apart 60 seats [for various other categories] is . . . not a reservation, but laying down sources for selection necessitated by certain overriding considerations such as obligations toward those who serve the interests of the country’s security, certain reciprocal obligations and the like.16

Since analytically a reservation is precisely that—a specification of sources from which a selection will be made—the distinction is not a

very illuminating one. A few pages further on, the Court takes up one of

these ‘“‘sources”’ (the provision for admission of “Children of Political Sufferers”). In the course of its discussion, the Court identifies the question as whether there may be “‘valid reservations, apart from those permissible under Article 15(4).”’16 It finds the ‘“‘political sufferers” provision justified by extending the principle of Article 15(4) to others who are handicapped in competition, in this case by the disruption and economic stress entailed by their parents’ involvement in the national 163.

/d., at

1769.

The

same

notion

would

seem

to fit the PUC

vs. graduates

distribution, as well. 164. Id., at 1770. The “sources” language comes in with Chitra Ghosh, discussed at nn. 143 ff., above. 165. Jd., at 1770. 166. Id., at 1775.

,

Scope and Extent of Preferences

435

struggle. The Court’s portrayal is persuasive, but it is clear that the

question is one of special entitlements for candidates whose personal characteristics make them especially deserving; it is not a question of arranging sources of recruitment to achieve any medical or educational

objectives. “[S]uch a classification has a reasonable nexus with the object

of the rules which can be nothing else than a fair and just distribution of seats.”167 The notion of single-minded devotion to the substantive policy at hand (here, medical education) has now given way to recognition that fair distribution of medical admissions is itself a legitimate

object of governmental policy.16 But this raises a host of questions,

which are only obscured by the formula that “a provision laying down a

source is not a reservation.”

Two questions emerge of profound import for policies of compensatory discrimination for the backward classes. First, what are the other legitimate grounds for government preferment (other than performance related to the matter at hand)? That is, what conditions of inequality and unfairness may government attempt to remedy? Is there any quantitative limit on the amount of such preferment? Specifically, when such preferential treatment takes the form of reservations in scarce facilities, is there any quantitative limit on the cumulative total of preferential treatment?

The

Full Bench decision in Jagdish Rai dramatizes the enlarged

doctrinal possibilities for preferential treatment supplied by Thomas and Chanchala. Reservation of 28% of posts for ex-servicemen was upheld on the grounds that Thomas had introduced a new dynamic concept of equality of opportunity which permitted application of general princi-

ples of classification under Article 16(1). Just as Chanchala permitted

preferential treatment for children of political sufferers who are handicapped but did not fall under Article 15(4), so it is “fair and equitable that a just proportion of posts” be given to ex-servicemen who suffer

peculiar handicaps in competing for jobs.17° For the State to “classify

167. Id., at 1776. 168. The narrower view is still alive. Thus the Court in Pradip Tandon says the only object of selection for medical college is to choose the “best possible student” (A. I. R. 1975S. C. 563 at 570). And Justice Khanna, dissenting in Thomas, A. 1.R. 1976 S.C. 490 at 508, adheres to the view that the only legitimate object of rules regarding recruitment is “to ensure efficiency of service.” 169. State of U.P. v. Pradip Tandon, A. 1. R. 1975S. C. 563 at 571. Provision ofa source is not a reservation. What is and what is not a “source” remains unfathomed, but some

categories are not. In Pradip Tandon the Court finds that “rural area in Uttar Pradesh cannot be said to be a source for reservation of the type in Chitra Ghosh’s case” (A. I. R.

1975 S. C. 563, 570).

170. Jagdish Rajv. State of Haryana, A. 1. R. 1977 P & H 56, 61. That the shift is 180° is

confirmed by the judgment’s words of commendation for the minority view in Sukh-

nandan (id. at 60).

436

The Judicial Process

them separately as a source of recruitment and reserving posts for

them” is “an extension of the principle of Art. 16(4) to those that do not fall under Article 16(4).”!71 Supplied with such abundant doctrinal

resources for justifying preferential treatment of virtually any object of governmental solicitude, is there any limit to the expansion of the compensatory principle?

Expansion of the compensatory principle to a sovereign remedy for

any falling short of the substantive equality promised by Articles 14-16

equips the government with an authorization so vast that the commit-

ment to the lowest social groups represented by Articles 15(4), 16(4), 46, 335, etc., is in danger of being overwhelmed by competing commit-

ments. In effect, any other groups thought to be the victims of “unde-

served want” are now entitled to have the reservation device used on their behalf, too. New competitorsfor special reatment will proliferate}72 and public patience with the use of reservations may be exhausted. These new categories cannot use the classifications forbidden by Articles 15(1) and 16(2), but the beneficiaries under these various

headings are overwhelmingly drawn from relatively advantaged strata

of society. They are the ‘temporarily or situationally disadvantaged

members of those sections of society which enjoy social, educational,

and economic advantages. Now all the kinds of disadvantage which afflict the better-off—diplomats, central government employees, retired army officers, physically handicapped children in well-to-do

families—are candidates for preferential treatment along with the

backward classes. It is doubtful that, in a setting of chronic shortage, the backward classes will fare as well where government is thought to have a general obligation to produce substantive equality as in a regime of formal equality, qualified by a single major exception for measures to overcome the disparities associated with the traditional social

hierarchy.?73__ Justice Krishna Iyer, in the course of his contribution to enlarging the notion of equality of opportunity in Thomas, discerned the potential for diffusion of the commitment to the backward classes and attempted to provide a formula for circumscribing the compensatory principle:

171. Id, atl. 172. Some of the categories commonly employed can be seen as attempts to exempt favored groups from the effects of restrictive eligibility requirements, adopted to appease localist sentiments, without discarding those requirements. As local restrictive-

ness grows (see Katzenstein 1979) so do provisions for out-of-state categories. 173. Before we can tell the net effect of these other categories of preference, we have to know whether they will operate as guaranteed minimums or on an “over and above” basis (see chap. 6, §B). If they are guaranteed minimums and the beneficiaries are from more or less advantaged groups, there may be little net effect; these reservations would be little more than a device for allowing authorities to take credit for admissions that would have been made anyway.

Scope and Extent of Preferences

437

The distinction would seem to be between handicaps imposed accidentally by nature and those resulting from societal arrangements such as caste structures and group suppression. Society being, in the broad sense, responsible for these latter conditions, it also has the duty to regard them as

relevant differences among men and to compensate for them whenever they operate to prevent equal access to basic, minimal advantages enjoyed by other citizens.174 :

The initial attractiveness of this formula is soon dissipated by reflection. Putting aside the question of whether medical admissions or high government posts should be distributed according to the same principles as “basic minimal advantages,” how sturdy is the asserted distinction between natural and social handicaps? Natural and social categories overlap: which, for example, is “woman”? or “handicapped”? Beyond this, why is compensation thought appropriate to offset conditions “resulting from social arrangements” but not those “imposed .. . by nature”? What does it mean to be “responsible” for former conditions? It was not the existing State that imposed them. Is it responsible for conditions imposed, supported, or tolerated by its predecessors? Or because it has powen to alter them and undo their results? But the State has the power to overcome “‘natural” conditions as well. Does it not have comparable responsibilities for failing to do so?!75 For example, is an accidental injury (or a form of blindness) that could have been prevented by known measures any less a matter of societal responsibility than ‘caste structures”? The former may be considerably more tractable to government action. If this attempt to erect a coherent theory of compensation proves wanting, it should not obscure the fact that the Constitution and a quarter-century of attendant policy-making do embody a sustained commitment to offset a specific historic constellation of inequalities implicated with the traditional social hierarchy in India. To generalize from this commitment may diffuse rather than strengthen it. E. THE NUMBER OF BENEFICIARIES Judicial control over the extent of reservations leads back to the

question of whom the State may designate as backward. Does a limit on the percentage of places that may be reserved imply a limit on the percentage of the population for whom they may be reserved? There is no explicit constitutional limitation on the number of persons that may constitute the backward classes. However, independently of any limita174. State of Kerala v. N. M. Thomas, A. 1. R. 1976 S. C. 490 at 538. 175. Cf. Calabresi and Bobbitt 1978.

438

The Judicial Process

tion on the percentage of places reserved, the courts have inferred some quantitative limit on the percentage of the population designated as beneficiaries. In the Kesava Iyengar case, the court did not object to the inclusion among the backward of over 95% of the population of the state.176 However, in Ramakrishna Singh v. State of Mysore, another bench

of the same court held invalid a scheme for the same list of beneficiaries on the ground that designation of 95% of the population as Backward

Classes ‘“‘is a fraud on the Constitution.” !77 Even in the absence of any

fraudulent intent, the use of the power bestowed by Article 15(4) in

order to benefit others than the backward was invalid. Finding that the Constitution did not intend that this power be used to benefit all but a few,

the

court

held

the scheme

an

unconstitutional

discrimination

against the excluded 5%. Similarly, fifteen years later, a Full Bench of the Allahabad High Court thought that to call backward the 80% of

Uttar Pradesh’s population residing in rural areas “would . . . be an appalling opinion, staggering to one’s imagination.”?”® The Supreme

Court, affirming on this point, thought this amounted to “‘a case of discrimination in favour of the majority of rural population.”179 In both

of these instances the putative beneficiaries were found not to be legitimate objects of preferential treatment, so the courts did not reach the question of numerical limits on the number of legitimate

beneficiaries.

Although the courts have recoiled from 80 and 90% figures, reservations for groups ranging from over 30 to over 50% have passed muster on several occasions without exciting judicial hostility. When some 20% for Scheduled Castes and Tribes are added, this means that bene-

ficiaries of preferential treatment are a clear majority of the population. No limit corresponding to the widely accepted 50% ceiling on reserva-

tions has even been hinted at, although existing doctrine provides some starting points for such an argument.

In the Balaji case, the Supreme Court did not address itself explicitly

to the number of beneficiaries clear that the Court’s strictures merely “less advanced” or only number of beneficiaries was an 176. A. I. R. 1956 Mys. 20. 178. A. 1. R. 1975 All. 1, at 7.

of the Mysore scheme. However, it is against preferences for those who were “relatively backward” indicate that the important point. Several of the Court’s 177. A. I. R. 1960 Mys. 338 at 348. 179. A. I. R. 1975 S. C. 563, at 571.

180. Thus in Balaram the Court upheld Backward Classes lists estimated to consist of

38% of the population of Andhra Pradesh (A. I. R. 1972S. C. 1375). In fanki Court approved of almost all of a list which apparently consisted of 46% of population (A. I. R. 1973 S. C. 930). In Periakuruppan, the Court upheld a Classes list supposed to contain 51% of the population of Tamil Nadu (A. I. C. 2303).

Prasad, the the state’s Backward R. 1971S.

Scope and Extent of Preferences

439

tests of backwardness imply severe quantitative limitations. The Court

holds that ‘only communities which are well below the state average. . .

can properly be regarded as educationally backward classes of citizens.”’181 The “state average” in question is the mean, and presumably

considerably less than half of the population, could qualify as ‘‘well below” the mean. Second, the Court limits those who may legitimately

be designated as backward to those whose “backwardness [is] compar-

able to Scheduled Castes and Scheduled Tribes.”’*? Such a standard, if

applied to any tangible criteria ofJiteracy, income, etc., would severely

restrict the number of possible beneficiaries. (The Balaji Court else-

where endorses an economic test of backwardness, but it is not clear

whether it means the “well below the state average” or the “comparable to the Scheduled Tribes and Castes” standards to specify the level at which an economic test must be set.) The relationship between the percentage of seats reserved and the percentage of the population who enjoy the reservation is a puzzling one. It seems to be generally assumed that a reservation should be at about the same level as the representation of that group in the population—thus, for example, seats or posts are normally reserved for the Scheduled Castes and Scheduled Tribes roughly in accord with

their percentage in the population. In Pandit v. State,1®3 the Bombay

High Court found the beneficiaries percentage of the state’s population “an objective and just test for determining the proportion of seats to be reserved in the medical colleges.” There is ‘perhaps . . . no better

basis for such reservation than the proportion of the population of the Backward

Classes

to the whole

population

of the State.”?85 Such

181. A. I. R. 1963 S. C. 649 at 660 (emphasis added). 182. Jd., at 668. However, in the same paragraph the Court observed that “{i]t was

realized that . . . there were other classes of citizens who were equally, or maybe somewhat

less, backward than the Scheduled Castes and Scheduled Tribes and it was thought that some special provision ought to be made even for them” (emphasis added).

183. A. I. R. 1972 Bom. 242.

184. Jd., at 250. The reservation in question operated as a guaranteed minimum. The obvious “representative” thrust of the Court’s argument may not hold in cases where the reservation is of places over and above those acquired by merit. See chap. 13, §B,

below. 185. Id., at 251. This proportionality theme provides the basis for Justice Faz Ali’s

argument that the 50% limit on reservations is not a definite limit but merely “‘a rule of caution”: if the government ofa state with “backward classes of citizens which constitute 80% of the population . . . reserves 80% of the jobs for them, can it be said that the

percentage of reservation .. . violates the permissible limits of . . . [Art. 16(4)]? The

answer must necessarily be in the negative,” since the object of this provisign is to secure adequacy of representation (State of Kerala v. N. M. Thomas, A. 1. R. 1976 S. C. 490 at

555).

440 The Judicial Process statewide ratios may be used even when the population in the catchment area ofa particular scheme of benefits may vary from it.

Many reservations for the Backward Classes are much smaller than the population percentage of those classes. Courts have responded differently to this discrepency. In Jacob Mathew the fact that the reservation was smaller than the population of the group impressed the judge as evidence that the selection was a prize of caste politics rather than a genuine attempt to assist the backward. On the other hand, in Vishwanath the Mysore court found the fact that 30% of the seats were reserved for

50% of the population a warrant of the reasonableness and moderation

of the measure. 187 Since, as we have seen, there is no necessary relation-

ship between the nominal size ofa reservation and its effect, it might be doubted whether either presence or absence of commensurateness to population ratios is in itself an indication of reasonableness. Furthermore, quite apart from the guesswork by which they are estimated, population figures for backward classes may be misleading characterizations of the number of beneficiaries. The population of backward classes and the beneficiaries of any given scheme cannot be equated. The dealing with a to First Form population in

relationship between them scheme for providing health schoolchildren, we might the backward classes as a

is problematical. If we are care or free books and slates take the percentage of the rough estimate of the per-

centage that might benefit from this scheme. But in what we have called

higher-echelon benefits, those which require previous training and accomplishment, there may in fact be few members of the backward classes in the pool of persons eligible to compete for a benefit such as a

high government post or a medical college admission. Even if 60% of the population are members of the backward classes, it may be that only

30% of the government officers eligible for promotion are from these

groups.

Symmetry between the percentage of places reserved and the percentage of the population who are backward may produce some curious

distortions. When effective competitors from these groups are few, to confine the backward population to the percentage of places reserved

might diminish the pool of competitors. Conversely, if because of population figures, the reservation is set high relative to the number effectively eligible, there may be a windfall pile-up of benefits on a few

fortunate individuals. 1°

But the abandonment of symmetry exposes the real problems of a

186. A. I. R. 1964 Ker. 39 at 61. 187. A. I. R. 1964 Mys. 132 at 137. 188. Cf. thie situation described in Harpartap Singh v. Union of India, 1970 S. L. R. 40; see also State of Punjab v. Hira Lal, A. 1. R. 1971 S.C. 1777.

Scope and Extent of Preferences

441

limit on the number of beneficiaries. First, as the beneficiary class approaches the total population there arises the question of whether the administration of eligibility requirements is justifiable. Is it worth its cost in money, energy, and attention? And, as a scheme becomes more

inclusive, at some point it takes on the character of singling out the omitted groups for unfavorable treatment rather than conferring special treatment on those who are included. Schemes conferring benefits on majorities may be hard to terminate or reduce.

The non-included are not the only victims of expansive lists of

Backward Classes. As the backward category expands, benefits may be diluted. Since many benefits are distributed among the backward on a merit basis inter se, the least prepared and advantaged will suffer in

merit competition with more favored groups. Large numbers of Backward Classes will provide political support for measures for these

groups, so category. backward of the less

that the most backward may benefit from enlargement of this But unless there is some internal sub-division, the most groups are likely to be squeezed out by the superior “merit” backward as well as by their superior access to officials and

rograms. Who is hurt by enlargement of the Backward Classes category will

depend in part on whether preferences are distributed on a guaranteed minimum or an “over and above”’ basis.1®9 Ifa reservation is merely a guaranteed minimum, enlarging the Backward Classes category will presumably shift places from the most backward groups to the best

prepared among the Backward Classes. On the other hand, if reserva-

tions are places over and above those gained by merit, enlargement of the backward category will increase the effect of reservation at the expense of non-beneficiaries as well as squeezing out the most backward. In either case, the most backward lose.

F. BENEFITS OTHER THAN RESERVATIONS The State is not limited to “reservations” as a method of preference. Only in the area of government employment is “reservation” constitu-

tionally specified, but even there reservations may be accompanied by

other preferential devices—fee concessions, waiver of age limit, etc. The Balaji case disposed of the notion that a limitation on the extent of reservations must rest upon the construction of the word “‘reservations” in Article 16(4) and established that such limits are applicable to 189. See chap. 13, §B, below.

442

The Judicial Process

reservations under Article 15(4) as well. The question arises whether

there are similar constitutional limitations on the quantity of other

kinds of preferences. Presumably, “reservations” would be covered wherever used. But the national interests which made the limitation imperative in the professional and technical colleges or in promotions in government service may not be present to require a similar limitation on the reservation device in other schemes. A percentage limitation of the sort applied to reservations would have

little meaning with respect to other kinds of benefits authorized by Article 15(4), such as expenditures for housing and education or protection against foreclosure for debt.19° The State is entitled to provide such things as housing facilities for the exclusive use of Scheduled Castes. 191 Can it not do the same for other backward groups and with other kinds of facilities? May it not have a health scheme for Scheduled Castes only or an agricultural program for Scheduled Tribes only? Presumably, the State may do any of these things. Is the State then limited in the percentage of expenditures for, e.g., housing, that it may devote to such purposes? Perhaps the notion that provisions for the advancement of the backward must be balanced against broader national interests

might be applied here. But the broad national interests in the limitation

of expenditure for the backward are more difficult to identify. In the case of government employment or admissions to medical colleges, there is a public interest in some share of these scarce resources going to the ‘“thaves” because, presumably, they will be more productive and

efficient, bestowing more benefit on the public. In addition there are the

guaranteed and enforceable fundamental rights of the merit competitors in these areas. But expenditures for such matters as housing and health do not involve specific and enforceable rights of others. Nor is it clear that there are parallel public benefits in the rewarding of the advantaged. There is not necessarily any accretion of public benefit from giving housing subsidies to those who already have good housing. Ifthe Devadasan majority’s notion that opportunity must be equalized annually were imported into this area, it would make impossible large-scale diversion of resources to meet the problems of backward groups. Unlike legislative seats, government posts, and medical college admissions, these welfare areas do not involve values of which government is the exclusive or prime dispenser. Valued resources in health,

190. Of course, it may prove a useful instrument of policy to set percentage goals for

expenditures. For example, the Karnataka Government recently instructed municipal bodies to earmark 18% of their annual budgets for programs for welfare of the Scheduled Castes and other weaker sections (Times of India. Aug. 22, 1977, p. 6). 191. AMoosa v. State of Kerala, \. 1. R. 1960 Ker. 335 Cf. Jagwant Kaur v. State of Bombay.

ALT. R. 1952 Bom. 461.

Scope and Extent of Preferences

443

housing, land, etc., lie largely in private hands. Government's role in these fields is not that of the major distributor of scarce opportunities, but rather that ofa countervailing re-distributor. There would seem to be no constitutional bar to prevent government from channelling all of its expenditure on, e.g., housing, to the backward classes.

The case for judicial control over the extent of such preferences is far

less clear than over reservations.

For schemes of reservation directly

involve the guaranteed and enforceable fundamental rights of other

citizens. But in the case of expenditures for such matters as housing and

health, there are no tangible rights of others involved, and the courts would have only a vague national interest to guide them—or at most the non-justiciable rights of other citizens contained in the Directive

Principles of State Policy. It seems clear that in making these non-

justiciable the purpose was to keep the details of policy involved in development out of the courts. Any detailed judicial review would inject the judiciary into an area in which both its authority and its competence

are not clear.

13

The Design and Operation of Compensatory Discrimination Programs: Constitutional Limits and Judicial Control A. RULEFULNESS AND DISCRETION We nave seen that preferential treatment may be initiated by Centre or states, by the executive as well as by legislative bodies. Courts have been vigilant to insure that reservations and other special provisions are made in a ruleful manner and by competent authorities.* A state may not employ a scheme of reservations where there has been no government order specifying the backward classes or fixing the proportion of places reserved for them.? Nor may a state reserve places in medical school admissions where, at the time of selection, no special

provision for this had been made by either legislative enactment or executive order.3 Nor may a state appoint Scheduled Castes or Back1, The recent decision in Sharda v. Visveswariah Engineering College, A. 1. R. 1978 Kar.~ 111 [NIC], departs strikingly from the requirement of rulefulness which imbues the

cases cited below. The university had permitted a Backward Class student to transfer from an electrical engineering to an electronics program in preference to other would-be transferees with higher marks. The single judge held that in the absence of any express order permitting such preferences, the university’s course might have been illegal, but since it was fair and substantially just, this was nota fit case for the exercise of the court's extraordinary power to issue a writ.

2. Triloki Nath Tiku v. State of Jammu and Kashmir [1], A. 1. R. 1967 S.C. 1283.

3. Lalita Shuri Tikku v. State of Jammu and Kashmir, A.1.R. 1966 J. & K. 101: “there must be some order or notification to show that state has applied its mind” (/d., at 104). Some formality must be observed, although the Supreme Court has left open the question whether formal publication is required to make an order effective (Triloki Nath

Tiku v. State of Jammu and Kashmir {11}, A.1.R. 1969S. C. | at 3). Where a valid list exists,

a particular community may not be omitted because of administrative oversight without express modification of the order “by a proper decision of the government” (Susila v. State of Madras, A. 1. R. 1970 Mad. 399, 400).

Design and Operation of Programs

445

ward Classes to posts in contravention of its own rules of recruitment, by invoking a broad power of the Public Service Commission to consider the case of any candidate with the prescribed qualifications, to the disadvantage of more-qualified candidates.* Similarly, preferential treatment, when properly authorized, must be implemented rulefully. Notice must be given to applicants that a selection is subject to reservations: implementation of an otherwise valid reservation (for women, in this case) was set aside where the announce-

ment of the Selection Board led applicants to rely on a merit competition unencumbered by reservations. A principal of a medical college may not decide that unfilled seats reserved for Scheduled Castes should be filled with members of Backward Classes rather than revert to the general pool, where this “carry-over” arrangement is not authorized by statute or rule.6 Nor may the State overlook merit within its specified

reservations: the State may not select Scheduled Castes candidates

ranking lower in the competition over those who rank higher.”

The Devadasan notion of limiting the future effects of unfilled reservations has been extended to a more general principle that reservations may not operate with retrospective effect ® on the interests of individual (non-beneficiary) appointees.? Thus it has been held that the State

might not withdraw a post from an employee already appointed in order to give it toa Scheduled Caste—even where the seat was originally reserved

and

was

given

to the non-Scheduled

Caste employee

because there was no applicant eligible for the reservation.1° 4. Channabasaviah v, State of Mysore, A. 1. R. 1965S. C. 1293. 5. Abodha Kumar v. State of Orissa, A.1.R. 1969 Or. 80.

6. Abdul Aziz v. State of Maharashtra, High Court at Bombay, Special Civil Application

No. 1363 of 1962, K. K. Desai and Chandrachud, J.J., decided 12 Oct. 1962; Mohd. Abdul Mabood Khan v. State of Maharashtra, High Court at Bombay, Special Civil Application

Nos.

1802 and

1829 of 1962, Tambe

and

Palekar, JJ. decided

19 Feb.

1963. 7. Lalita Shuri Tikku v. State of Jammu and Kashmir, A. 1. R. 1966J. & K. 101 at 105; Kushma Joshi v. Pro-Vice-Chancellor, A. 1. R. 1969J. & K. 136 at 139. 8. Similarly, a debtor whose caste was added to the Scheduled Caste list by the 1979

amendment of the lists was not entitled to benefit from debt relief provisions applicable to loans subsisting at an earlier point in time (Mangilal v. Mittilal-Radhaylal Rastogi, A. 1.

R. 1978 M.P. 160 [NIC}).

9. The vast corpus of case law on service matters contains indications that it is open to government to use the seniority provisions to offset temporary delays in filling reservations. Where reservations go unfilled, the government may assign seniority according to its rotational system, so those who belatedly take the reserved places take seniority over those who were hired (or promoted) because of the temporary lack of persons to fill the reserved places (Chauhan v. State of Gujarat, A. 1. R. 1977 S. C. 251, 263). 10. In Sudama Prasad v. Divisional Superintendent. A.1.R. 1965 Raj. 109. a railroad

employee appointed as Chief Clerk obtained an order setting aside the order of the

railway authorities who had demoted him in order to replace him with a less senior

446

The Judicial Process

But beneficiaries of reservations who are disadvantaged by retrospective changes in reservation policy have had little success in securing

judicial relief. A Scheduled Caste typist who lost his chance for promo-

tion when the basis for computing seniority was changed was held to have no vested right to promotion on the original basis.11 A member ofa Backward Class promoted to a temporary post in accordance with a state order reserving promotion posts might be reverted to his original post when the state government reversed its policy and withdrew reservations from Backward Classes. This case involved a, temporary post, and the original order was unconstitutional and therefore void.12 It is not clear whether the rule would apply to permanent appointments or to appointments pursuant to orders which are constitutionally valid.

But it has been applied to a government servant who passed up a de-

sirable selection post on “merit” to take a more desirable promotion

post reserved for Backward Classes. When the government reversed its policy, he was reverted out of the promotion post; the other post had in

the meantime been filled and he lost out on both.

While the courts have required regularity in the designation of backward classes and the authorization of reservations, they have moved slowly in controlling broad personal discretion at the later stages of the selection process. The operation of selection committees provides a vivid example. In several states, the portion of the competitive score for college admissions allotted to interview by a selection committee is so substantial that it creates the potential for what is in effect a covert

reservation.

employee who had belatedly established that he was a member ofa Scheduled Caste. But cf. Mahendra Nath Pathak v. State of Assam, A. 1. R. 1970 A. & N. 32. The original

settlement was restored when the putative Scheduled Tribe member failed to establish his tribal membership.

11. Pramanick v. Union of India, A. 1. R. 1969 Cal. 576. The basis for computing

seniority was changed from date of comfirmation to “merit as determined by Public Service Commission or results of training tests.” Scheduled Castes had gotten confirmation earlier than other employees. Even though the court says there was no right to complain ofa change, it repels the claim of discrimination by observing that “no reservation or special provision in favor of members of the Scheduled Castes can be carried to the length of impairing the efficiency of the administration” (id., 577). It is not clear whether the same result would have been reached had the change of bases been in the opposite direction.

12. Dasa Rayudu v. Andhra Pradesh Public Service Commission, A. 1. R. 1967 A.P. 353.

13. G. Satyanandhan and others v. State of Andhra Pradesh, etc. Writ Petition Nos. 278, 1352,

etc., of 1964. High Court of Andhra Pradesh, Basi Reddy and Gopal Rao Ekbote, JJ.. 8

March 1965. Such cases as these were numerous in Andhra when the government withdrew its Backward Class list and reservations in promotions. Some departments permitted those promoted when the reservation was in effect to retain their new posts, others reverted them. The Supreme Court denied special leave to appeal on this point.

Design and Operation of Programs

447

Selection committees were a favorite device in pre-Independence

Madras for assuring the proper communal balance.!* When the Madras system of outright communal quotas was struck down by the

Supreme Court in 1951, the new modest reservations that replaced them were accompanied bya revival of selection committees. Committees consisting of local notables, presided over by the Vice-Chancellors, were empowered to take into consideration physical fitness, sports activities, special aptitude, and the backwardness of the applicant’s family in the matter of higher education. In their original incarnation these procedures “had the intended effect of reducing drastically the

proportion of Brahmin students.”15 An enthusiastic defender of selec-

tion committees leaves little doubt that this remains a major element in

their purpose: The

sole criterion of highest

Brahmins,

has gone.

Physical

merit

based

on marks,

so beneficial

fitness, special aptitude, and

to

educational

backwardness of the family, will all operate against the Brahmins. The test of physical fitness for once will eliminate puny creatures with no personality from becoming engineers and doctors in the future, and this is a step in the

right direction.'®

The potential for communal bias in the operation of the Madras selec-

tion committees is greatly enhanced by the maintenance of an entry for caste in the Secondary School Leaving Certificate books which must be

supplied by applicants.17

The combination of modest reservations and use of selection committees in Madras had the effect of discouraging litigation, so that the Madras system of selection long remained unchallenged. When writ petitions were finally brought in the late 1960s, the first one to challenge the use of selection committees was P. Rajendran v. State of Madras.1® Three-man selection committees were to award up to 75 marks (out of 300) for extra-curricular activities under five heads. The three members

did not sit together, but each interviewed each candidate separately 14. “When the Justice Party was in power they appointed Selection Committees for admitting students into the Colleges, not being able to trust the Principals, who were mostly Brahmins to do justice” (Dharmalingam 1951: 79). The committees flourished from 1922 until they were abolished in 1938 (Irschick, 1977). See Irschick 1969: 238 ff.

Selection boards are credited by knowledgeable observers with effecting the virtual

exclusion of Brahmins from schoolteaching in Maharashtra. (Interview, Poona, 1966.)

15. Rama Rao 1967: 74.

16. Dharmalingam 1951: 82.

17, Rama Rao (1967: 74) reports that after Independence the Madras Goyernment

“thas felt bound to abandon requirement of information of caste in application forms,

[but] the S. S. L. C. [Secondly School Leaving Certificate] books are being maintained

as previously, offering an unfailing clue about the caste of all candidates”

18. A. I. R. 1968S. C. 1012.

448 The Judicial Process and gave up to 25 of the total of 75 marks. After the selection was over

the mark-sheets were destroyed. The Supreme Court, striking down the scheme on other grounds, found both the individual interviews and the destruction of the mark-sheets decidedly “odd.”!9 Nevertheless, the

Court found the rule “indicates what matters have to be taken into

consideration” and was

not prepared to accept that the Committee did not follow the criterion

indicated in Rule 10(d) in allotting the marks provided in that rule. . ..We

do not think that it is possible to provide any further guidance in the matter

and the rest must be left to the Selection Committee.2°

The selection committee problem had been more fully considered in earlier Mysore litigation.24 In 1963, when the Mysore Government abandoned its community-wise list of backward classes and switched to

an occupational-cum-means test, it suddenly exhibited an enhanced

concern for the extra-curricular accomplishments of applicants to pro-

fessional colleges.22 The State decided that the interview marks should comprise 25% (75 marks) of the maximum marks on the examination in

the optional subjects.23 The two selection committees, each operating over half the state, were authorized to allot marks for interview having

PP

EN

regard to...

General knowledge. Aptitude and personality. Previous academic career, including special distinctions. etc. N.C.C. [National Cadet Corps], A.C.C. [Auxillary Cadet Corps], etc. Extra

curricular

dramatics etc.

activities

including sports,

social service, debating,

According to the Committee’s chairman, each “interview” generally took two or three minutes: generally 3 to 4 questions were put to each candidate to test his general

knowledge, aptitude, the practical application of scientific knowledge to 19. Id., at 1017.

20. Id.

21. Chitralekha v. State of Mysore, A.1.R. 1964S. C. 1823, aff'g Viswanath v. Government of

Mysore, A. 1. R. 1964 Mys. 132.

22. Previously, selectioris committees who interviewed candidates were to award a maximum of 20 marks (out of 300): “A maximum of 20 marks is fixed as interview marks to be awarded by the selection committee for assessing the suitability of the candidate for engineering profession, having regard to the following factors: (i) aptitude, (ii) general ability and physical fitness, (iii) sports or N. C. C. activities or scouting or other similar activities, (iv) extra curricular activities such as social service etc’ (quoted at A.

1. R. 1964 Mys. 142).

23. A. 1. R. 1964 Mys. at 141.

24. Id.

Design and Operation of Programs

449

everyday lite, and mental alertness and ability to express himself clearly, the last two factors being relevant for the criterion of personality. While each candidate was interviewed, one of the members of the Selection Committee was examining the Certificates produced by candidates in support of their extra-curricular activities like N.C.C. or A.C.C., sports, debating, drama-

tics and social services.25

The committee’s allotment of marks on this basis was attacked as an arbitrary, unguided, and uncontrolled power. The High Court in Vis-

wanath v. Government of Mysore denied that is was unguided power, for in the

absence of specific allocation it must be presumed that the government intended that “‘each one of those heads should carry 1/5th of the ‘Interview’ marks.”26 Since the Committee had “acted under the impression that 25% of the marks were left in their hands to be given in any matter

they considered best,” they did not deal with the several heads separately and gave only consolidated marks.

Instances were brought to our notice where ... applicants who were proved to possess at least 3 out of the 5 tests . . . have secured . . . 10 marks

or even less out of 75 marks.27

This failure to use “objective standards” was a misuse of the power conferred on them.?8 The court quashed the interview and directed the selection committees to interview petitioners afresh in accordance with the directions. On appeal to the Supreme Court it was argued that selection by

interview is repugnant to the doctrine of equality embodied in Article 14, because it introduces a strong subjective factor into selection, giving ample room for discrimination and manipulation.29 The Court conceded that manipulation was possible in interviews as in other selection procedures, but that it is capable of abuse is not a ground for quashing it. So long as the order lays down relevant objective criteria and entrusts the business of selection to qualified persons, this Court cannot obviously have any say in the matter. In this case the criteria laid down by the Government are certainly

relevant... 2° 25. Id., at 143.

26. Id., at 144.

27. Id., at 144.

28. Id.

29. The argument was also made that giving weight to the interview was changing or raising the qualifications and therefore outside the power of the State, since the co-ordination and determination of standards ofa University is a subject reserved to the Union by VII Schedule, List 1, Entry 66 of the Constitution. The courts have indicated that the fixing of standards was pre-empted from State authority only where the State’s action would make impossible or difficult the exercise of Parliament’s power. Here, there was no such encroachment (A. I. R. 1964 S.C. at 1839-40).

30. A. I. R. 1964S. C. at 1831.

450

The Judicial Process

The Court seemed to leave the door open for.a further look when it concluded: We

cannot,

therefore,

hold

without

better and

more

scientific material

placed before us that selection by interview in addition to . . . the written examination is itself bad as offending Article 14. . . 34

But the Court took no notice of the disparity between the criteria and the time in which qualified persons could possibly reach such complex judgments. In a later Mysore case, counsel convinced the High Court that the interview methods employed were “highly rudimentary,” but the very disparity between the immensity of the committee’s task and

the paucity of its resources convinced the court that under the circum-

stances the interviews were satisfactory.5? Subsequently, in Jammu and Kashmir, a Full Bench encountered an even more flagrant use of interviews.33 Medical school admissions were distributed entirely on the basis of marks given by a committee (consisting of two members of the Public Service Commission) which in-

terviewed each student for two to three minutes. The committee in-

terviewed only those who had obtained aggregate marks of 45%. It gave

no weight to the academic accomplishment of those who came over this threshold. The committee avowed that it allotted marks as follows:

physical fitness, 10; aptitude,

10; general knowledge, 40; personality,

40. In what is surely the ultimate tribute to the political primacy of

medical

school

admissions,

the committee’s

recommendations

were

submitted to the Chief Minister, who issued the order of admission.

The High Court found that it was well-nigh impossible for the

committee to assess merit objectively within such a short time; that its

failure to take academic merit into consideration made the selection

patently unreasonable, arbitrary, and subjective; that ignoring merit was “inherently abhorrent to the notions of a civilized and progressive state of society.”’>* The court politely concluded that what the committee did was not what government intended it to do (i.e., to give due weight to academic merit along with interview marks) and struck down the admissions.35 31. 32. 33. 34. 35.

Id., at 1832. Subhashini v. State of Mysore, A. 1. R. 1966 Mys. 40 at 47. Subhash Mohan Jalali v. Principal, Medical College, A. 1. R. 1967J. & K. 106. Id., at 110. The Full Bench spelled out the acceptable ways in which such a selection might be

made. These included (1) use of aggregate marks at the first qualifying examination,

with or without

an interview;

(2) adding

interview marks

to the marks in the last

qualifying examination; (3) interviewing only candidates with high scores at the last qualifying examination and allotting marks at the interview, giving due weight and consideration to “the academic merit of the candidates as revealed by their perform-

Design and Operation of Programs

451

When the selection committee issue returned to the Supreme Court in Periakaruppan v. State of Tamil Nadu,>6 the bench included the judge

who had written the High Court judgment in Viswanath. The state had set up regional selection committees to interview applicants to medical college. Of the total of 275 marks, 75 were to be awarded on the basis of the following tests: (1) (2) (3) (4)

Sports or National Cadet Corps Activities; Extra Curricular special services; General Physical condition and endurance; General ability and

(5) Aptitude. The committees were composed mostly of medical practitioners, and the interviews lasted roughly three minutes for each candidate. The Court reflected that 75 out of 275 marks seemed excessive for a threeminute interview, but felt itself unable to hold it beyond the power of the government to provide such a high portion.37 Although there was

“disturbing” discrepancy between interview marks and other measure-

ments of performance, the Court declined to conclude that either the committees or the government were manipulating the interview

marks.38 Nor was it fatal that no objective criteria were established: the ances in the last qualifying examination” (id., at 109-10). When it was later charged

that these acceptable modes of selection had not been followed, the High Court called

for all the relevant records of the Selection Committee and found that “‘by and large candidates have been selected purely in order of merit by adding the marks secured by them at the interview to the marks fetched by them at the last qualifying examination.” The documents revealed that “the Selection Committee did not try to plump up marks

ofa particular candidate so as to tilt the scale of one candidate in favour of another. The

marks awarded . . . appear to be fair and reasonable and are fully consistent with the aggregate fetched by the candidates in the last qualifying examination” (Sardool Singh v.

Medical College, A. 1. R. 1970 J. & K. 45 at 47). It appears that once tamed into tulefulness, selection committees with only a few minutes to spend can do little that makes the outcomes very different than those produced by the examinations.

36. A. I. R. 1971 S.C. 2303.

37. Some courts have been unperturbed by the perfunctory character of interviews.

In Daya Ram v. State of Haryana, A. 1. R. 1974 P. & H. 279, 280, petitioners claimed that

in the selection of sub-inspectors the interview was an empty formality: “The candidates entered through one door and were made to get out from the other. No question whatsoever was put to them.” The selection board denied this and said that “only 253 candidates were interviewed on the said date between 7: 00 A.M. and 3: 00 P.M.” Assuming that the board worked continuously, this would leave 1.7 minutes per

candidate.

38 A. I. R. 1971 S. C., at 2307-8. “[N]umerous students whose performance in the University examinations was none too satisfactory nor their past records creditable . . . secured very high marks at the interview. It is also true that a large number of students who had secured very high marks in the University examination and who performed very well in the earlier classes had secured low marks at the interview” (id., at 2307).

452

The Judicial Process

Court found the specification of the five headings “‘sufficiently objective in character.”39 However, even though the rule did not prescribe separate marks for separate heads, this could not be read as permitting the committee to allot marks as it pleased. The Court found the failure

to divide the interview marks under the various heads and giving them “Gn a lump” was clearly illegal.4° In addition, the committee departed from the rules by taking into consideration irrelevant matters and failing to consider those required

by the rules. Thus under N.C.C. training the committee judged candidates not on their records as cadets but on the response to questions

about the role of N.C.C. in national life. And physical condition and

endurance were not judged by medical examination: “It was in the nature of things not possible for the selections committee, though

composed of eminent doctors, to find out the physical condition and

endurance by a mere look at the candidate.”’*1 From the facts available,

the Court concludes that “it is clear that the candidates were not interviewed in accordance with the rules governing the interview.”"42

Pursuant to the Court’s direction, the state constituted a new selection committee to fill the remaining 24 unfilled seats, but again the petitioner failed to be selected, because of low interview marks. Although the Court was obviously sympathetic to petitioner’s claim that

he was the target of retaliatory hostility, it found that he failed to carry

the burden of proving mala fides on the part of the committee. But, it

found that the procedures followed by the selection committee contravened the Court’s earlier judgment by departing from a strict allotment of interview marks among the five heads and by proceeding on “wholly wrong premises while granting marks under the head ‘Na-

tional Cadet Corps activities.’ ”’3 The selections were quashed, and the

government

was

ordered

to appoint

yet

another

fresh

selection

committee. Subsequently, in Janki Prasad v. State of Jammu and Kashmir, another

attempt of the Jammu and Kashmir Government to use the selection

committee device to salvage a system of communal quotas was struck down by a constitution bench of the Supreme Court. This was the 39. Id., at 2308.

40. In Gudigar v. State of Mysore, 1972 (2) Mys. L. J. 202 [NIC], the court declined to apply this mechanical allotment requirement where the interview was designed to test suitability rather than to examine for knowledge. Mention of desired traits was “by way of guidance” to the selection committee, and assurance against arbitrariness was to be had by ensuring that the selection was entrusted “to persons of known experience,

integrity and capacity” (id., at 215). 41. A.LR. 1971 S.C. at 2308.

42. Id., at 2309.

43. A. Periakaruppan Chettiarv. State of Tamil Nadu, A. 1.R. 1971 S.C. 2085, 2089 [NIC]. 44. A. 1. R. 1973 S. C. 930.

Design and Operation of Programs

453

fourth round of Supreme Court litigation in a protracted and bitter struggle over the communal quotas within the Jammu and Kashmir

schools.*5 Promotions to the post of headmaster had been found invalid

on the ground that they were distributed on the basis of community. To comply with a Supreme Court order to make new selections, the state

reverted all of those earlier promoted and required all of the 1,100

teachers who were eligible for promotion to appear for an interview by a Departmental Promotion Committee. (Four hundred Kashmiri Pan-

dits boycotted the interview because they “did not have any faith in this system of selection.”)*© The previous appointments (invalidly made on a communal basis) had given 178 posts to Muslims and 134 to Jammu

Hindus. Selecting anew on merit, the selection committee chose 177 Muslims and 134 Hindus. But finding some substance to the state’s claim that things might have been different if the Kashmiri Pandit

teachers had participated, the Court declined to “go merely by the

coincidence.””47

But the interview was in two other ways “thoroughly unsatisfacory”’. First, the selection was based entirely on the interview, without any reference to the service records, confidential reports, or any other information about the candidates. The Court thought that in such a case where there are relevant records that should be consulted, “the

interview cannot be made the sole test.’’48 The state’s failure to avail of

this other information was improper. Second, the committee fixed a “cutting score” or minimum that was

in theCourt’ 's view absurdly low: “those who got more than 30% marks and more than 20% marks from the [outside] expert were declared eligible for selection. This is indeed a travesty of selection. . . a selec-

tion made on such a poor basis cannot be called a real selection at all"?

The whole selection process, then, was “wrong and unsatisfactory and must be set aside.”’5° The Janki Prasad court takes judicial monitoring of

45. In Triloki Nath Tiku v. State of J. & K. [I], A.LR. 1967 S.C. 1283, the Supreme

Court found that it had insufficient material to decide if the resrvations of 50% for Muslims and 40% for Jammu Hindus were indeed for Backward Classes and directed the High Court of Jammu and Kashmir to gather the necessary material. In Triloki Nath

Tiku v. State of J & K [II], A.I.R.

1969 S. C. 1, the Supreme Court found the

distribution of posts not within Art. 16(4), but an unconstitutional distribution on the

basis of community and residence. The state’s initial evasive response was set aside in Makhan Lal Waza v. State of J. & K., A.1.R. 1971 S. C. 2206. 46. Id., at 934. 47. Id., at 934.

48. Id., at 934. The Court was clearly of the view that the government's failure to avail of these records was an exercise of bad faith, finding it “impossible to conceive” that the reports were not available as the government somewhat lamely claimed. Ifa

few were missing, said the Court, to decline to use any of them “would be ridiculous” (id., at 935). 49. Id., at 936. 50. Id., at 936.

454

The Judicial Process

selection an important step further than Viswanath and Periakaruppan.

In those cases it was the failure to conform to the rules prescribed for the selection that was the focal point of unconstitutionality. Here the rules themselves are held to fall short of minimal constitutional standards by disregard of relevant information and adopting standards that abandon minimal concern with quality. So rulefulness is not enough where crucial substantive commitments are subverted. The history of this litigation does little to dispel the conclusion that the selection committee was here yoked to the unconstitutional purpose of communal quotas. It is then difficult to be sure how much the State might vary the sources of information or minimal scores in a scheme of preference that is legitimate in its scope and in its intended beneficiaries. In spite of its dubious history, the selection committee does not lack proponents. Doubts about the relevance of grades as the predominant basis for selection are compounded by concern for the social bias incorporated in those grades.51 If in principle intellectual and leadership qualities can be better ascertained in personal encounter than by written examination,

it remains

unclear that this promise could

be

fulfilled without an investment of resources far beyond that evident

from the cases.

As the cases suggest, the selection committee is a device with an enor-

mous potential for abuse. It is abuse that is especially difficult to correct

because it is difficult to prove or even to discern. The potential for

distortion of merit by selection committees is at least as great as that introduced by reservations. It is now firmly established that the standards of overtness and rulefulness applied to other phases of the selection process are also applicable to selection committees. Just as courts have looked beneath

the

formal

shell of expansive

reservations

to find

that schemes

amounted to “fraud on the constitution,” so they have begun to look beyond the forms of the selection to assess the operative realities.5? 51. Thus the Havanur report proposed that interviews be alloted weight equal to the

written examination: “The merit ofa candidate can be assessed better by his alertness,

intelligence, intellectual outlook, his presence of mind and nervous equipoise. Emphasis on the interview will help offset the deficiencies of the candidates coming from the rural areas, and the peasant, artisan and low castes who do not have adequate opportunities to prove their ‘talents’ at the written examination. Better performance at the University or qualifying written examination is due to urban biased education. The present system of open competition has tended to deprive the backward communities, drawn mostly from the rural parts, of their adequate share in the administration, and the administra-

tion is not representative of the society at large” (Government of Karnataka 1975: I, 212-13). 52. The courts are not the only influence operating to curtail capriciousness, as is illustrated by the situation in Mysore. By 1967 the scoring of the interview had been

Design and Operation of Programs

455

Judicial scrutiny of the adequacy of selection procedures may contribute in an important way to the implementation of reservations policies. For interviews and selections committees can be abused, not only to create what are in effect illegitimate reservations, but to frustrate and

undermine legitimate ones.5? So far, though, the persistent complaints of

beneficiaries that they are unfairly eliminated in interviews have not been brought before the courts.

B. RESERVATIONS AS GUARANTEED MINIMUMS Any scheme of preference must be administered so as to operate in

Savor of the backward class. In Raghuramulu v. State of Andhra Pradesh,

admission to a medical college was denied Backward Class candidates,

although their scores were “general competition,” on Backward Classes had been as a ceiling on admissions backward classes prevented

higher than candidates admitted in the the ground that the 15% reservation for filled. The interpretation of the reservation “instead of advancing the cause of the some members of those classes from getting

seats which they would have otherwise gotten if all the seats were brought under a common pool.”55 Such rules, said the Andhra Court, were to be confined in their operation to cases in which the backward

classes were benefitted, and their application in cases where they did not operate for the advancement of the backward classes violated the

fursdamental rights of the latter. Article 15 (4) authorized the abridgment of fundamental rights in order to make special provision for backward classes. Thus, if backward classes secure more seats by merit than are reserved for them, the reservation cannot be used to limit them

to a prescribed number. The Supreme Court has upheld the position that reservations are not to be construed as compartments. A Scheduled Tribe candidate who had made the special declaration required to stand for the reserved seat in a double-member parliamentary constituency was held not disquali-

fied from being elected to the general seat. The Supreme Court found

that the reservation (authorized by Article 330) was to be construed asa abandoned and the interview reduced to a check on the documents which support the application. Informants suggested that this was because the scoring provided little additional separation of applicants, but exposed the committee to unwelcome pressures and charges of corruption (letter from Glynn Wood, dated Bangalore 21 Dec. 1967).

53. Cf. the persistent complaints of Scheduled Caste candidates that they are unfairly

eliminated in interviews and personality tests in competing for higher posts in the services. See chap. 4, §B, above. 54. A. 1. R. 1958 A.P. 129. 55. Id, at 131

456

The Judicial Process

guaranteed minimum and not as a barrier in election to additional unreserved seats.56 This principle that reservations are guaranteed minimums, rather than maximums, ceilings, or closed compartments,

seems equally applicable to reservation of posts under Article 16(4),

which empowers the State to make reservations “in favour of’ backward classes.5”

On the other hand, members of backward classes are not constitutionally entitled to a system of selection which transforms a reservation into an award of a guaranteed number of seats over and above those

obtained by merit.58 In Puppala Sudarsan v. State of Andhra Pradesh,5? an unsuccessful Backward Class candidate for admission to medical college contended that a higher ranking Backward Class candidate should

be given a seat from the general pool rather than a reserved seat, thus enabling petitioner to obtain a reserved seat. The High Court, however,

said that the selection should not be made in separate compartments in such a way as to allow Backward Class candidates to compete for beth

general and reserved seats to the detriment of others. Reservations are to operate as a guaranteed minimum. Backward Class candidates, it was held, may obtain additional seats by merit but cannot require that reservations operate as a guarantee of seats over and above those

acquired by merit. In Chamaraja v. State of Mysore, a narrowly unsuccessful Backward

Class candidate for a medical college seat argued that the reservations provided under Article 15(4) were distinct from, and in addition to, the

right of successful merit candidates, whose right to admission was

guaranteed by Article 29(2). The court held that Article 15(4) only empowered reservations but did not require them, leaving to the State broad discretion about how it uses this power, including an option to mesh its reservation with the guarantees of Article 29(2), as it had done

56. V. V. Giri v. D. Sura Dora, A. 1. R. 1959 S. C, 1318, affg. A. I. R. 1958 A.P. 724.

Double-member constituencies were abolished in 1961. This abolition seems to have been inspired principally by the concern of politicians with the greater difficulty and expense of campaigning

in a double-sized district. But this was compounded

by

resentment of the infrequent but well-publicized instances in which the unreserved “general” seat as well as the reserved seat was won by Scheduled Caste or Scheduled

Tribe candidates. Dushkin (1972: 192) says the Dora decision triggered the move for

abolition.

57. Harpartap Singh v. Union of India, 1970S. L. R. 40, 43: an objection to reservations

on the ground that Scheduled Castes already held a higher percentage of the promotion posts was based on “complete misunderstanding of the situation.”

58. This “over and above” arrangement was employed in Mysore until the decision in the Partha case, discussed below at n. 61. In the 1960 professinal college admissions,

Backward Classes who enjoyed a 22% reservation secured 40% of the seats (Government of Mysore 1961 : 23).

59. A. I. R. 1958 A.P. 569.

60. A. I. R. 1967 Mys. 21.

Design and Operation of Programs here.

457

In effect, the State might make reservations in the form of

guaranteed minimums. In Partha v. State of Mysore,®* it was held that a reservation could not be

used to set up two separate competitions; it must operate as a guaranteed minimum within a general competition for all seats. Allowing members of the backward group to compete separately for reserved seats and for general (i.e., unreserved) seats was not permissible, since it abridged the fundamental rights of the others. The only permissible method, said the court, was to prepare a single consolidated list of candidates in order of their merit, to draw a

tentative line under the

number of places open, to count the number of members of “privileged” groups above the line. In the event that some privileged group had a number of successful candidates which was less than the number of

seats reserved, additional members of that group were to be selected in

order of merit and an equal number of others deleted in inverse order of

merit.

The very different results that different methods may lead to may be

seen by a simplified tabular representation (table 23). Let us take the simplest case of a reservation: a double-member constituency in which one of the seats is reserved for member of group R and the other seat may be contested by any member of the general public (G).© Iftwo Rs

and two Gs contest the two seats, we may observe that under some

circumstances different methods of declaring winners lead to different results,

Methods A and B lead to identical results. If there is a single uncompartmented competition it does not matter whether the reserved place

is declared first and then the rest of the places in order of merit; or

whether all the places are distributed in order of merit and the reservation is employed only in the event that the reservation is not filled with merit candidates (the Partha method). By either procedure, the Rs will

obtain the number of reserved places or the number of merit places, whichever is higher; the Gs will get all the places earned on merit except

for the number of reserved places unfilled by merit. The other methods, C and D, lead to very different outcomes. Each of them involves a division into separate competitions. In method C, the selection is treated as two separate and simultaneous competitions, one for a reserved seat (in which only Rs compete) and one for the general

seat (in which only Gs compete). In case 4 this procedure leads to the 61. A. 1. R. 1961 Mys. 220.

62. Id., at 235.

63. Although double-member constituencies are no longer used for parliamentary and legislative elections, multi-member constituencies are in use for local and panchayat

elections, and the same problems of method are very much alive. Cf. Shanshikalabai v.

Election Officer, A. 1. R. 1968 Bom. 445, for some anomalous results.

_

458

The Judicial Process

anomalous result that the higher ranking R2 is rejected in favor of G1. This procedure has been rejected by the Supreme Court and several High Courts, because it penalizes rather than helps the backward

classes.

Method D also treats the selection as two separate competitions, not

simultaneous

but successive, with the general to be decided first, in

which the Rs are eligible in both. In cases 5 and 6, this leads to the

anomalous selection of R2 rather than the higher ranking G1. This is what we have referred to as the “over and above merit” method of

reservations. TABLE

23

Possts-e Metuops oF DECLARING WINNERS IN A DouBLE~MEMBER ConstrITuENCY Case.

Case

Case

Case

Case

Gl G2 RI R2

Gl RI G2 R2

Gl RI R2 G2

RI R2 Gl G2

RI Gl R2 G2

1

Order of candidates Ist 2nd 3rd 4th

Method A single

competition after

filling reservation

Method B single

competition with reservation as floor

Method C

separate and simultaneous competitions

MethodD

separate successive

competitions, with general decided

RI

2

RI

3

RI

Gl

Gl

Gl

Gl

Gl

GI

RI Gl

RI GI

RI GI

Gl

Gl

Gl

RI

RI

RI

RI

RI!

RI

4

RI

R2 RI

5

RI

GI

RI

Case 6

RI Gl G2 R2 RI

GI

RI

R2

Gl

Gl

RI (GI

RI Gl

RI Gl

RI

RI

R2

(R2)—

RI

(R2)

first, Rs in both

64. V. V. Giriv. D. Sura Dora, A. 1. R. 1959S. C. 1318 (double-member parliamentary constituency); Diganbar Rao Bindu v. Dev Rao Kamble, 15 E. L. R. 187 (Bombay High

Court, 1958); Raghuramulu v. State of Andhra Pradesh, A.1.R. 1958 A.P. 129 (reserved

places in medical college).

Design and Operation of Programs

459

Some important schemes, notably the central government's job re-

servations for Scheduled Castes and Tribes, operate by the guaranteed-

minimum method. But a large number of reservations by the states are of the “‘over and above” type. Thus we find “over and above” arrangements in admissions in Andhra Pradesh,®5 Punjab, Tamil Nadu,®7

and U.P.® “Over and above” type of reservations of government posts are

found

in

Andhra

Pradesh,®

Kerala,?°

Punjab,7?

and

Tamil

Nadu,”? also in Maharashtra,73 where it is accompanied by a ceiling. Merit may expand recruitment from specified classes above the 24% reserved for them, up to 40% of the total vacancies. At this point, pre-

sumably, the guaranteed-minimum principle would reassert itself. This

is in effect a combination of the two methods: it is a 24% reservation on in “over and above” basis, but as the success in general and reserved

pools reaches 40%, it is transformed into a guaranteed-minimum reservation of 40%. “Over and above” schemes have been urged on the cen-

tral government as a way of augmenting its reservations, but have not

been accpeted.74

In recruitment of Scheduled Castes and Tribes, such provisions until now have had little effect in most areas. But where combined with an expansive definition of Backward Classes, ‘‘over and above” reservation for Backward Classes may have major distributive consequences. 65. The Andhra Pradesh Backward Classes Commission (1970: 102) recommended

restoration of the “over and above” method that had been abandoned a decade earlier.

The new arrangement is found in Balaram v. State of Andhra Pradesh, A. 1. R. 1972 S.C.

1375. 66. In Ramesh Chander Garg v. State of Punjab, A. 1. R. 1966 Punj. 476, we find a

reservation for “students from recognized backward areas ofPunjab utunable to get

admission on merit.”

67. Tamil Nadu Backward Classes Commission 1971: II, 196-97. 68. E.g., the U.P. medical admissions scheme challenged in Dilip Kumar 0. Government

of U.P., A. 1. R. 1973 All. 592, 594 specified that ‘candidates belonging to the reserved categories who qualify . . . for . . . general seats on the basis of merit will . . . be treated as general and admissions against the reserved seats ... will be made only from amongst those candidates of reserved categories who do not qualify . . . for admission

against the general seats.” (The reservation for women was excepted from this provi-

sion.) This aspect of the admissions scheme was not challenged in Dilip Kumar or the other cases which overturned it on other grounds. 69. Andhra Pradesh Backward Classes Commission 1970: 115. 70. Rule 14(b) of Kerala State and Subordinate Service Rules 1958 (as amended to 1970), App. IX in Kerala (Backward Classes Reservation Commission) 1971: H, 169. This was present but not challenged in Hairharan Pillai v. State of Kerala, A.1.R. 1968 Ker. 42. 71. RCSCST 1973-74: 99.

72. Tamil Nadu Backward Classes Commission 1971: I, 32. The Commission notes

(id., at 98) that “it is expected that the better . . . aspirants will get into the unreserved quota of Government service on their own merits without the protection of the

reservations.”

73. RCSCST 1973-1974: 99.

74. Id.

460 The Judicial Process

Thus in Periakaruppan, the Supreme Court observed that “candidates of Backward Classes have secured about 50% of the seats in the general

pool....”75 In a guaranteed-minimum method, this would have liquidated the 25% reservation for Backward Classes, but under the “over and above” method, it had no such effect. Returning to our, schema, we may summarize the constitutional acceptability of the ‘alternate methods as follows: separate compart-

ments (method C) are an unconstitutional method of administering reservations; the guaranteed-minimum methods (A and B) have been commended; the" ‘over and above” method (D) has been held not to be

constitutionally req uired. One court found the over-and-above scheme to be constitutonally prohibited, but since that 1961 case this kind of scheme has come before courts on several occasions without exciting judicial condemnation.” In Periakaruppan, the Supreme Court’s reaction to heavy success by Backward Classes in the general pool was that this showed that “the time has come for a de novo comprehensive ex-

amination of the question [of who are the Backward Classes}.””77

The matter was raised even more pointedly i in State of Andhra Pradesh v. Balaram, where it was argued that since more than the reserved quota

of Backward Class candidates had secured seats on merit, there could

be no further selection against the reserved quota.”8 The Court declined to address this as a question of constitutional stature: “we can only state

fier responded] that it is the duty of Government to review the question of and

er reservations of seats for such groups.””? In both Periakaruppan

Balaram, it was pointed out to the state government that the question

of designation was open to judicial review, but neither court seemed inclined to take up the method of administering reservations as a constitutional question. In view of the broad authorization of Article 15(4), which empowers the State to make “any special provision” for the advancement of the backward, and of Article 16 (4), which permits “any

provision” for the reservation of posts for the backward, it does not appear that the State is constitutionally debarred from employing an

75. A. 1. R. 1971 S.C. 2303, at 2311. 76. In Jacob Mathew v. State of Kerala, A. 1. R. 1964 Ker. 39 at 63, there is dictum to the effect that reserved places may be calculated on such an “over and above merit” basis.

It appears that this conclusion is based on a misreading of the Andhra cases discussed in nn. 54 and 59 above. 77. Periakaruppan v. State of Tamil Nadu, A. 1. R.

1971 S. C.'2303, at 2311. This

judgment was delivered while the Tamil Nadu Backward Classes Commission was at work. The Commission submitted its report two months later, but its termsof reference

did not include a mandate to address the question of who should be included among the

Backward Classes (Tamil Nadu Backward Classes Commission 1971: I, 2). 78. A.I.R. 1972S. C. 1395 at, 1400.

79. Id., at 1400.

Design and Operation of Programs

461

“over and above” merit scheme of reservation. The Constitution itself employs a variant of the “over and above” mechanism in the case of reservations of legislative seats.® . The “guaranteed minimum” type of reservation has several advantages. First, it insures that the amount of effective reservation is somehow commensurate with the backwardness that inspired it. That is, the less able the members of the backward group are to compete on merit, the more preference they get as a result of a given reservation. If there are reservations of 20% for backward group A and 20% for backward group B and the As win 5% of the places on merit and the Bs 15%, the net effect of the reservation (assuming that sufficient candidates secure the minimum acceptable marks) is 15% in the case of the As and 5% in the case of the Bs. The more backward the group, the greater the effect of .the reservation. More important, such a reservation is selfliquidating. As increased educational opportunities lead to better performance by members of Backward Class A, the effect of the reserva-

tion declines as they increase their success in merit competition. The

disadvantage of such “guaranteed minimums” is that they may overstate the effective amount of reservation, leading to complications with the 50% limit in the courts and having adverse effects on the morale of the non-backward, who may be led to underestimate their own opportunities to compete on merit, and of the backward whose entrance on

merit is labelled as a reservation. -

An “over and above” reservation may be easier to administer than a guaranteed minimum in situations like, e.g., elective office from singlemember constituencies. In recruitment to services where a roster of rotation is used to designate reserved turns, it is obviously simpler to be able to ignore the success of backward class candidates on the unreserved turns.

Another advantage of the “over and above” method is substantive as

well as administrative. A guaranteed minimum requires that everyone in the competition be identified as belonging to the preferred group or not, so that all members of that group who succeed on merit can be sub-

tracted from the reservation. Thus, it is necessary to collect identity datd for everyone, instead of only for those members of the preferred group who wish to be considered for the reserved places. The “over and above” method, in contrast, is compatible with the privacy of nonbeneficiaries and those backward class members who prefer not to disclose their identity. And the ‘over and above” method may afford some welcome flexibility in policy in providing preferential treatment to a sizable group who are already getting a number of places on merit. 80. I.e., there is no diminution of reserved seats as a result of additional seats won in

open competition in unreserved constituencies.

462

The Judicial Process

Suppose a state provides a reservation of 20% for backward class A which makes up 40% of the state’s population at a time when As are winning only 5% of the places on merit. In the course of time, improvement of their condition leads to a situation in which As are winning 20%. Now the net effect of the reservation is nil. If the state still considers the As backward and deserving of preference, it has two choices: it may raise the reservation—say to 30%, which would mean that the net preference would be 10% —with all of the problems of general morale that this would entail. Or it could substitute a lower reservation that was not a guaranteed minimum, but of the “over and above”’ type— 10% again. This 10% reservation would have the same net effect as the 30% guaranteed minimum. It has the advantage that

the public knows just how much preference the As are getting, and protects against adverse effects on public morale of exaggerated notions of the amount of reservation. But the “over and above” method has the

decided disadvantage of rigidity. What may be reasonable one year may very shortly be obsolete. This method does not have the selfadjusting and _ self-liquidating features of the ‘guaranteed minimum” and therefore requires closer judicial scrutiny to prevent abuses. But the principal argument®? advanced by the proponents of “over

and above” reservations is that it would insure that some benefits go to the really backward.8? This is a particularly acute problem when the reservation is uncompartmented and the beneficiaries include some groups or strata that are highly qualified relative to other groups or

81. A curious argument by Kuppuswamy (1978: 111) that guaranteed minimum type reservations will lower the motivation to excel (“each group will then perpetuate itself as being backward and depend wholly on reservation for its prospects”) is unconvincing in its assumption of a high degree of group management over the performance of individual candidates. 82. See, e.g., the plea for the “over and above” merit method (D) contained in the

resolutions of the 1959 meeting of Andhra Pradesh Backward Classes Association: “It is highly regrettable that the Government are still following the procedure . . .of selecting candidates first for the reserved seats and then for general seats [Method A

above]. . . . This procedure strikes at the very root of the principle of reservation. The

principle of reservation is meant for enabling candidates belonging to the Scheduled Castes, Scheduled Tribes and other Backward Classes, who are otherwise qualified but

are unable to compete for general seats, to come up educationally and enter Government services. This does not prevent candidates belonging to these classes from competing for general seats. The procedure now followed by the Government in this respect is against all canons of fairness and justice. So, this . . .Body earnestly urge

upon the Government to restore the old procedure of selecting candidates first for general pool and then for reserved seats.” (Andhra Pradesh Backward Classes Association

1959: 2). A decade later the Andhra Pradesh Backward Classes Commission

(1970: 102) agreed.

Design and Operation of Programs

463

strata. “‘Over and above” reservations prevent the success of the former on merit from liquidating the reservation. But this protection of the most

backward/from

the effect of disparities among

the backward

classes will obtain only in the unlikely event that the relatively advanced groups and strata succeed on merit but do not also cluster at the top of the list for the reserved places. If the less advanced groups within the backward classes are to be protected from these disparities, it can be more efficiently accomplished in the long run by more careful designation of the backward classes and in the short run by arrangement of compartments within the reservation to assure that the top groups within the backward classes do not obtain all the benefits. C. COMPARTMENTS

AND LAYERS

The requirement that preferences must operate for the benefit of the backward raises the puzzling problem of differentiated treatment of the backward—a problem that dramatizes the difficulty of implementing the principle of compensatory discrimination. This problem arises in several forms. One is the permissibility of compartmental treatment. Is the State confined to one aggregate reservation for all whom it designates as backward classes? Or may it make separate reservations for component parts of the total backward group? It also arises in the form of layers or slabs: must the State bestow the same preferential treatment on all of the backward? Or may it designate some of the backward to receive more preferences or to have first call on limited preferences? Opinion is divided about the desirability of these devices. A breakdown of reservations has been specifically recommended on the ground that a single uncompartmented reservation would help only the rela-

tively advanced groups among the backward classes and would be disadvantageous to the most backward.®3 On the other hand, arrangements for compartments or layers of backward classes have been officially opposed on the ground that such distinctions promote caste feeling. 83. 1 BCC xxi, 141, Cf the recommendation of the Lokur Committee that “the various castes and tribes in the lists should be administratively . . . catagorized so as to give higher priority in planning and development to the more needy, and lower priority for

the comparatively advanced” (Dept. of Social Security 1965: 8). 84. RCSCST 1957-58: I, 9. This opinion was apparently shared by the Ministry of Home Affairs. Yet. the Commissioner agrees that it is necessary to compartmentalize benefits tor Backward Classes on the one hand and Scheduled Castes and ‘l'ribes on the other, lest the former benefit from the preferences intended for the latter. RCSCST 1952: 70.

464

The Judicial Process

A number of schemes involving compartments are in use, ®5 and some

of these have come before the courts. It is clear that differentiated treatment of the Scheduled Castes, the Scheduled Tribes, and the

Other Backward Classes is permissible, and perhaps it is even constitutionally required.® So far the courts have not settled conclusively the permissibility of the use of compartments within the Scheduled Castes or within the Backward Classes. Several early decisions suggested that any such compartmentalization of reservations is unconstitutional. The first court to address itself explicitly to this issue was in Ramakrishna Singh v. State of Mysore.®7 Twenty percent of seats in the

state’s professional and technical colleges were reserved for members of the Scheduled Castes and Scheduled Tribes; 45% were reserved for Other Backward Classes. One hundred and sixty-four communities

(castes, sub-castes, and religious groups) were listed as Backward

Classes. These communities were arranged in 14 groups (ranging from 2 to more than 100 communities) and each group was assigned a portion of the reservation (from 1.2% to 8.5% of the total seats).

Among the features of this scheme that the High Court found to be invalid are two which concern the validity of compartments. First, the arrangement of the groups so that “‘each group of backward classes includes one relatively forward class” made it possible that “the limited percentage of seats reserved” for each group “would be captured by those communities who are more forward

. . . leaving the really back-

ward classes with no chances of getting any seats... .”®® The court held that such an arrangement, by diverting benefits from “really backward” to “relatively forward” groups, failed to meet the requirement of the Raghuramulu case® that the arrangement be for the benefit of the backward ‘classes. Here, the court’s holding is based on the conclusion that the favored groups are not legitimate recipients of preferences at all. But the court’s reasoning would seem to invalidate any scheme of compartments which combined, in a single compartment, groups of disparate backwardness—even where both groups were legitimate reci-

pients of benefits.

This argument does not dispose of the question of compartments, for

85. E.g., the communal rotation among various Backward Classes in Kerala. In Jacob

Mathew v. State of Kerala, A. 1. R.

1964 Ker. 39, a “sub-rotation” among various

Backward Classes was struck down because the compartments were drawn wholly on

lines of caste and were not commensurate with the relative backwardness of the respective groups. The rotation was later restored by the Division Bench, when it

upheld the classification (State of Kerala v. Jacob Mathew, 1. L. R. 1964 (2] Ker. 53).

Sub-rotation among the major groups of Other Backward Classes was not challenged in Hariharan Pillai v, State of Kerala, A. 1. R. 1968 Ker. $2.

86. See below, §D. 88. Id., at 351.

_ 87.4.1. R. 1960 Mys. 338. 8Y. A. 1. R. 1958 A.P. 569.

Design and Operation of Programs

465,

it is really only an objection to the arrangement of the groups. But the court has a second and more basic argument against compartments. Under the Mysore scheme, members of each Backward Class “‘can only compete for the seats . . . reserved for that group and are not eligible for the remaining seats reserved for the backward classes.

. . . [T]hey are

debarred from capturing the said remaining seats in open competition amongst the members of backward classes....”%° Extending the principle of the Raghuramulu case that the reservation must operate to the benefit of the backward, the court holds that a backward class must

be able to compete not only for the unreserved seats (as in the Raghuramulu case) but also for all of the reserved seats.91 That is, while the fundamental rights of the general public may be abridged by State

pow er to permit-preferences for members of Backward Class A, the court

ds that Article 15(4) does not authorize abridgement of the fundamental rights of Backward Class B in order to permit preferences for Backward Class A. Since compartments restrict the rights of the Bs to compete for the seats reserved for the As, they are unconstitutional. The two arguments against compartments, both inferred from the principle that reservations must be for the benefit of the backward, are ultimately inconsistent. The first objection— the unfairness of competition between members of groups of unequal backwardness—could be remedied only by more precise and equitable arrangement of compartments; the second objection— that all of the backward, whatever dis-

parities exist among them, must be free to compete for every reserved seat—can be met only by complete elimination of compartments. But the notion that compartmental arrangements are impermissible because members of the backward classes must be free to compete for all the reserved seats encounters both practical and theoretical problems. Does this principle make it unconstitutional for the state to provide, for example, that some preferential scheme applies only to backward classes in a given geographical area? Would backward classes throughout the state bé automatically eligible to participate? Would it invalidate the separate treatment of Scheduled Castes, Scheduled Tribes, and

Backward Classes— giving each of these a right to participate in any schemes for the benefit of another of them?92 90. A. I. R. 1960 Mys. at 351. 91. Cf. Partha v. State of Mysore, A. 1. R. 1961 Mys. 220 at 235. 92. An argument for separate compartments for Scheduled Castes and Tribes might

be found in the wording of Art. 15(4) itself, which mentions them separately from and

alternately to backward classes. Art. 16(4), which does not mention Scheduled Tribes and Castes separately, does not supply any such textual ground for a distinction in the government employment field. But presumably such a distinction could be founded upon Arts. 46 and 335. -

466

The Judicial Process

Beyond these, this conclusion lacks any apparent constitutional foundation. There is nothing in Articles 15 or 16 to indicate that State power to make provisions for backward classes falls short of power to limit preferences to one backward class in order to benefit another. There is no indication that persons who are designated as backward do not share with other citizens the possibility of having their fundamental rights limited to the extent that it is for the purpose of benefitting other

backward groups.

The Supreme Court also intimated some doubts about the constitu-

tionality of compartments. In Balaji v. State of Mysore, the Supreme

Court rejected a scheme under which separate reservations were pro-

vided for lists of “Backward Classes” and of “‘More Backward Classes.”

Article 15(4), the Court held, authorized special provisions only for the “really backward classes.” Finding that the Mysore scheme benefitted not only these classes but also classes that were merely “less advanced, compared to the most advanced classes in the State,” the Court held that “the classification of the rwo categories . . . [was] not warranted by

Article 15(4).’%3 Since the Court clearly indicates that the scheme would have been equally bad had all its beneficiaries (i.e., both legitimate and

illegitimate recipients of preference) been grouped together, its rejection of the compartments is only dicta. The Court indicated that one of the objectionable features of the use of “‘such relative tests” is that they would entail “several layers or strata of Backward Classes . . . each one [of which might] claim to be included under Article 15 (4).”™ But it does not explain why provision for a plurality of backward classes is thought to lie outside the scope of Article 15(4). A decade later, compartments did not excite the same concern. The Andhra Pradesh Backward Classes Commission (1970) recommended parcelling out reservations among four groups in proportion to their estimated populations: 7% to 37 aboriginal and nomadic tribes; 1% to Harijan converts; 13% to 21 “vocational groups;” and 9% to 33 “other classes.”95 The Commission never explains the recommendation for separate reservations beyond stating that it would afford them “equal opportunities” and would be “useful for watching the progress made by each group.” In Balaram the state’s apportionment of the reservation along these lines was attacked as an instance of the division (into more

and less backward) condemned in Balaji. The Supreme Court repelled

this attack by distinguishing this “distribution of seats among the

reserved classes in proportion to their population.” 93. A. I. R. 1963 S. C. 649 at 661.

94. Id., at 658.

95. Andhra Pradesh Backward Classes Commission 1970: 103.

96. Id.

97. A. I. R. 1972 S. C. 1375 at 1399, A distribution among Scheduled Castes (and converts to Buddhism), Scheduled Tribes, Denotified Tribes, and Other Backward

Design and Operation of Programs

467

Compartments may be made sufficiently fine-grained that they ap-

portion seats among communities. Thus in Kerala the reservations for Backward Classes in professional colleges were apportioned among

Ezhavas (9%); Muslims (8%); Latin Catholics (2%); Other Backward Christians (1%); and Other Backward Hindus (5%).9® This feature

has not attracted any challenge in court. The Backward Classes Reservation Commission proposed an even more elaborate sub-rotation of reserved posts among nine groups. Presumably, compartments require some justification beyond apportionment of seats among communities, but so far the courts have not produced

an analysis

of the circumstances

under

which

they are

permissible. But that they are sometimes permissible seems not to be in doubt. But if groups can be sorted into compartments for purpose of apportioning reservations, can they not be sorted into “layers” or ‘“‘slabs,” some of which receive more preferential treatment than others? Such differentiated or layered preferences have long been in use. For example, Uttar Pradesh at one time had two lists of Backward Classes: the first received educational concessions only, the second received conces-

sions in both education and state employment.1°! Uttar introduced economic “ayers” of Scheduled Castes for certain welfare measures.1°2 The central government schemes for sweepers and scavengers, who are the most Scheduled Castes.1 (And, of course, schemes based on

Pradesh also purposes of has special needy of the an economic

Classes was challenged and upheld in Pandit v. State of Maharashtra, A.1. 242. 98. Kerala (Kumara Pillai Commission) 1966: 76. 99. Kerala (Backward Classes Reservation Commission) 1971: 1, 124 100. One recentiHighCourt decision says they may not. In Chhotey Lal [NIC], A. I. R. 1979 All. 135, 154 [NIC], the court declares flatly

R. 1972 Bom. ff. v. State of U.P. that “making

classification of the backward classes into two sections . . . backward classes and .. . moat backward classes isitself unconstitutional.” This rule is announced here in the

context ofa finding that the state has failed to establish the backwardness of the groups in question. It is purportedly drawn from Balaji but ignores the tensions in Balaji as well as intervening judicial development of the point. 101. RCSCST 1957-58: I, 9.

102, For the purpose of assuring that the:most needy ‘receive the most benefit from welfare schemes, the Scheduled Castes have for this purpose been divided into three ps: those with monthly incomes under Rs. 100; those with monthly incomes under Rs. 250; those with high incomes. Members of the last group are not eligible for benefits

from welfare schemes; members cf the second group may receive such benefits only after

all the members of the first group have been provided for (RCSCST 1961-62: I, 28). 103. Id., at 98 ff. Special preference for these groups and for the most backward

among the Scheduled Tribes has been suggested by the Commissioner (RCSCST 1960-61: 148).

468

The Judicial Process

test often sort out the recipients into layers—as has the central government’s post-matriculation scholarship scheme for Other Backward Classes since 1963.)1 Layers, like compartments, may of course be used for purposes other than protecting the most backward. The Kerala Backward Classes Reservation Commission, proposing a series of successively higher income cut-offs in connection with reservations for higher posts, expressed concern that the Balaji pronouncements about layers might cause difficulties.1°5 Of course, the layers it proposed are not for the purpose of preserving some share of benefits for the least well-off. Quite the opposite! They are to enable benefits to be extended to highly placed government officers who are doing well by any allIndia standard. 16 But, for the most part, schemes for preferences are unaccompanied by provision for compartments, and in many cases there is a massive “creaming” of benefits by the more advanced groups within the back-

ward category.1°7 Thus in Tamil Nadu, a group of Most Backward

Classes were from 1957 listed separately for purposes of some educational fee concessions but not treated separately for purposes of admissions or government posts.?° These castes (including such occupations as barbers, washermen, basket-makers, stone—cutters, and fishermen)

were estimated to make up about 44% of the population of Backward Classes.1° In 1970 they formed 9.4% of the Backward Classes students admitted to engineering colleges and 11.0% of those studying in medical colleges.11° They made up only 13.9% of Backward Classes candidates selected for Class II government service.1!1 If their extreme backwardness prevented them from competing for such high-echelon benefits, it might be thought that the benefits for them would be concentrated at earlier-stage preferences. But in 1970 they formed

only 18.9% of Backward Classes students appearing in the Secondary

School Leaving Certificate examinations.1!2 And they accounted for only 22.4% of the expenditures on scholarships for Backward Classes.

In fact, the per capita expenditure on scholarships for them was only 78 104. See chap. 6, §C, above.

105. Kerala (Backward Classes Reservation Commission) 1971: 1, 128 ££.

106. A more detailed account of this proposal is discussed in chap. 8, §G, above. 107. Cf. the observations of Justice Krishna Iyerat N: M. Thomas v. State of Kerala, A. 1. R. 1976 S. C. 490 at 531-32.

108. Tamil Nadu Backward Classes Commission 1971: I, 18. The report says that

these groups were selected because they were “almost as backward as Scheduled Castes . . . [but] did not satisfy the criterion of untouchability.” But a dissenting member of the Commission suggests that the list was copied from the list of “Most Backward” groups compiled by the Kalelkar commission (I, 241). 109. Id., at 176. 110. Zd., at 160. 112. Id., at 160. 111. Id., at 150.

Design and Operation of Programs

469

paise, compared with Rs. 2.10 for the remainder of the Backward Classes. 133 As one focusses on smaller groups, this creaming effect and the corresponding exclusion of the most backward is cven more pronounced. Table 24 summarizes data supplied by the Commission comparing the flow of benefits to nine castes who are prospering under the protections for the Backward Classes and a group of seven numer-

ous castes (all but one from the Most Backward Classes list) who are not

doing nearly as well.

TABLE 24 A Comparison oF Upper AND LOWER STRATA OF BACKWARD CLASSES 1N Taso Napu

Nine Castes Percentage of Backward Classes Population

11.3

12.1

443°

18

47.3¢

0.4¢

30.3¢

3.5¢

44.0°

18°

37.3¢

1.9¢

48.2°

0.9¢

Rs. 4.63

[Rs.0.78]°

Students admitted, Engineering, 1970 Studying first-year MBBS 1969-70 (one college omitted)

Students appearing in SSLC examination, March 1970

Candidates selected for Group II Services by TNPSC

1964-1969

Non-Gazetted Posts in all departments Gazetted Posts in all departments

Per capita expenditure on scholarships

Seven Numerous Castes*

Source: Tamil Nadu Backward Classes Commission, Vol. I. Notes: ® Six from Most Backward Classes list. » This figure is for the entire Most Backward Classes group (I, 156). The Commission does not provide a figure for its seven numerous castes, but the figures it gives (I, 157) for five of the seven

(0.64, 0.28, 0.50, 0.45, 0.14) suggests that an aggregate figure would be lower than the MBC

average.

© Percentages of all Backward Classes present.

113. Id., at 156.

470

The Judicial Process

These disparities point in two directions. First, they point back to the

question of designation. Are groups and strata that are succeeding in such ample measure properly included among the Backward Classes? Should certain groups be removed? Or can income limits be employed to eliminate the “upper crusts” within a group? Thus the Tamil Nadu Backward Classes Commission suggests the adoption of an income cut-off to exclude the “upper crusts” within the Backward Classes.11* But there is a second problem. Even if the most affluent and successful are removed, the Backward Classes category will still span groups with a wide disparity of competence and resources. Some segments of the Backward Classes may need training or other special measures to increase their ability to utilize benefits. But even if such disparities are reduced, it is unlikely that they will be eliminated. One way of providing that some of the benefits flow to the least advantaged sections of the backward is to provide a special sub-reservation for them. Thus the Tamil Nadu Commission proposed that almost half of the reserved rtion of admissions and posts be set aside separately for the Most ckward Classes. 115 There is no definitive pronouncement by the courts on such schemes of layers. In the Balaji case, as we have seen, the Supreme Court indicated that it found objectionable the division of the recipients into « two categories of “backward” and “more backward.” But there was no attempt in the Mysore scheme to apportion benefits in accordance with the need of the several classes. Since the scheme was found unconstitutional because it covered those who were merely relatively “less advanced” as well as the “really backward classes,””""® the permissibility of the layers as such was not before the Court. But the Court clearly

indicates its objection to ‘‘several layers or strata of Backward Classes

... each one of which might claim to be included under Article 15(4)."17 The Supreme Court’s objection to compartments, along with its insistence that beneficiaries of preferences be both socially and educationally backward,118 amounts to a proposal that the Backward Classes are some single, undivided, and unvarying group who are to be treated uniformly. But backwardness is not a single isolable trait; it exists in kinds and degrees. The backward are not a uniform group. By whatever standards the “Backward Classes” are selected, some of them will be more backward than others; and, ergo, some will be less backward.

Whatever arrangement is proposed for the distribution of preferences among then, it will work to the advantage of some and the detriment of 114, Id.,at 152. 117. Id., at 658.

115. dd., at 21. 116. A.LR. 1963 S.C. at 661. 118. See chap. 8, §C, above.

Design and Operation of Programs

471

others; e.g., the “most backward” would be disadvantaged by a single competition, while the less backward would be disadvantaged by layers and compartments. The whole system of compensatory discrimination rests on the notion that because of disparities in resources and background, the backward must be protected against open competition with

the general public. But a single uncompartmented reservation for all of

the backward tends to reproduce within that group the same kind of unfairness that protective discrimination is designed to eliminate. If the backward are to be protected against open merit competition with the general public, why cannot the most backward be protected against such competition with other sections of the backward? Again compensatory discrimination is aimed at channelling benefits according to

need. But uncompartmented competitions may well give the least benefit to those with the greatest need.

The ultimate objection to compartments and layers seems to rest on the assumption that there is some constitutionally pre-ordained group of

“really backward” persons who may be the beneficiaries of preference

but may never be disadvantaged by preference for others. But this has

no evident constitutional foundation. Articles 15 and 16 seem to antici-

pate a plurality of backward classes. The wording of the articles does not indicate that every class which is found backward for one purpose must be treated as backward for every purpose or that all classes found

backward for a particular purpose must be treated uniformly. As the High Court in the Partha case points out, “the classification of .. . [backward classes] may vary from time to time and with reference to the

nature of . . . [the] backwardness which is sought to be remedied by

special provisions made in respect ofit.” 119 Thus there appears to be no

constitutional bar on the use of the compartments and layers by the State. Such arrangements seem appropriate once it is recognized that among the backward, some are more backward than others in particu-

lar respects and are more necessitous and/or deserving of help of certain kinds. The Constitution seems to envisage that preferences for Sche-

duled Castes and Tribes will be more extensive than those for Backward Classes, a distinction corresponding to their relative lack of resources and opportunities. It would be strange were the government prohibited

from distinguishing among Scheduled Castes}2° and among Backward Classes in order to make preferences commensurate with the backwardness of the recipients. The broad discretion vested in the State by

Articles 15(4) and 16(4) would seem to cover such arrangements. It 119. A. 1. R. 1961 Mys. at 235.

120. Preferential treatment for the most backward sub-castes within the Scheduled

Castes had been recommended, for example, by the Central Advisory Board for Harijan Welfare (Times of India, 15 Apr. 1959).

472

The Judicial Process

may be asserted that the use of compartments, even if constitutionally permissible, has the undesirable feature of providing what are in effect guaranteed communal quotas. To some extent, this is inherent in any

scheme of preferences in which the recipients are designated along

communal lines, and it cannot be eliminated by merely eschewing compartments but only by substituting other criteria of backwardness. The main danger in the use of compartments and. layers is that their arrangement will reflect, not the relative need of the various backward groups, but their relative political power to secure arrangements favor-

able to their interests.122 Again, this can be remedied, not by eliminating differentiated treatment, but by insuring that the compartmental arrangement corresponds in some measurable way to the relative back-

wardness of the groups. Ifcompartments or layers of preference are constitutionally permissible, courts may be faced with another kind of attack on these arrange-

ments. It may be claimed that a particular scheme of compartments or

layers actually operates to the disadvantage of a “more backward” group while benefitting the less backward. In the Ramakrishna Singh and Balayi cases, the courts indicated that schemes which benefitted the “relatively less backward” at the expense of the “really backward” were invalid. The implication was that the former were not legitimate recipients of preferences at all. However, it is easy to visualize the case

in which both groups are “really backward” but one is clearly more

backward than the other. In such a case, are the courts to intervene to see that the most backward of the groups are not disadvantaged by the

scheme? For example, may a group complain that its relative backwardness entitles it to inclusion among the groups receiving the greater quantum of preferences? Such questions could be decided only if the courts were willing to apply some objective standards of backwardness. But the willingness of the court in Ramakrishna Singh to invalidate a scheme of preference on the ground that unequals were treated equally indicates that the same standards might be applied to invalidate schemes where the quantum of preference is incommensurate with

relative backwardness.

D. RELATIONS BETWEEN MULTIPLE RESERVATIONS The constitutional provisions for special treatment envision a plurality of backward groups. When there are reservations for more than one 121. As was so dramatically demonstrated

A. I. R. 1960 Mys. 338.

in Ramakrishna Singh v. State of Mysore,

Design and Operation of Programs

473

category of backward classes, the question arises as to whether the reservations are entirely independent of one another or whether they are

articulated. One problem of articulation is that of immunity from

“bumping.” Is a member of one backward group immune from being

“bumped” in favor of a member of another backward group? Suppose there is a competition for 10 seats among 20 candidates and that there are two backward groups, A and B, for each of whom one seat is reserved. Suppose, further, that in open merit competition, two As score

respectively first (A-1) and tenth (A—10) in order of merit, and that the

highest B is eleventh (B-11). Since reservation is not a ceiling, the

reservation cannot be used to limit the number of As. Since the Bs have

not obtained their quota of reserved seats by open competition, B-11 is to be given a seat. This requires that one of the successful candidates be eliminated. Is A-10 immune from elimination? Should he be passed

over and the lowest of the successful non-backward candidates be elimi-

nated in his stead? To make this more concrete, it is easy to imagine that

the As are the Scheduled Castes and the Bs the Scheduled Tribes or

Backward Classes. So far, there is no indication of how this dilemma is

to be resolved. If the Constitution supplies no answer to this dilemma, it is difficult to see that the State’s power to make special provision for members of a backward group does not encompass a power to make seats that members of this group gain on merit immune from “‘bumping” by members of other backward groups. At least one state has a rule

to this effect: the Mysore rule provides that deletion be of “applicants not belonging to any of these catagories.””122 One may imagine a further elaboration of the “bumping” problem. Suppose an income cut-off provision makes the most affluent members of Backward Class X ineligible for reservations, and suppose further that an affluent X is the last person on the merit list and that an insufficient number of backward Xs gain places on merit. To fulfil the reservations for the Xs it will be necessary to eliminate one of those on

the merit list. Should the reservation fora Backward Class X be filled by taking the place that would have gone to the affluent X? This is ordinarily the way that a reservation as guaranteed minimum would work. May the State immunize affluent Xs who secure seats on merit against being

displaced by Backward Class candidates for reserved places? Absent

some finding of their backwardness, to choose between the affluent Xs and other successful competitors might involve forbidden discrimination. May affluent Xs be granted a kind of intermediate status in which they are sufficiently backward to be exempt from being bumped but not

sufficiently backward to entitle them to share in the reservations? Such 122. Chamaraja v. State of Mysore, A. 1. R. 1967 Mys. 21, at 22.

474

The Judicial Process

an arrangement would seem consonant with the recognition of the

relative and variable nature of backwardness.123 It might also be regarded as a measure for the advancement of the backward, even if the beneficiaries are not themselves backward, since the affluent Xs attain-

ment of valued positions might, through ties of kinship and group

loyalty, redound to the benefit of the less-affluent backward Xs.124

The reverse side of the bumping problem is whether, if the reserva-

tion operates as a guaranteed minimum, the affluent Xs should be

taken into account in calculating whether the reservation is filled. Literally, one would think not, for they are not members of the Backward Class that is entitled to the reservation. They would then be counted in for immunity from bumping but out for purposes of calculat-

ing the reserved seats to be awarded.

It is quite common for reservations to go unfilled, especially those for

Scheduled Tribes and Scheduled Castes. States not uncommonly pro-

vide for a ‘“‘carry-over” in which the seats unfilled by one backward class are to be filled by another and only revert to the general pool ifno

candidates from the latter are available. In Partha, a “carry-over”

arrangement, whereby any places reserved for Scheduled Castes and Scheduled Tribes and not filled by them were to be filled with additional members of the Backward Classes, was held unconstitutional,

since these were clearly distinguishable classes and the giving to the Backward Classes of the extra seats other than by open competition was an unreasonable restraint on the fundamental rights of other

citizens,125

In Garg 8 seats (2%) were reserved for Backward Classes and 80 (20%) for the Scheduled Castes and Tribes, but only 10 Scheduled Caste and Tribe candidates were eligible. Under a “carry-over” provision, the 70 surplus seats were given to Backward Classes, raising their total to 78 (19.5%). The High Court struck this down as arbitrary, and

“contrary to the spirit and ratio of [the Balaji] judgment.”6 In its

view, specified reservations should represent the maximum limit to which members of given class might be admitted on considerations

other than merit.

In Sukhdev,127the selection was first divided into regional ones, and

123. See the discussion at §C, above.

124. On the broad interpretation of “advancement,” see chap. 11, §B. 125. A. I. R. 1961 Mys. at 235. But some kind of “‘pourover” arrangement seems to have been present in the situation described in Venkataramana v. State of Madras, A. 1. R.

1951S. C. 229 at 229, where Harijans received fewer, and Backward Hindus more, posts

than originally were reserved for them. No objection was raised on this ground.

126. Ramesh Chander Garg v. State of Punjab, A. 1. R. 1966 Punj. 476, at 478. 127. Sukhdev v. Government of Andhra Pradesh, 1966 (1) Andhra W.R. 294 (1963).

Design and Operation of Programs

475

16% within each regional selection was reserved for the Scheduled Castes and Tribes and 25% for the Other Backward Classes. It was then provided that where there were insufficient candidates. in one of these groups in a region, those seats should not revert to the general pool in that region, but should carry over to the corresponding group in another region. Besides the vagueness and arbitrariness of this provision,it was held to militate against the principle of the regional scheme. The court held it to be discriminatory and a violation of Article 14. Ordinarily, the government has treated Scheduled Castes, Scheduled Tribes, and Backward Classes separately, making separate provision for each. Courts have looked favorably on this.128 In Janardhan Surraraya v. State of Mysore,9 the Supreme Court indicated that its decision in the Balaji case did not invalidate Mysore’s reservations for Scheduled Castes and Scheduled Tribes, which were not in any manner irregular or unjustified by Article 15 (4). Reservation for these groups, the Court indicated, ‘‘is distinct and separate from, and independent of

the other reservation which was challenged.” 3° The High Court in the Partha case found the ‘“‘carry-over” arrangement objectionable precisely on the ground that it lumped together as a single category the Scheduled Castes, Scheduled Tribes, and Other Backward Classes. Such

treatment struck the court as opposed to the clear constitutional differentiation among them.131 The opposite tack was taken by the one court to squarely uphold a “carry-over” provision. In §. G. Pandit v. State of Maharashtra,*3? the state had reserved 34% of the seats for four groups: Scheduled Castes

and Buddhist converts from them, 13%; Scheduled Tribes, 7%; Denotified and Nomadic Tribes, 4%; Other Backward Classes, 10%. It was

provided that “reserved seats remaining vacant in any of the above . .. . groups for want of students in that group should go to the other groups ...-1433 The court found this unobjectionable because “{a]ll the four groups form one category of socially and educationally backward citizens. . . . The sub-division into the four groups is made obviously only to allocate the reservation ... so that the comparatively brighter

students in one group may not keep out the students of the other

groups.’’1434 In effect the court reads this scheme asa single reservation compartmented to protect the less educationally robust among the beneficiary groups. But the desirability of compartments rests on differ-

128. Pradesh 129. 131. 133.

Ramakrishna Singh v. State of Mysore, A. 1. R. 1960 Mys. 338; Dasu Rayudu v. Andhra Public Service Commission, A. 1. R. 1967, A.P. 353. A. I. R. 1963S. C. 702. 130. Id., at 703. A. LR. 1961 Mys. 235. 132. A. I. R. 1972 Bom. 242. [d., at 244. 134. Id, at 251.

476

The Judicial Process

ences among groups of beneficiaries, not on their similarities. And it requires some further argument

to justify a carry-over from one to

another. Is there some affinity among the groups so that one benefits

from

places

gained

by

the other? Are

the carry-overs

in fact re-

ciprocated? Are the numbers so small that there is a need to protect the

reservation from temporary fluctuation in the number of qualified candidates in each of the groups? Where this is not the case a carry-over rule creates the possibility that some beneficiary groups will enjoy reservations of vastly greater extent than could be authorized for them. (Thus in the Garg case, the carry-over rule amplified the reservation for Backward Classes from 2% to 19.5%.) Carry-over provisions can help to deliver preferences without undue fluctuations. But whether they are likely to do so in a given instance depends on the size of groups and the distribution of qualified candidates among them (which in turn may depend on the kind of benefit that is at issue). It can hardly be deduced in advance from notions of the conceptual difference or affinity between the various beneficiary groups, since they of course not only are different but are also joined by the circumstance of being beneficiaries of

compensatory discrimination policies.

Some states have provided composite reservations for Scheduled Castes and Scheduled Tribes*35—a practice which has never been found objectionable; nor have provisions for reciprocal “carry-over” of unfilled reservations between these two groups, which are used both by

states and by the Centre. Although it is clearly open to the State to recognize such a differentiation between Scheduled Castes, Scheduled Tribes, and Other Backward Classes, it may be wondered whether the Constitution requires

that it do so. In the Balaji case, the Supreme Court stresses that “the Backward Classes . . .contemplated by Article 15 (4) are [those whose backwardness is] comparable to Scheduled Castes and Scheduled

Tribes.”’!37 If this is the case, presumably it would be reasonable for the State

to treat

them

alike.

And,

since

preferences

are, it is hoped,

temporary expedients, these provisions should be construed with an eye

to preventing impediments to ending them. It may well be that the amalgamation of these separate categories will be a step toward the

eventual dismantling of the system of compensatory discrimination. There would seem to be no constitutional warrant to rule out such a

step.

135. E.g., in Kerala as in Jacob Mathew and Hariharan.

136. E.g., the “carry-over” provision in the Central Secretariat was left standing after

the Devadasan case (Tewari v. Union of India, A. 1. R. 1965S. C. 1430). 137. A. 1. R. 1963 S. C. at 658.

14

The Setting and Incidence of Judicial Intervention AFTER ouTiininc the background and contours of the preference policy, we have examined in some detail the working out of the compensatory discrimination principle (and its articulation to other principles) within one institution—the courts—and by one means— litigation. We shall proceed to assess the impact of the courts on the preference policy and to reflect more generally on the capabilities of the

common-law constitutional court to handle such tasks and forward such policies, But first it is necessary to examine the setting, institu-

tional and cultural, in which this judicial activity takes place. A. COURTS AND JUDGES

The judicial encounter with compensatory discrimination occurs almost entirely within the higher reaches of the judicial system.? At the apex of India’s unified, hierarchical judiciary is the Supreme Court, consisting in 1980 of a Chief Justice and thirteen puisne judges.? The

1. Except for a few cases involving protections (debt relief, etc.) and election cases,

which until 1966 were initiated in Election Tribunals, all cases involving compensatory discrimination originated in the High Courts or the Supreme Court. In 1966 the High Courts were made the courts of first instance in election complaints (Representation of the People (Amendment) Act, 1966, §38). It was this arrangement that provided the setting for Mrs. Gandhi’s election case, which provoked the 39th Amendment. That Amendment

enacted, inter alia, a new Article 329 A, which prohibited the calling into

question in any court the election of someone who was at the time of the election Prime Minister or Speaker of the Lok Sabha, or who was appointed to either of those positions

subsequently. The Amendment and the Election Laws (Amendment) Act 1975 (Act 40 of 1975), which empowered the President of India tb decide whether someone convicted

of corrupt campaign practices should be disqualified from an election, were upheld by the Supreme Court in Raj Narain v. Indira Nehru Gandhi, A. 1. R. 1975 8. C. 2299. 2. It has grown from 7 judges at its formation in 1950 to 11 in 1956 to 14 since 1960.

477

478

The Judicial Process

Court sits in benches of two, three, five (in constitutional cases), and

(extraordinarily) seven or more judges.3 The Supreme Court is the highest appellate court in all matters of cri-

minal, civil and constitutional law. Its authority extends to state as well

as central law.‘ Its appellate jurisdiction extends to the High Courts and through them to all inferior courts and to all quasi-judicial tribunals.5 It has a limited original jurisdiction, notably to issue writs to prevent violations of Fundamental Rights.® In addition it has an advisory jurisdiction.” There is no lower central judiciary.* Each state has a single judicial hierarchy with a High Court at its head.? The High Court exercises an 3. Sitting in benches is sanctioned by Art. 145, but Art. 145(3) requires that a

minimum of five judges sit to decide “any case involving a substantial question of law as to the interpretation of this Constitution” or to exercise the Court’s advisory jurisdiction. In 1976, the Forty-second Amendment inserted the requirement that only a bench of seven or more should determine the constitutional validity of any central or state law and that a law should not be declared constitutionally invalid by less than

two-thirds of the judges sitting. This provision (Art. 144A) was repealed before it came

into effect by the Constitution (43rd Amendment) Act, 1977. §5.

4. The 42nd Amendment added to the Constitution Art. 32A which withdrew the jurisdiction of the Supreme Court to hear writs under Article 32 for determining the constitutionality of state legislation. This provision was repealed before it came into effect by the Constitution (43rd Amendment) Act, 1977, §3. 5. Articles 132, 133, 134, 136. These broad grants of jurisdiction are subject to some

explicit limitations in regard to compulsory acquisition of property (Art. 31{6]) and delimitation controversies (Art. 329). 6. Art. 32, Art. 131.

7. Art. 143. The advisory jurisdiction has not been involved in any of the cases concerning compensatory discrimination, with the exception of In re Kerala Education

Bill, 1957, A. 1. R. 1958 S. C. 956, 982-83, which touches on the power of the State to

make reservations in non-governmental schools. This case is not counted in the various computations in this chapter.

8. There are no lower central courts, but there are various tribunals which are part of the central government: incpme tax, labor, industrial, etc., tribunals. Beyond this, the

High Courts themselves occupy a mixed status: the Constitution discussed them under the heading of “‘The States,” but the-judges are paid from the Consolidated Fund of India; appointment is by the President; removal by impeachment is by the central Parliament; judges may be transferred from one High Court to another by the central Ministry of Home Affairs. On the other hand, staff are appointed and administrative expenses borne by the state government. 9. There are 18 High Courts, one for each state, except that Assam and Nagaland share a High Court (since the latter’s formation in 1962) as do Punjab and Haryana (since the latter’s formation in 1966). Delhi has had its own High Court since 1966.

Three of the Union Territories have courts of Judicial Commissioner with analogous power; the other Union Territories are included within the jurisdiction ofa High Court. In four states the High Courts have permanent benches sitting in more than one location. The High Courts of Bombay and Calcutta also exercise an extensive original

jurisdiction in those cities, which is not of concern here. (H. P. Dubey 1968: 386 ff.).

Judicial Intervention

479

extensive jurisdiction to review the lower courts?® and other tribunals,

and it shares with the Supreme Court an original jurisdiction to issue

writs against violations of Fundamental Rights! (although since 1976 it cannot hold unconstitutional a central law). The High Courts, which

vary in size from two judges (in Sikkim) to forty (in Uttar Pradesh), sit as single judges, Division Benches of two judges, or, occasionally, as Full Benches of three, five, or even seven or more judges. In some instances an appeal or a reference on a point of law is possible from a smaller to a larger bench of a High Court. Judges of the High Courts are recruited directly from the bar (usually the High Court bar) and by promotion from the subordinate judi-

ciary.’? Although the Constitution provides for direct appointment of

lawyers and of “jurists” to the Supreme Court, so far all appointments to that Court have, with but three exceptions, been by promotion from the High Courts.3 Seniority has played a prominent role in selection.14 Just over half the appointees have been Chief Justices of their High Courts. The average age of appointees is fifty-seven. The preponderant role of seniority in making these appointments and the compulsory retirement age of sixty-five combine to produce a relatively rapid turnover on the Supreme Court.15 The average service of judges on that

10. Below the High Courts are the Subordinate Courts, usually consisting of two somewhat overlapping levels. In criminal matters, there are various grades of magistrates, graded by the scope of their penal powers, and above them Sessions Courts, which hear appeals from magistrates and which are trial courts for more serious offenses. From the Sessions Courts there are appeals, revision petitions, and references to the High Court. On the civil side, there are various classes of Subordinate Judges; appeal from them and original jurisdiction over larger suits lies with the District Judge.

There is appeal from the court of District Judge to the High Court. There are also a

variety of Small Cause Courts, Revenue Courts, judicial panchayats, and a rapidly increasing number of administrative and quasi-judicial tribunals (Dubey 1968; Law Commission 1958; Khosla 1949).

11. Art. 226. Cf. Art. 227. Petitioners must elect to pursue their writ remedy in either the High Court or the Supreme Court. A petitioner who obtainsa decision on the merits in an Art. 226 proceeding cannot then file an Art. 32 petition in the Supreme Court

(Daryao v. State of U-P., 1962 [1] S.C. R. 574).

12. Judges in the lower courts are members of the state judicial service, chosen either by competitive examination among recent law graduates or holders of other government posts, or by selection from members of the bar of three to six years standing. 13. See Gadbois 1968-69: 326.

14. On the preponderant role of seniority in these appointments, see Law Commission 1958: I, 209. 15. The retirement age in the High Courts is sixty-two, which insures a comparable

turnover there. Compulsory retirement has been accounted a political threat to judicial independence, since many judges look to government for post-retirement reemployment and appointments. See S. Dhavan 1964: 344; Seervai 1968: 1007. Cf. Gadbois 1968-69:330; and R. Dhavan 1977: 29 ff., on the post-retirement occupations

of Supreme Court judges.

480

The Judicial Process

Court has been 67 months, and the turnover seems to be increasingly rapid: judges appointed from 1950 to 1964 served an average of 82 months; those appointed after 1964 served only 52 months.*6 Judges of the High Courts typically serve ten years or more.17

Appointments to the higher judiciary are formally by the President after consultation with the Chief Justice of India and, in the case of the High Courts, with the Chief Justice of the High Court and the Governor of the State.18 No legislative approval is required. In practice, the President acts on the recommendation of the Home Ministry, the government department in charge of services. Chief Justices of the High Courts have exercised initiative in appointments, but this is subject to the concurrence of the Chief Minister of the State, who has effective

control over appointment to the High Courts.1* To the dismay of at least the upper echelons of the profession, there is in effect some admixture of considerations of regional and communal balance in the selection of High Court judges. The emphasis on seniority and regional balance transmits a similar profile to the Supreme Court. Once appointed, a judge ofa High Court or the Supreme Court enjoys complete

security of tenure and can be removed only for cause and only by a

special parliamentary charge.2° The Chief Justices of the Supreme Court and the High Courts occupy positions of strategic importance, because they are consulted on appointments and because they distribute work among their colleagues,

deciding who should sit on which panels. A convention under which Chief Justices are appointed in order of strict seniority on the Court was broken by Prime Minister Gandhi in 1973 and re-established with the 16. The average is calculated for all of the 44 judges appointed after the court was formed in Jan. 1950 who had retired, resigned, or died in office through Feb. 1978. (Ifwe set aside the 4 judges who resigned in the supercession controversies, the average tenure was 66.7 months.)

17. This assertion is based on the following computation: From the ages and dates of appointment of High Court judges sitting in 1977, it was possible to ascertain the expected length of their service, unless shortened by resignation, promotion to the

Supreme Court, or untimely death. Such length of expected service was computed for each of the four High Courts with ten or more cases in our universe of compensatory discrimination cases. (See n. 45 bélow.) The expected length of service on these High

Courts was Karnatak, 15 years; Punjab and Haryana, 14.5 years; Patna, 12.8 years; Andhra Pradesh, 12.6 years (Ministry of Law, Justice and Company Affairs 1977). I

know of no reason to believe that these courts are unrepresentative.

18. On the formalities of selection, see Art. 124 (Supreme Court) and Art. 217 (High Courts). 19. On the practice of selection, see M. Gupta 1964: 368; Gadbois 1968-69: 335; Law Commission 1958: I, 34, 69. 20. Art. 124(4). Art. 218 makes this provision applicable to the High Courts.

Judicial Intervention

481

appointment of Chief Justice Chandrachud in 1978.21 In the Supreme Court, this has led to a rapid turnover of Chief Justices. From 1950 to early 1978, there were 14 occupants of the Chief Justice’s seat—an average tenure of just two years.2 Although judges and lawyers share a common legal education and a common professional milieu, members of the higher judiciary possess some distinctive characteristics. They tend to be drawn from prosperous and

high status families,23 to have been educated

in elite insti-

tutions,” to have had distinguished professional careers and, with few exceptions, very little political involvement. Indian judges, especially in the superior courts, have a tradition of relative isolation from political life and from governmental connections.25 Since Independence there has been some decline in this ‘tradition of isolation and aloofness.”26 21. The chiet” justiceship of the Supreme Court came up for appointment in April 1973, shortly after a momentous decision holding that Parliament’s power to amend the Constitution did not include a right to alter its “basic structure” aggravated the government’s simmering displeasure with the judiciary (Keshavananda v. State of Kerala, A. I. R, 1973 S. C. 1461). Departing from the seniority convention, the government passed over Justices Shelat, Hegde, and Grover, all of whom had opposed the government’s position in Keshavananda, to appoint A. N. Ray the Chief Justice of India. The three “‘superseded” judges promptly resigned in protest. A fierce controversy ensued and spawned a

vast polemical literature—e.g., Palkhivala

1973. In January

1977,

Justice Beg was appointed Chief Justice, in spite of the greater seniority of Justice Khanna, who resigned in March 1977. After the demise of Mrs. Gandhi's Emergency Rule, the Janata Government in 1978 pointedly adhered to the convention of appointing the seniormost judge, notwithstanding grumblings that the judge in question had not been sufficiently stalwart in his opposition to the misdeeds of the Emergency.

22. Had the seniority convention not been broken by Mrs. Gandhi, the number of Chief Justices during this period would have been 16 (barring untimely deaths). 28. Scheduled Castes and Scheduled Tribes were virtually absent from the higher judiciary during the period discussed mn this book. Of the 352 high court judges sitting in 1977 there were only four members of Scheduled Castes and not a single member of the Scheduled Tribes (RCSCST 1975-77: 1,11). 24. For example, they are much more likely to have had legal training in England. Fully one-third of the judges appointed to the Supreme Court until 1967 had been called to the bar in England. On the careers of Supreme Court judges, see Gadbois 1968-69;

R. Dhavan 1977. 25. Gadbois

1968-69: 330. Tripathi (1966: 28-29) attributed the “divergence bet-

ween national aspirations and judicial pronouncements” to the recruitment of judges

from among those who “generally speaking kept themselves aloof from this historic

upheaval [the independence movement] and continued to function in the imperial and princely

governmental

institutions

as administrators,

lawyers,

judges

and

civil

servants.” 26. Judges are relatively little given to extra-judicial expression through speeches, lectures, and books— except for inspirational talks to bar meetings. Chief Justice Subba Rao shocked professional opinion by his active candidacy for the presidency in 1967. More recently, some judges have earned a reputation as advocates of (pro-government)

causes off the bench.

482

The Judicial Process

As arbiters of the constitutionality of governmental action, equipped

with broad powers of intervention, the judiciary was inevitably embroiled in battles between contending political forces. In applying the broad generalities of the Constitution to the actions of the government,

the judiciary emerged as critic and monitor of government—a role enhanced by the fact that other organs of criticism and redress were often feeble or overburdened. So the higher judiciary served not only as vindicator of Fundamental Rights but as the single most persistent and significant agency of criticism of government operation of terms of constitutional norms.?? The courts have not been noted for a marked solicitude for government, which according to Gadbois’ calculation came out the loser in about 40% of the cases in which it was a party. Nor did the courts shrink from employing their power to strike down offensive legislation. For example, from 1950 to 1967, the validity of legisla-

tion was attacked in 487 Supreme Court cases; the Court struck down all or part of the impugned enactment in 128 cases.?8

The higher judiciary in India enjoys enormous respect. Judges are seen as honest, independent, and immune to the partiality and narrowness that are attributed to most actors in public life. The judiciary had

to undertake these expanded responsibilities as critic and monitor of

government at a time when its own prestige and that of the bar was undergoing a relative decline. The emphasis on economic development and the burgeoning of elective politics enhanced the power and pres-

+ tige of other callings, accelerating the long-term drift away from the bar as a focus of ambition for the educated. Judges and lawyers were often

seen

as mere

word-mongers

who

had

little to contribute

to national

development— when they were not obstructing it by overzealous protection of vested interests. The judiciary was severely criticized for thwarting governmental social and economic policy, for being too solicitous of the claims of property, for being too conservative in labor matters (and, in other quarters, for being too ready to bear with the executive).29 Sensitive to criticism, judges have from time to time

employed their broad contempt powers to curb prominent detractors.

In the early 1970s, critics aligned with the government mounted a campaign for a ‘“‘committed judiciary’—generally taken to mean 27. Gadbois 1977. 28. Gadbois 1970a: 3. 29. See M. Gupta: 1964: 363 ff. The impatience with the bar and bench in the early

years of the Constitution was shared by Prime Minister Nehru, who in the course of the

debate over the First Amendment observed to Parliament: “Somehow we have found

that this magnificent Constitution that we have framed was later kidnapped and

hoined by the lawyers” (Parliamentary Debates, Vol. XII-XIII [part I1], col. 8832

17 May 1951).

Judicial Intervention 483 judges favorably disposed toward the “socialist” policies of the government. In spite of outspoken complaints about the formalism of the law and the conservatism of judges, and against a background of.some mistrust of courts and lawyers, attitudes toward the higher judiciary remain remarkably respectful. Among those who deplored particular decisions,

there has been little talk of outright defiance; even talk of evasion is not

respectable in Indian public life. During Emergency rule, the government took care to curtail the authority of the courts by formal legal means, rather than by breaching the fabric of legal authority. Although the Supreme Court disappointed many critics of the Emergency, the

High Courts emerged from Emergency tule with their reputation enhanced as guardians of citizens’ rights. The extraordinary regard in

which the higher judiciary is held is evidenced by theirfrequent use for extra-curricular ‘“‘integrity jobs.” Hardly any situation in which there is

political stalemate or loss of confidence in other authorities arises without aggrieved parties calling for—and often obtaining—an inquiry by a High Court or Supreme Court judge.°

On the bench, Indian judges tend to take an active, sometimes assertive, role in the taking of evidence and the controlling of argument of counsel. (It is my impression that less initiative in the management of the case is left to the lawyers than is the situation in the United States.) Indian judges feel free to introduce arguments and to consult sources which are not put before them by the lawyers.31 But such initiative is limited by their enormous work load. There is an emphasis on verbal skills; opinions are sometimes delivered orally and otherwise are dictated to stenographers.3? The higher courts are heavily overworked,

and often judges have little time for editing the dictated version. The predominant style is formalistic.33 Social phenomena are addressed by

fitting them into a framework of pre-existing legal categories. The Indian version of formalism emphasizes the autonomy and the determinative character of legal learning. Judges expatiate on the inherent meaning of words; they elicit definitive answers from textual passages; they portray themselves as controlled by inexorable rules of procedure and precedent. Judges apply pre-existing rules; that this 30. E.g., four of the commissions appointed by state governments to review the controversial matter of identifying the Backward Classes were chaired by retired High Court judges: the Kumara Pillai Commission in Kerala, the Jammu and Kashmir

Commission, the Andhra Pradesh Commission, and the Gujarat Commission. 31, See, e.g., n. 98, below.

32. Since the early 1970s the Supreme Court has had a few research officers. But the judges do not have clerks (in the American sense) or other professional assistance. 33. On legal formalism, see Kennedy 1973; Friedman 1966.

484

The Judicial Process

involves choices is rarely emphasized, for it is assumed that such choice will be guided by formal principles (e.g., canons of statutory interpretation) rather than by assessment of consequences. Therefore, making choices is a matter of professional knowledge and skill rather than of personal values or political commitments. ‘The prevailing style of judicial work has impressed many foreign observers as less than optimal. The Indian judicial process, one distinguished comparativist put it, “is in an interesting, though not satisfactory, position.”34 Speaking of law in India generally, he observed

that “the difficulty is after all one of intellectual attitude and conception— the failure to approach legal problems functionally and

its corollary, a static conception of law.”°5 And a close student of the

Indian Supreme Court recently characterized the Indian legal culture as one of “procedural nitpicking,” “hair splitting legalisms,” “literal © interpretation,” “narrow, technical, and mechanical,” expressing “concern for form, not for policy or substance.”’36 One can recognize in this unflattering portrait some of the pervasive stylistic features of the judicial process in India. But our compensatory discrimination cases suggest that the formalistic style should not automatically be equated _ with an absence of policy concern on the part of the judges. We find instances of judges smoothly blending the idiom of the formal style with reflective pursuit of substantive policy.37 We even find instances of tacit but incisive pursuit of policy by strict invocation of the canons of formalism.3® (And, as everywhere, judges may use the canons of formalism to torpedo unworthy or unappealing claims.)39 Just as purposive forays are possible within the formal idiom, abandonment of that idiom does not guarantee that chosen policies will be effectuated.“ Although the formalistic idiom undoubtedly constricts the range of 34. Von Mehren 1963a: 279. 35. Von Mehren 1963b: 15.

36. Gadbois 1977.

37. E.g., Chatturbhuj Vithaldas Jasani v. Moreshuar Parashram, 1954 S.C.R. 817; Shantha Kenar v. State of Mysore, 1971 Mys. L. J. 21; Urmila Ginda v. Union of India, A. 1. R. 1975

1. 115. § 38, Periakaruppan v. State of Tamil Nadu, A.1.R. 1971 S.C. 2303, discussed in chap. 13, A, above.

39. Consider Ramalingam v. Boddu Abraham, A. 1. R. 1970 S. C. 741, 743, where the

losing candidate failed to prove his allegation that the winner had converted to Christianity (and was therefore not a Hindu), because it was the winner's parents who had converted. Where “because of clumsy blundering the petitioner undertook a much greater burden than the law required him to take” (i.c., proving winner’s conversion rather than that he was a non-Hindu), he lost because of “the vagaries of litigation

which was to be carried on according to rules.” 40. Cf. the analysis of the imponderable consequences of the “purposive” decision in the Thomas case, chap. 1:1, §D, above.

Judicial Intervention

485

policy considerations that judges take into account, it does not exclude considerations of substance or policy entirely. The restrictive effect of judicial ideology cannot be separated from the other features of the judges’ situation that militate against purposive concern with policy. The enormous work load that faces them, the fact that they encounter

isolated pieces of complex situations, the lack of information or analysis on which to base a decision, and the imponderable effects of many decisions surely impose limits at least as formidable as attachment to the formal style. The isolation of the judiciary from the purposive effectuation of policy has sources deeper than judicial style. But this

style has important effects as a cultural model, if not as a structural

barrier. By educating lawyers about what are acceptable arguments (and ultimately educating claimants about what are justiciable claims), formalism stabilizes the role of the judiciary: it elicits claims and arguments that courts respond to in the formal idiom. Judges and lawyers, whether trained in India or abroad, have been exposed to legal education which stressed formal analytic treatment of legal doctrine, rather than considerations of policy. They have virtually no exposure to social inquiry in or out of law college. There is little in their professional life to offset this emphasis. India has not had a flourishing legal scholarship. As one reflective High Court judge ob-

served, “[t]he one great weakness of the Indian judicial system today

[1964] is that [it] lacks. . . theoretical nourishment.”1 Since then a critical literature has begun to emerge,‘ but there is little to indicate that it has reached many of the lawyers and judges. Most judges are not voracious readers, and if they read any legal periodical, it is more likely to be an English or American than an Indian one. In their judgments, the judges do not often venture outside the boundaries of case law. Standard legal works are sometimes cited, as are government reports. But reference to current scholarly comment is rare. In the litigation about compensatory discrimination there has been only a single instance in which a judgment referred to scholarly comment on this topic.*3 In some of the group membership cases, judges have referred to ethnographic works for evidence of tribal usage. But where judges venture beyond the boundaries of Indian case law, the sources are likely to be Anglo-American rather than Indian and conceptual rather than descriptive. In our compensatory discrimination cases, the most elaborate citation of the external sources is found in the

various judgements delivered in the Thomas case. Although the judges 41.8, Dhavan 1964: 348, 42. As reported (R. Dhavan 1977: 71) and exemplified by that judge’s son.

43. In Sagar v. State of Andhra Pradesh, A. 1. R. 1968 A.P. 165 at 189, Jaganmohan

Reddy, J., cites Radhakrishnan 1965 on the interpretation of Chitralekha.

486

The Judicial Process

spend most of their time on the Indian case law, several refer copiously

to American cases as well. Justice Ray cites a California Law Review article by Tussman and ten Broek; Justice Mathew cites Plato, Aristotle, Harold

Laski,

Bernard

Williams,

R.

H. Tawney, John

Rees,

a

Harvard Law Review comment, and the Moynihan Report. Justice Krishna Iyer draws upon Charles Black, Wolfgang Friedmann, An-

thony Lester in the Modem Law Review, Ralph Ellison, and the off-thebench writings of former Chief Justices Hidayatullah and Gajendragadkar, and he provides what appears to be the single reference to any work analyzing Indian social phenomena.® Clearly, the connection between judicial decision-making and systematic empirical learning about Indian society remains to be established. Overall the courts have been harsh critics of governmental arrangements for compensatory discrimination. A search of the law reports turned up a total of 113 reported litigations about compensatory dis-

crimination from 1950 to 1977.45 Putting to one side the 32 electoral disputes, consider the outcome in the other 81 cases. The High Courts

44, His statement, A. I. R. 1976 S. C. 490 at 531, that “research conducted by the

A. N. Sinha

Institute of Social Studies,. Patna,

has revealed a dual society among

Harijans . . .” apparently refers to the research published as Sachchidananda (1977). 45. For purpose of this and subsequent computations, the universe consists of all

reported cases in the High Courts and the Supreme Court in which a substantial claim

was asserted as to the existence, method, or extent of a policy of compensatory discrimination, or the designation or membership of the recipients of preference. It does not include attempts to extend doctrine from this area to other kinds of fact situations. Reported cases seems a reasonable place to draw the line because they are part of the corpus of authority and because someone with a grasp of the Indian legal scene made the decision that they were worth reporting. Thus, I have omitted some unreported cases

which clearly fit the description above, along with some election cases which apparently

did not proceed beyond the Election Tribunal and a few cases in which the claim in regard to preferences seems marginal. I suspect that a few cases, reported only in local law reporters, have escaped this enumeration, but there is no reason to believe that they would in any way alter the general profile. The same is true for the few reported cases discussed in this book but discovered too late for inclusion in these computations and

marked [NIC].

Obviously, there is no assurance that this sample is not biased because of unre; cases. Although much trivia is reported, in a group this large over this period of time, I think it is safe to assume that the reported cases fairly represent what the courts have done. A partial check was permitted by examining the 6 unreported cases in my possession and by examination (in 2 cases), and indirect reconstruction (in the others)

of 16 unreported judgments in litigations where later stages were reported. I could detect no important difference in subject or disposition, except for a greaterPreponderance of election disputes among the unreported cases.

Although for convenience they are referred to as “cases,” it should be uted that the

unit here is not reported case as such but a

litigation—i.e., a dispute involving any

number of parties and going through one or more courts. Otherwise, the accident of a

case being reported two or even three times could bias our computations.

Judicial Intervention

487

struck down governmental action in whole or in part in 27 cases while

upholding it in 40. State rather than central arrangements drew the most critical response from the High Courts: in 57 litigations concern-

ing preferential treatment conferred by the state governments, they

ruled against the state in 24. But in ten cases involving central preferences, they ruled against the Centre in only three. While all of the

central preferences were for Scheduled Castes and Tribes, 16 of the 24 invalid state arrangements were for the Backward Classes. Central schemes and schemes for Scheduled Castes and Tribes fared better in the Supreme Court as well as in the High Courts. Putting aside the electoral cases, the Supreme Court gave decisions in 24 litigations, in which it upheld the government's preferential scheme in only 8 cases and struck it down in whole or part in 16 instances. Only 4 of these cases involved central schemes and they were upheld in 3 of 4 cases. But state

schemes were struck down in 15 of 20 cases. The Supreme Court was

particularly censorious of government in the exercise of its original writ

jurisdiction. Eleven Article 32 petitions were brought against state

government schemes; they were struck down in 10 cases.“ The Supreme Court was more permissive when it sat on appeals from the

High Courts. In 10 appeals it reversed the High Courts six times.*7 In 5 of these, the Supreme Court held that the preference conferred by the state was permissible where the High Courts had said that it was not.48

Three of these 5 concerned Scheduled Castes— including the 2 major cases enlarging government power to confer preferences on them.”

Cases involving compensatory discrimination have, at times, placed some strain on the judges because of the intense political and social

controversy involved. This is an area in which judges’ values, ideals, and perceptions of their society cannot but affect their reactions. But

the relative infrequency of such litigation, the system of sitting in

46. Perhaps this reflects the Court’s ability to pre-screen Article 32 cases at the admissions stage. 47. The six cases are General Manager v. Rangachari, A. 1. R. 1962S. C. 36; Sarda v. Mize

District Council, A. 1. R. 1967 S.C. 829; State of Punjab v. Hira Lal, A. 1.R. 1971S, C. 1777; State of A.P. v. Balaram, A. 1. R. 1972 S. C. 1375; State of U.P. v. Pradip Tandon, A.1.R. 1975 S. C. 563; State of Kerala v. N. M. Thomas, A. 1. R. 1976 S. C. 489. There were also

three reversals in electoral cases (Punjabrao v. Meshram, A.1.R. 1965 S. C. 1179; Abhay

Pada Saha v. Sudhir Kumar Mondal, A. 1. R. 1967'S. C. 115; Laxman Siddappa Naik v.

Chandappa, A. I. R. 1968 S. C. 929), in each of which the High Court had held that a su

was.

1 candidate was not a Scheduled Caste, but the Supreme Court found that he

48. Only in Sarda is the governmental action not upheld, and the preference element there is not clear. In Pradip Tandon the Court upholds the more permissive of the High

Court responses to the state scheme and overrules an earlier, more restrictive, High

Court case.

49. L.e., Rangachari and Thomas.

488 The Judicial Process benches, and rapid turnover on the courts have allowed few judges to sit in enough cases involving such schemes to articulate in their judgements an explicit and comprehensive point of view on compensatory discrimination. For example, if we take the Supreme Court, we can identify 39 cases dealing with the compensatory discrimination policy from 1951 through early 1977—an average of just one and a half cases per year. Of the 59 judges who served on the Supreme Court from 1950 through early 1977, 50 sat in one or more of these cases. Because of benches of varying sizes, this amounted to 160 sittings by these 50

judges—an average of only 3.2 sittings per judge. Fifty-two judgments

were delivered— just more than | per judge. (If we exclude the 14 cases which involved only questions of group membership, we find 44 different judges sitting 113 times in 25 cases—an average of 2.6 sittings and 0.84 written opinions per judge.) While some pronounced differences in approach are evident, there are not enough instances to make any quantitative comparisons. No judge has emerged as the champion of compensatory discrimination; Justice (later Chief Justice) Gajendragadkar probably comes closest by virtue of his opinions in Dora and Rangachari, two of the small number of Supreme Court cases in which the scope of preferential treatment was given a broadening, expansive interpretation. A few other judges sat often enough to display a consistent tendency to curtail its operation. Overall quantitative scoring is not very revealing here, for judges’ views tend to vary from issue to issue: the same judge may take a very different stance’ in regard to different aspects of the preference policy. Thus Justice Subba Rao, the only judge who took a broad view of governmental power to make extensive reservations in Devadasan, took a very restrictive view of the State’s power to employ the caste criterion in

Chitralekha. Again, the unreliability of quantitative scoring is increased by the fact that many of these cases involve a whole bundle of issues, so that there is no necessary connection between the doctrine pronounced

on

any

one

point

and

the

outcome

of the

case.

In

Chitralekha,

animadversions against the caste basis were set forth at length, although no state action of this sort was before the Court. On the other hand, in P. Rajendran the Court was quite permissive on the use of caste by the state, but the scheme was struck down on a different ground. In the High Courts, too, few judges have had sufficient exposure to these issues to facilitate quantitative comparison. Even in Mysore, which had a disproportionately large share of these cases (19 in 18 years), the largest number of reported cases sat on by any judge was four and the median was two. Possibly, some High Court judges may have encountered many more such cases at the unreported admissions stage (but interviews with lawyers and judges did not uncover any lore about this).

Judicial Intervention

489

Styles or tendencies of some of the High Courts as a whole can be more readily discerned than those of individual judges. Thus in Kerala there was a reluctant willingness to go along with the state on the composition of the Backward Classes while expressing misgivings and pressing the government for reforms; in Andhra, on the same issues and during the same period, we find a less accommodating strictness on _ the part of the High Court. The homogeneity of judges’ backgrounds and experience and strong pressures for, unanimity are reflected in the infrequency of nonunanimous judgments. Dissenting and concurring opinions are considerably less frequent in India than in the United States. Gadbois reports that from 1950 to 1967, 91% of the judgments in the Supreme Court were unanimous; in the court’s first twenty years, Dhavan reports, there were dissents in only 7.4% of all cases decided. He points out that 70% of the dissents are of a single judge and non-programmatic in character: Dissent is usually on a specific non-recurring matter, or of a token nature, where the judge merely expresses another point of view and having made his

point, makes no attempt to try to get the others to adopt his opinion.5?

In the 39 cases involving compensatory discrimination in which judgments were delivered by the Supreme Court, 50 different judges sat—a total of 160 sittings. Eight of these judges registered dissenting opinions in 6 (16%) of these cases— which suggests that this might be an area of law somewhat more volatile and unsettled than most. This impression is confirmed if we exclude those 14 cases involving only questions of disputed group membership. In the remaining 25 cases, which deal with the contours of the preference policy, we have non-unanimous courts in 6 out of 25 cases (24%)—a rate three times the Supreme Court’s owerall rate. (Of the dissents, five were objections to the Court’s “ex: pansion” of preferences and a sixth was in the same spirit; two objected to undue restriction of governmental power to make preferences.) There is other evidence of division and disagreement. Dissents are even rarer on the High Courts. In part, this reflects the narrower opportunities—a court that disagrees with an earlier court is bound either to follow it or to refer the matter to a larger bench; a divided

Division Bench is required to refer the case to a larger bench. In the 7

preference cases decided by a Full Bench, we find 3 dissents (and 1 separate concurring opinion)— but, of course, these are a very select 50. Gadbois 1970a: 2. Non-unanimous decisions of the United States Supreme Court

increased from 11% in 1930 to 76% in 1957 (ZoBell 1959: 205). Non-unanimous decisions (in full opinion cases) were 66% in 1968, 72% in 1969, and 81% in 1970

(Grossman and Wells 1972: 165). 51. R. Dhavan 1977: 34.

52. R. Dhavan 1977: 32.

490

The Judicial Process

group of cases; the Full Bench was empanelled precisely because of

their difficulty. In the 13 cases in which there was appeal from the

decision of a single judge to a larger bench, we find 4 reversals. And in the 24 appeals from the High Courts to the Supreme Court, we find 9

reversals.53

Strain or change in the law cannot be measured only by dissent or reversal. For structural features— sitting in small and shifting benches

by overworked judges whose short tenure provides few opportunities to encounter these issues and who enjoy few resources to effect a comprehensive synthesis of learning on these matters— increase the possibility of shifts from case to case that are not reflected in these measures. Most change appears not in dissents but within majority opinions, occasionally by overruling of earlier cases,55 but more often by elaborating conflicting tendencies within the corpus of case law. The

fluctuation of the courts with regard to the permissibility of caste units

and the disparate strands in their handling of group membership problems are examples of the way that different approaches may flourish in

spite of the absence of articulated disagreement. In spite of powerful

unifying influences, the Indian judicial system contains features which seem to amplify the normal tendency ofa case-law system to accommodate inharmonious strands of judicial doctrine.

The opportunities for diversification without confrontation have in-

creased over time. As the case loads of the courts have increased, the number of judges has grown. Thus, the Supreme Court’s case load increased tenfold from 1951 to 1971,56 and the number of judges in-

creased from 5 in 1950 to 14 in 1980. Their tenure on the Court has

decreased.57 After 1954, the Court started sitting in benches (at first the minimum mémbership in a bench was three; in 1966, the minimum was

lowered to two).5® The average number of judges sitting in our com-

pensatory discrimination cases has fallen from 5.4 in the 1950s to 4.1 in the 1960s to just 3.6 in the 1970s. B. AUTHORITY AND PRECEDENT: FORMAL STRUCTURE AND OPERATING NORMS Indian courts, like those of other common-law systems, are governed

by the doctrine of precedent, or stare decisis. All courts in India are 53. The 62.5% rate of affirmance is very close to the 60% rate for all appeals from.

High Courts from 1950 to 1967, calculated by Gadbois 1970b: 38. 54. 55. ment 56.

Cf. R. Dhavan 1977: 32. Cf. R. Dhavan 1977: 44. But overruling is more difficult since the 42nd Amendimposed a requirement that reversal be by a larger bench. R. Dhavan 1977: 116-17.

57. See above at nn. 15-17.

58. 1966 Supreme Court Rules O. VII, R. I.

Judicial Intervention 491 bound by the judgments of the Supreme Court, not only on questions of constitutional and central law, but on questions in every field of law, central or state, on which the Supreme Court has passed.59 The courts are bound not only by the narrow holding of the Supreme Court’s

decisions, but also by the broader expressions and obiter dicta of the Supreme Court. That Court, however, is not bound by its own prior

decisions, but it usually treats them as authoritative.! Where there are

conflicting Supreme Court decisions, the decision of the larger bench controls.®? If the conflicting benches are of equal size, the later decision is controlling®— at least in theory. But as we have seen in the handling

of Balaji and Chitralekha, a later decision with a distinctively different

stance may be ignored or assimilated to the earlier decision. Neither the High Court ina state, nor the subordinate courts there, are bound by the decisions of the High Courts in other states, but usually these decisions are considered of persuasive authority unless their own High Court has ruled differently. Within a High Court, a

59. Art. 141. The courts other than the Supreme Court are also bound by decisions of the predecessor Federal Court (1937-1950) and the Judicial Committee of the Privy

Council delivered before Jan. 26, 1950—e.g., Kishan Chand v. Ram Babu, A. 1. R. 1965

AIL 65. This includes obiter dicta of the Privy Council as well as its holdings—e.g., Dass

Bank v, Kali Kumari, A.1.R. 1958 Cal. 530. Judgments of the Privy Council delivered after that date are of merely persuasive force as authorities. Decision of the Privy

Council and the Federal Court are not binding on the Supreme Court, although they may carry persuasive force (Srinivas v. Narayan A.1.R. 1954 S.C. 379; State of Bihar v. Abdul

Majid, 17 S.C,J. 300 [1954]). See, e.g., Gurunath v. Kamalabai, AIR. 1955 S.C. 206 (followed where settled character of law important).

60. Municipal Committee v. Hazra Singh, A. I. R. 1975 S. C. 1087, 1088-89. A lengthy

compendium of High Court citations may be found at Saxena 1963: 208 ff. This includes

advisory opinions (Ram Kishore v. Union of India, A.1.R. 1965 Cal. 283).

61. E.g., see Income Tax Officer v. Nadar, A. 1. R. 1968 S. C. 623; Superintendent v.

Corporation of Calcutta, A. 1. R. 1966 S. C. 997.

62. State of U.P. Ram Ghandra, A. 1. R. 1976 S. C. 2547, 2556.

63. Superintendent 0. Ram Ajodhya, A. 1. R. 1965-Cal. 348; Bhavan v. Commercial Tax

Officer, A. 1. R. 1961 Mys. 3.

64. Subordinate courts in a state are bound by all judgments of the High Court of that state: “omission [to follow High Court precedents] is as much dereliction of duty as

omission to refer to sections of the statute. The disregard of authority is, however,

something still more objectionable. It amounts to an act of insubordination” (Rex v, Ram

Dayal, A. 1. R. 1950 All. 134, 137). Courts are divided on whether subordinate courts are bound by the obiter dicta of their High Court. They are not, according to Vinayak Shamrao v. Moreshwar Ganesh, A. 1. R. 1944 Nag. 44, 541; Baula v. Chief Executive Officer, A.1. R.

1955 Nag. 49, 63. But see, e.g., Basunta Kumar v. Chief Electrical Engineer, A. 1, R. 1958 Cal. 657, 660. 65. Emperor v. Dholraram, A. 1. R. 1941 Sind 221. A court may not follow decisions of other High Courts when binding decisions of its own High Court are available (Bisanchand Lalchand v. Union of India, A. 1. R. 1961 M.P. 220; Ramaswami v. Chandra Kotayya, I. L. R. 48 Mad. 693 [1924]). But a High Court is not bound by obiter dicta of its

own court (Trilochan Choudhury v. Dayanidhi Patra, A. 1. R. 1961 Or. 158; Ram Narain v.

Director of Consolidation, A. 1. R. 1965 All. 172).

492

The Judicial Process

smaller bench is bound by the decision of a larger bench or a bench of equal strength. Thus a Division Bench is bound by a Full Bench or an earlier Division Bench; a single judge is bound by a Full Bench, a Division Bench, or an earlier single judge.® If a Division Bench or a single judge disagrees with a prior decision of a court of equal strength, it should request that the case be referred to a larger bench.®7 A Full Bench can be overruled only by another Full Bench— possibly only by a larger one.® Where there are conflicting decisions of courts of equal strength, it is unclear whether a subsequent court may choose or is bound to follow the earliest decision.® Nor is it clear whether binding authority is retained by those points in overruled decisions which were not decided on by the overruling bench.7° Usually there is a single written opinion in a case decided by a higher court.71 Where there are several opinions of equal strength, as where both judges of a Division Bench deliver separate opinions, they are presumably of equal authority. There is considerable uncertainity about the precedential value of concurring judgments of the Supreme Court.72 Similarly, the pronouncements of dissenting Supreme Court & + Dhanki Mahajan 0. Rana Chandubha, A.1.R. 1969 S.C. 69.

fai Kaur v.Sher Singh, A. 1. R. 1960 S. C. 1118; Mahadeolalv. Administrator-General,

. 1960 S. C. 936; Jaisriv. Rajdewan, A. 1. R. 1962 S. C. 83. There are slips: withina nomth ‘ofeach other twoJ aivision benches of the Allahabad High Court came to opposite conclusion regarding the constitutionality of reservations for candidates from rural and

hill areas. There is no indication that the second (Dilip Kumar, A. I. R. 1973 All. 592, which struck them down) knew of the first (Subhash Chandra, A. 1. R. 1973 All. 295, which upheld them). In Pradip Tandon (A. I. R. 1975 S. C. 563, 565) the Supreme Court

noted reprovingly: “It is desirable from the point of view of judicial propriety to refer to earlier decisions of the same High Court.” ~ 68. Chandhlal v. Babulal, A. 1. R. 1952 M. B. 171. However, in practice, it seems that full benches can be overruled by subsequent full benches (Maktul v. Manbhari, A. 1. R. 1958 S. C. 918).

69. Compare Rama Suddarayalu , Rengammal, A.1.R. 1962 Mad. 450 (court should

follow that which seems more correct) with Januma Rai v. Chandradip Rai, A. 1. R. 1961

Pat. 178 (earlier prevails) with Agurchandv. Deochand, A.1.R. 1960 A.P. 101 (more recent

and more consistent with general principles should prevail).

70. Compare Shri Prithvi Cotton Mills v. Broach Municipality, A. 1. R. 1968 Guj. 124 (decision reversed on point still remains good authority on another point) with Daya

Krishnan v. Assessing Officer, A.1.R. 1966 Punj 490, (no longer good law).

71. In the first few years of its existence, there were some instances in which justices of

the Supreme Court delivered individual opinions. By 1953 it had become the invariable practice to have one justice deliver a single opinion for the court. 72. Doubt about less than majority decisions arises from Art. 145(5). See Shri Prithvi

Cotton Mills v. Broach Municipality, A. 1. R. 1968 Guj. 124 (minority judgment not

binding); Jayantilal v. Rana, A. 1. R. 1963 Guj. 80 (concurring opinion ofa High Court judge not binding).

Judicial Intervention

493

justices which are uncontradicted by the majority may have some weak value as precedents.73 The body of precedent consists of the judgments of the Supreme

Court and the High Courts— formally only of those officially reported,

but in practice those reported anywhere.”* There is no reporting of the decisions of the subordinate courts. Nor are all of the judgments of the higher courts reported. There is no reporting of proceedings at the admissions stage, which are so crucial to writ petition litigation. Judges are empowered to determine which of their judgments should be reported. Many cases which turn on questions of fact or which are felt merely to reiterate established principles are not reported, although the

practice differs in the various High Courts. Inquiries to lawyers and Judges turned up only a few unreported cases in the compensatory

discrimination area, none of which involved any important departure from the reported body of material. Reporting is slow—a year or sometimes more may elapse before the decision is widely available. Indian lawyers and judges have an intense respect for English decisions, and the latter are often cited. Courts do regard such decisions as illuminating and sometimes persuasive guides, but not as binding by

virtue of the doctrine of stare decisis.?5 In the field of preferences, as in many other areas of Indian constitutional law, there is little English precedent to go on. In dealing with constitutional questions, courts and lawyers do look to American (and to a lesser extent to Australian) precedents where they are relevant. Again, such precedents are not

binding and are not always persuasive. The practice of looking to American cases reached its zenith in the mid-fifties and has gradually

been declining since then as a body of Indian precedent was built up. Although “equal protection” is an area in which American learning is

available,7® there is little from America—or at least there wasn’t until 73. Venkatasubramanian (1965) says courts are bound by the views of dissenting judge where the majority did not express any view of the matter. Rayi v. Deputy Tahsildar,

A. I. R. 1963 Ker. 155. Cf. Ranjit Kumar v, State of West Bengal, A. 1. R. 1968 Cal. 551.

74. Although the Indian Law Reports Act, 1875 (Act XVIII of 1875) makes only the

official report a source of binding authority, in practice those unofficially unreported are freely cited—and even unreported judgments are treated as good authority. Cf. Baty

Nath Prasadv. State of Bhopal, A. 1. R. 1957S. C. 494.

75. E.g., Income Tax Commissioner v. Vazir Sultan, A.1.R. 1959 S. C. 814. In some fields

at least, English

law is the residual source of law for Indian courts (Phulchand v.

Hukumchand, A. 1. R. 1960 Bom. 438).

76. Art. 14 (equal protection) is an area in which Indian courts have been receptive to American case law. Cf. the observation of the Supreme Court that: “It may reasonably be assumed that our Constituent Assembly when it enshrined the guarantee of equal protection in our Constitution, was aware ofits content delimited by judicial interpretation in the United States of America. In considering the authorities of the superior courts in the United States, we would not therefore be incorporating principles foreign

494

The Judicial Process

recently —or elsewhere that offers any guidance to Indian courts in the field of preferential treatment.” This is a purely Indian area of the law and one erected from the bare constitutional texts in the last twenty-five years. In expounding the meaning of these constitutional provisions, the courts have been reluctant to go back to the Constituent Assembly

Debates to determine the meaning of the text— though

this reluctance

may be waning.78 (Indian courts do not consult legislative history to ascertain the meaning of particular enactments.)79 One may wonder how the relatively few cases in this field could have generated so much law. Indian courts, unlike their more parsimonious American and British counterparts, tend to decide all—or at least a great many—of the issues of fact and law which arise in the cases before them. For example, where one of several alternative defences is upheld, an Indian court will often proceed to consider the remaining defenses on their merits, even though they are strictly speaking, not needed to reach a decision. The prevailing view is that all of these “alternative trains of reasoning” in a particular judgment have equal binding

authority, even though the case could have been disposed of on only one

ground.®° For example, in Balajt’s case, the Supreme Court found three

to our Constitution or proceeding upon the slippery ground of apparent similarity of expressions or concepts in an alien jurisprudence developed by a society whose approach to similar problems on account of historical or other reasons differ from our own’ (State of Uttar Pradesh v. Deoman, A.1.R. 1960 S.C. 1125 at 1131-32). An eminent commentator observes that Art. 14 “thas been given content by practically taking over the theory of classification developed by the U.S. Supreme Court” (Seervai 1968: 34).

77. American cases are cited to bolster the “classification” argument in State of Kerala

v. N. M. Thomas, A.1.R. 1976 S.C. 490. See discussion in chap. 11, § D, above. Earlier,

purported American analogies had not proved persuasive when put forth in preference cases. See, ¢.g., Sukhdev v. Government of Andhra Pradesh, | Andh. W.R. 294. In Jagan Nath

v. State of Jammu and Kashmir, A. 1. R. 1958 J. & K. 1, American cases were relied upon—but the holding there was directly under Art. 14 because at the time Art. 16(4) did not extend to Jammu and Kashmir. Ironically, one early preference case cited Plessy v. Ferguson, 167 U.S. 537 (1896), for the proposition that the burden lay with the petitioner to show that the classification is unreasonable (Om Prakesh v. State of Punjab, A.

IT. R. 1951 Punj. 93.). 78. See Alexandrowicz 1957, chap. 1; Merillat 1967. 79. State of Mysore v. Bidap, A. 1. R. 1973 S. C. 2555, 2558 (“There is a strong case for whittling down the Rule of Exclusion followed in the British courts . . ..Where it is plain, the language prevails, but where there is obscurity or . . . other special circumstances it may be legitimate to take external assistance”); Anandji Haridas & Co. v. Engineering

Mazdoor Sangh, (1975) 2S. C. J. 415.

80. Even where the case could have been disposed of on one point. See, e.g., Jaiwant Rao v, State of Rajastan, A. 1. R. 1961 Raj. 250; Reddiv. Savitramma, A. 1. R. 1961 A.P. 305;

Janardhana Rao v. Deputy Transport Commissioner, A.U-R. 1965 A.P. 115. But it is not unknown

to read the previous case so that one train is dictum and not binding (Ram

Narain v. Director, A. 1. R. 1965 All. 172).

Judicial Intervention

495

independent grounds upon which the Mysore provisions violated the Constitution: first, the exclusive reliance on caste in measuring social

backwardness; second, the inclusion among the Backward Classes of groups that were only relatively backward; third, the magnitude of the

reservation. The courts that have cited Balaji have considered each of these pronouncements of binding authority. The corpus of authoritative doctrine is enlarged by another feature of precedent in India: the assignment of greater weight to the dicta of the Supreme Court than to the holdings of the High Courts. The Constitution commands that all courts apply the law “‘as declared by the Supreme

Court”’.®1 This is usually taken to include obiter dicta, a view that has not

completely restricted the flexibility of High Courts in handling Supreme Court dicta.82 The more assertive High Court judges find little difficulty in sidestepping Supreme Court pronouncements; the less bold, however, are likely to take the matter as settled by a Supreme Court dictum. This orientation to the peak of the system is reflected in lawyers’ perceptions and responses. They tend to be relatively unobservant about the doings of other High Courts. For example, the pre-Balaji litigation in Mysore clearly invalidated a number of arrangements widely employed elsewhere, especially in southern India. But there was no initiative taken elsewhere until the Supreme Court intervention, which was quickly followed by litigation in several states. The Indian judicial structure, as we have seen, is more unified, more

pyramidal, and more heavily weighted at the top than, for example, its American counterpart. First, unlike the American, the unified Indian

judiciary deals with all questions of state as well as central law.*3 There is no exclusion of issues of local and state competence as disputes move up in the judicial system. Generally there is less narrowing of the issues as the case proceeds to higher levels in the system. The tendency for many issues in a case to remain open, and the generosity with which 81. Art. 141. A view taken by a majority on which several judges reserved opinion is law declared by the Supreme Court and overrides earlier decisions of the court (Bhavan v. Commercial Tax Officer, A. 1. R. 1961 Mys. 3). 82. Compare, e.g., Kalinath v. Nagendra Nath, A. 1. R, 1959 Cal. 81, 83, where the Supreme Court dictum was followed even though the decision there was on other grounds, since “the Supreme Court Justices took a definite view on this point . . . itis binding on us,” with Pardhasaradhi Rao v. Srinavasa, A.1.R. 1959 A.P. 512, 516, where the

court carefully distinguished the Supreme Court case as “only authority for what it actually decides and the generality of expressions in it are not intended to be expositions of the whole law but are governed and qualified by the particular facts of the case... .””

See Popkin 1962: 247 ff. 83. On the Forty-second Amendment’s attempt to constrict Supreme Court jurisdiction over state law, see n. 4 above.

496

The Judicial Process

{ndian judges decide points, make for doctrinal profusion at the same time that the system is institutionally unified. The broad appellate jurisdiction of the Supreme Court and the multi-

plicity of appeals make that Court a potent instrument of unification

and integration.®4 The centralizing tendencies are augmented by the relatively closely circumscribed recruitment practices and the lack of localizing devices (like juries, locally elected judges and prosecutors, local financial responsibility for courts and police).85 The top-heaviness is emphasized by the great social and professional distance between peak and lower agencies and by promotion on grounds of seniority and conformity. Like other formal structures in India, the lower echelons are severely constrained in their capacity to innovate and to accommodate new situations.® (This top-heaviness is expressed in the reluctance of many lower court judges to hear elaborate legal arguments on the ground that any innovation is a monopoly of the High Courts.) All questions can be reopened at higher levels;87 there is little apportioning of power to make final decisions in different matters to subordinates at

various levels. The competence of the higher authorities is viewed, not as another

specialized

skill, supervisory

or managerial,

but as a

generalized potency which encompasses all of the skills of the subordi-

nate staff. It is as if the lower courts were mere conduits through which the legal potency of the higher courts flows. In spite of their broad jurisdictional scope and the multiplicity of

appeals, the role of the Supreme Court and the High Courts in making

law uniform and predictable is undermined by a number of features: sitting in benches, high judicial turnover, overwork, and slow reporting.

Discrepancies in the case law are attributable not only to ordinary

judicial vagaries, but to the constant turnover of personnel who decide similar cases, sometimes unaware that a similar issue has been decided

differently elsewhere. Judges are overworked, having little time for 84. The Supreme Court’s jurisdiction extends to cases where the High Court certifies

that the case involves substantial constitutional questions; if refused, the Supreme

Court may grant special leave to appeal (Art. 136).

85. Cf. another centralizing factor, the right of the State to appeal acquittals in criminal cases, discussed in Law Commission 1958: II, 810 ff. The austere absence of overt channels through which local concerns and attitudes can gain expression may not

be unrelated to the prevalence of covert devices by which the law is bent to local views.

86. A similar feeling about the relation of subordinates and superiors informs the Indian notion of “due process” or reasonableness which includes having a sufficiently high

official make the decision. See, ¢.g., Dwarka Prasad v. Stateof U.P. (1954) S.C. R. 803. 87. Cf. Gledhill’s (1964: 13) observation that English procedure has been so adapted

in India that “it seems based on the assumption that the lower courts are incompetent todo anything more than prepare a record of evidence and that the real contest is in the court of first appeal.”

Judicial Intervention

497

reading and reflection (and, as we shall see, little critical stimulation). A

distinguished leader of the bar observed:

Those practicing before the [Supreme] Court know that on an average probably each of the Judges has to prepare in the course ofa week about four judgments apart from his sitting in Court hearing cases and reading a number of miscellaneous matters during the five working days in a week. In

the result the Court is often unaware of its own decisions and is apt to make inconsistent pronouncements.®°

This was epitomized in an incident in 1959 when on the same day two three-judge benches of the Supreme Court encountered the same question (of construction of a section of the Industrial Disputes Act) and decided it in opposite ways.®9 Notwithstanding their hierarchic organization and the formal doctrine of precedent and authority , the Indian courts diverge in practice from the model of tight pyramidal hierarchy in which legal rules are declared or fashioned precisely and unambiguously at the higher levels and applied automatically and inexorably at subordinate levels. It is a useful counter to this hierarchic image to imagine the courts as a collegial body whose members, by virtue of selection, socialization, habits of deference, and common

work pressures, share affinities of

outlook and style. The collegial image should be meaningful to anyone who has ever served on a committee made up of like-minded (ifindividualistic) peers

which has to make a lot of decisions in a short time—e.g., an admissions committee. Decisions tend to,get made on the basis of a limited set of factors, often those conveniently found in the file, rather than by individualized weighing of all the particularities of the case after a fresh and exhaustive investigation. With insufficient time for everyone to brief himself thoroughly about each issue, there is an inclination to go along with what seems sensible and workable and to accept some variability among different sub-groups. Similarly, both formalism and tolerance for inharmonious strands of doctrine can be seen as a reflection of the work situation of the judges, who are asked to make quick and imponderable decisions on the basis of the limited materials supplied by the legal culture.

88. Setalvad 1969: 1.

89. The appeals turned on widely divergent readings of the term “gratuity.” In Brachmachari Research Institute v. Their Workmen, A.1.R. 1960 S. C. 257, the workers received only gratuity, since “gratuity in the context of the award was held to mean nothing more than retrenchment compensation.” But in Indian Hume Pipe Co. v. Their Workmen, A. 1. R. 1959 S. C. 1081, the workers received both gratuity and retrenchment compensation on the ground that these were distinct benefits meant to redress different burdens on retrenched workers.

498

The Judicial Process

In spite of all the unifying tendencies, there are many factors tending to allow the doctrinal system to be overloaded with conflicting principles, offering great leeway for subsequent courts and great opportunity for selective reading of the corpus of authority. In the compensatory

discrimination area, where we start from conflicting principles, the

institutional setting permits the co-existence of competing themes without overwhelming pressure for definitive resolution. This doctrinal indeterminacy accentuates the passive or reactive position of the courts.91

Whaf they do depends very much on the cases that parties bring to the courts and the arguments that lawyers make to them.

C. THE PATTERN OF LITIGATION With the exception of election appeals, almost all of the cases in the field of preferences have originated in the High Courts or in the Supreme Court under their special jurisdiction to issue writs, orders, and directions for the protection of Fundamental Rights.9? The writ

jurisdiction is the most striking legal innovation in independent India.93 Parties are permitted to approach the highest courts directly with

a petition that the court either invalidate legislation or set aside an

administrative order that contravenes the petitioner’s Fundamental Rights. The writ jurisdiction has proved extremely popular with litigants, and lawyers exercise all of their ingenuity to put their client’s claims in the form of a writ. The number of writ petitions in India is immense: by 1956 the total in all of India was about 15,000.% In 1973 nearly 4,000 were instituted in Karnataka alone.°5 The term “writ” has entered popular speech; the client who has no inkling of the substantive

provision under which he seeks redress will insist to his lawyer that he

requires a “writ.” Its popularity is not difficult to understand: writ petitions are expeditious and they are relatively cheap. The filing fee is almost nominal, compared with the high ad valorem court fees which 90. In the High Courts and Supreme Court, cases are argued, decided, and reported in English. Varying standards of English probably increase the distortion in communication, blunting precision and screening out subtleties. However, the effects of this are more pronounced below the level of the High Court. 91. See Black 1973. 92. Arts. 32, 226. On this jurisdiction and its use, see Moothan sion 1958: II, 657 ff.

93. Alexandrowicz 1956: 35-45: Moothan 1963. 94. Law Commission

1958: II, 659-60. Moothan

1963; Law Commis-

(1966-67:261) observes that in all

of England the annual number of applications to the High Court for prerogative writs is about thirty. Gledhill (1964:12) gives an even lower estimate of twenty.

95. I. L. R. 1973 Kar. (Annex) viii.

Judicial Intervention

499

must be paid in advance in an ordinary civil suit or appeal.9* Lawyers’ fees (the costliest item in such litigation) and- other costs are lower because the expensive process of presenting evidence is eliminated?”— the facts being established by affidavit. Starting at the top; there is less likelihood of being embroiled in subsequent appeals. Perhaps the most attractive feature of the writ is the direct access that it affords to a high and respected authority. Although the writ jurisdiction has put an enormous burden on the higher courts, most observers credit this device with inspiring. a widespread feeling that redress is available against governmental improprieties. (On the other hand, not only are the courts inundated, but the widespread use of writs may undermine the authority of other institutions—e.g., universities.)

The procedure for hearing writ petitions differs from court to court. Some High Courts hear all writs by a Division Bench; others hear them by a single judge, with a possibility of appeal to a larger bench. In all courts there is some kind of screening process—an admissions stage in which writ petitions are examined to see whether they should be heard on the merits. A large majority —one knowledgeable lawyer estimated 80% —are eliminated at this stage. Those eliminated include cases involving disputed questions of fact. In writ proceedings no evidence is taken. In theory the facts are determined by the affidavits of the parties. Generally, writs are not admitted if there are serious questions of fact. However, in our compensatory discrimination litigation, we observe

that courts vary considerably in their willingness to undertake ex-

tended factual inquiries.

96. The Indian judiciary, as Alan Gledhill (1964: 8) has quipped, “‘is the most successful of the nationalized industries.” The court fees paid by civil litigants support the entire judicial system and in many states yield a surplus. Avoidance of civil remedies is by no means confined to writ petitions; there is massive use of the criminal law to settle what might be regarded as “civil disputes” — thus avoiding the delay and high court fees. Cf. the observations of Walsh (1930: 26) that “the Indian much prefers, if he can seek its assistance, the criminal court for the redress of his civil wrongs.”

97. Under Indian practice, lawyers’ fees may be awarded to the winning party as part of his costs. But fees are allowed according to a schedule and do not necessarily reflect actual fees charged. For example, a winning petitioner in a writ case in a High Court may be awarded Rs. some rough estimates 98. Thus, in Kajari that “this Court in its

100, although the of lawyers’ fees in Saha v. State, A. 1. Constitution Writ

lawyer may charge much more than that. For this kind of litigation, see §D, below. R. 1976 Cal. 359 at 369, the judge emphasizes Jurisdiction is not a Court of facts. This Court

does not sit in appeal over the decision of an authority whose order is challenged in this

proceeding.” But cf. Sardool Singh v. Medical College, A. 1. R. 1970 J. & K. 45, at 47: “In

order to test the validity of [petitioner’s] argument [about the proceedings of the selection committee] we called for all the relevant records of the Selection Committee

showing the marks fetched by the candidates at the last qualifying examination and

those awarded to them at the interview by the Selection Committee. After an examina-

500 The Judicial Process In virtually every instance, it is the government or a public body that

is the opposite party, since the Fundamental Rights, with few excep-

tions, run only against the government

(in the broad sense, including

public corporations, nationalized banks, universities, and other compo-

nents of the growing public sector) and do not afford protection against

the restriction ofthese rights by private action.9?

The use to which the writ jurisdiction is put varies in its emphasis froni state to state. For example, in Mysore a high percentage of writ petitions were “service matters” concerning the respective rights of government servants; most of these are questions arising out of the 1956 reorganization of the state and the consequent amalgamation of ser-

vices from previously separate areas. Service matters are less prominent

in other High Courts. In Bombay, for example, it is estimated that about 80% of the writ petitions are concerned with rent and tenancy of agricultural land; in effect, the writ petition has transformed the High Court into a court of appeals from a non-appealable land tribunal. The number of writ petitions concerned with the policy of preferences is insignificant in every Indian state, even in Mysore, which has had more

writs concerned with this subject than any other state.

Courts in India are viewed with a curious ambivalence; they are

simultaneously fountains of justice and cesspools of manipulation.?© Litigation is widely regarded as infested with dishonesty and corrupt manipulation. But courts, especially the High Courts where this compensatory discrimination litigation takes place, are among the most respected and trusted institutions.1°? Compensatory discrimination

tends to display the benign face of the Indian legal process. It is relatively inexpensive and quick; it focusses on genuine disputes—

there are few distracting false issues. And with the partial exception of

some

election

cases,

there

is little occasion

for the

fabrication

of

evidence. The other face of Indian litigation peeks through * in the

tion of these documents, we find that by and large candidates have been selected for

admission to the Medical College purely in order of merit.

... A perusal of these

documents would further reveal that the Selection Committee did not try to plump up

marks ofa particular candidate so as to tilt the scale. . . .”

99. Shamdasani v. Central Bank, A. 1. R. 1952 S. C. 59. It remains an open question

whether the courts would fashion writ remedies to protect against violations of those

Fundamental Rights which run against private individuals rather than against the State (Articles 15[2], 17, 23).

100. Since there are no records kept of writs filed but not admitted, it is impossible to

say how many more attempts there have been. Interviews with lawyers suggest there were not many unadmitted cases in this field. 101. Kidder 1973, 1974; Morrison 1974. Cf. Galanter 1972b.

102. Gadbois 1977. 103. E.g., in Narasimha Reddy v. Bhupathi, 31 Election L. R. 211, 219-20 (1967), the judge refers to evidence that was “artificial and improbable” offered by a witness who

Judicial Intervention

501

controversies over group membership and is fully exposed in some of the election controversies. Virtually all of the litigation about compensatory discrimination has involved reservations, even though preferences in the form of provisions of facilities, resources, and protections directly affect a much larger number of recipients. (See table 25.) A rough indication of this concentration is that in 113 reported litigations about compensatory discrimination which I have found,’ only 12 have concerned matters other than reservations—chiefly land acquisition and debt relief. This litigation has, by and large, been initiated, not by the beneficiaries of protective discrimination, but by others complaining of schemes which affect their interest. Thus of the 113 cases, if we put aside the 32 battles over reserved seats,1°5 we find that only 27 of the

remaining 8] cases involved claims brought to the courts by would-be beneficiaries and that 54 were initiated by those challenging the existence or operation of the preference. As tables 25 and 26 suggest, litigation about compensatory discrimiTABLE 25 Supyect Matter oF RePorTED LITIGATION ON COMPENSATORY Descrnanation 1950-1977 Cases

Subject Matter of Preference Education

Government Employment

Elective Office Other Tora.s

Involving Reservation Device 41

26

Not Involving Reservation 1

1

31 3

1 9

101

12

Total :

42

27

32 12 3

would “not lift his head and was looking down all the time while deposing as if he was ashamed of telling lies.” Indeed, the court concluded that “‘his evidence is nothing but a tissue of lies.” And in Rajgopal v. Anomugum, 1972(2) Mys. L.J. 240, 253-54, the judge refers to one side’s “‘belated inventions for the purposes of defence” and a bit of “‘tampering” with some documentary evidence by the prevailing party. 104. On this universe, see n. 45 above. 105. Electoral cases are omitted from this calculation because almost all of them are group membership problems in which rival claimants from the beneficiary group dispute as to the boundaries of that group. It is difficult to identify those claims in terms of the classifications used here.

502

The Judicial Process

nation has not been spread evenly over all applications of that principle.

It has concentrated heavily on the use of the reservation device. But it

has not been spread over all uses of reservations. Instead, litigation has focussed on a few kinds of opportunities. There are a few persistent scenarios.

The most commontof all litigants in these cases is the prospective student seeking admission to a higher technical or professional college—in most cases a medical college—who is excluded because of what he regards as an unwarranted reservation. Of our 42 education cases, 41 concern reservations in medical and engineering colleges in 11

states. (These are in turn only a fraction of an even larger number of “admission writs” that attend the yearly selection of candidates to these coveted institutions. As one judge in Mysore, where litigation on this has been thickest, put it, the admission writs had become an “annual

feature” in that High Court.)! Although some are field by single petitioners, these petitions often come in large batches, so that these cases involve many hundreds of students. The largest in Mysore in-

volved 473 petitioners. Although all members of the affected class might eventually benefit from a revision of the government’s order, interim relief is available only to the individual litigant.1°7 Thus those inTABLE 26 Inrriatine AND SuccessFut Parties in ComPENSATORY Discrnanation Lrrication 1950-1977

Initiating Party Seeks To Subject Matter of

Preferential Treatment

Education Government Employment

Other Tora.s

Overturn, Restrict,,

Curtail the Government’s

Preference

Enlarge, Restore, or

Amplify the Coverage

of the Government's Preference

Successful

Unsucessful

Successful

Unsucessful

12 10

12 10

4 0

14 7

5 27

5 27

0 4

106. Viswanath v. Government of Mysore, A. 1. R. 1964 Mys. 132 at 134.

2 23

107. Some High Courts have followed the English practice requiring separate writ applications from each party, while others have allowed joint or common applications in cases where the claimed right to relief arises from the same act or transaction and there is a common question of law or fact (Annam Adinarayana 0. State of A.P., A.1.R. 1958 A.P. 16). Such joint or common writs have not so far played a prominent part in litigation about compensatory discrimination.

Judicial Intervention

503

terested in obtaining admission in the current year would be inclined to file a writ petition of their own, and typically many coattail-riding petitions are filed whenever a college admission is under attack. All of these admissions cases are concerned with reservations by the

state governments rather than the Centre, And virtually all of them are concerned with reservations for the Other Backward Classes, not for the

Scheduled Castes and Scheduled Tribes. As table 26 suggests, those student petitioners seeking to overturn or curtail schemes of reservations have enjoyed a high degree of success.1% (Their 50% rate of success score is higher than that for any classes of parties in Gadbois’ survey of Supreme Court litigation.) As we have seen, they have

succeeded in establishing a variety of limits on the designation of Backward Classes, the extent of reservations, and the standards for the

operation of preferences.

Not all of the professional-college litigation is initiated by nonbeneficiaries. In 16 of our cases, it was initiated by putative be-

neficiaries who claimed that government policy. wrongly confined the scheme or excluded them from it. Not only were they less frequently successful, winning only 3 of 16 cases, but their successes were of a limited and correctional nature, holding the government to its announced policy rather than broadening that policy.11° A less numerous class of litigants are competitors for government jobs who feel that they have been unfairly denied employment or promotion by unfounded or excessive reservations. Actually, there has been surprisingly little litigation in regard to initial recruitment into government service. In 24 years there has been only a single reported case concerning initial recruitment into central government service! Of the 27 government employment cases that I have found, 19 were concerned with reservations in promotions (and a promotion element seems to be present in at least 2 others). Of the remaining 8 cases, 5 involved the post of munsif (minor civil judge); that is, the applicants were lawyers, presumably more inclined and better equipped to litigate. (At least one of the promotion cases also involved a legally trained petitioner.) Thus 108. Gadbois (1970b: 48) calculated the rates of success for claimants under various

Fundamental Rights provisions. Overall, the individual claimant won in 36 of 120

decisions involving the Constitution’s equality provisions (Arts. 14-18). 109. This was the same as the average for all Fundamental Rights claims. Gadbois also has rates of success for different classes of parties (1970b: 46). Only employees and unions with a 55% success rate did better in contests with the government than our medical applicants. For example, political candidates prevailed only 21% of the time in contests with the government.

110. The three successful “‘affirmative” cases are identified in n. 126 below. The list

there includes a fourth success (Mohan Rao), but that is a group membership case and

hence excluded from the present calculation.

504

The Judicial Process

the employment cases may be viewed less as communities fighting over entrance into government service than as government servants (and lawyers) fighting over opportunities for advancement. The battle focus-

ses not at the top, where reservations have little impact on the chances of others, nor at the bottom, where they are hardly needed (and the

others might not easily afford litigation), but in the middle— particularly in the upper ranges of Class III and the jump from Class III to Class IJ.111 Naturally such litigation clusters where reservations are present and most effective—e.g., within the central government in those offices which put into practice the provision for reserva-

tion in promotions. Eleven of our 13 central government cases concern

the Railways, Posts and Telegraphs and the Central Secretariat. All of the litigation about central schemes concerned reservations for Scheduled Castes and Scheduled Tribes, but 9 of the 14 cases about

state schemes concerned reservations for the Other Backward Classes.

Not all of these were attempts by governments servants to fend off reservation. In 7 cases, putative beneficiaries took the initiative and attempted to get courts to extend (in 2 cases, reinstate) preferences. None were successful. The government servants, on the other hand, were successful in limiting the extent of reservations and eliminating

practices unfavorable to them.1!2 In several states they succeeded in knocking out reservations for Other Backward Classes (at least tempo-

rarily), and in a number of states they brought about major revisions in state policy, including severe curtailment of the scope of preferential

treatment.

A third and much smaller and scattered group of litigants are those licensees, creditors, or property owners who object to particular government schemes for backward classes (debt relief, land acquisition) that

interfere with their property or business. There were only 10 of these cases, all concerning various state schemes, of which 5 were successful.

No pattern is evident here.

Finally, there are the litigants in electoral cases—usually losing candidates, but in a few cases would-be candidates whose nomination papers were rejected because of dispute about their membership in the 111. Cf. our earlier estimate (chap. 4, §B, above) that it is precisely in Class III that reservations are most effective and it is only there that their effectiveness is declining. It is not suprising that this is where the serious resistance is to be found. 112. Comparison with Gadbois’ analysis of Supreme Court litigation provides a

rough indication of just how successful they have been. He found that the overall success

ratio of government employees litigating with the government was less than one in three (1970b: 53); claimants under the various equality provisions of the Constitution prevailed in only 30% of cases (1970b: 48). Claims of the type considered here are then a successful sub-group of both these classes.

Judicial Intervention

505

preterred group. Three cases involved the question of reservations as

compartments or guaranteed minimums; in all, the courts decided in

favor of the mimimum. All the other cases involved claims of group

membership. Again, no pattern is evident in the gross outcomes.

Except in the election disputes and a few cases involving protections (debt relief, etc.), the government is the opposite party. In many cases, where appointment or admission of others would be affected by the decree, it is necessary for the petitioner to join them as respondents.

This is part of the reason that litigation is more popular when the

number of others affected is small (e.g., in promotions rather than, for example,

an annual

recruitment of clerks by the state government,

which might involve hundreds of potential parties) .11> Most of the litigant success has been at the expense of the state

governments rather than the Centre. If we set aside the electoral cases,

we find that in the remaining 81 cases, government loses (wholly or partly) in 30. But in only 3 of these is it the central government that loses. (Two of these 3 were High Court judgments on limited technical matters of administering preferences;14* only the single Supreme Court case, Devadasan, was of prime importance—and its major hold-

ing is now in doubt.)115 By and large, the courts exercised little direct control over preferential treatment by the central government. On the

other hand, the courts struck down 27 arrangements (20 for medical admissions) in eight states. Twenty of these cases concerned preferen-

tial treatment for Backward Classes. Litigation about preferences for Scheduled Castes, though less fre-

quent, was spread fairly evenly about the country. But litigation concerning preferences for Backward Classes was, in the 1950s and early

1960s, largely a South Indian phenomenon. (See table 27.) But after 196 the pattern changes, and Backward ‘Classes issues are litigated in the North as well—a trend which the recent surge of reservations in northern states promises to continue. The general pattern that we have detected in compensatory discrimination litigation emerges with even greater clarity if we consider only that litigation which reached the Supreme Court.

As table 28 indicates, all the cases involving the Other Backward

113. A further reason encouraging litigation about munsifs is that typically the

number of successful competitors is small, so that it is feasible to make all of them respondents (in contrast to clerks, where there may be hundreds of affected respondents

who are difficult to track down because of transfers, etc.).

114. Director General v. Natarajan, 1971 M. L. J. 79; Sudama Prasad v. Divisional Superintendent, A. 1. R. 1965 Raj. 109. 115. Deoadasanv. Union of India, A. 1. R. 1964S. C. 179.

506

The Judicial Process

TABLE

27

Recronat Disrripution oF LivicaTion aABouT (OTHER)

BACKWARD CLASSES

Four Southern States

7d

wt

oww

1951-55 1956-60 1961-65 1966-70 1971-75

Rest of India wVsSrn-o

Period

Note: The units here are litigations, as defined in n. 45, above. Times are dates of the final decision.

Classes arose from state schemes; virtually all of the litigation about

Scheduled Castes and Tribes arose from central government schemes. The cases decided by the Supreme Court deal almost exclusively with three kinds of disputes: (1) problems of reservation by state governments in South India for admission of Other Backward Classes to professional (especially medical) colleges (10 cases); (2) problems of reservation of promotion posts in the central services (especially Railways TABLE

28

Susyect MATTER oF Supreme Court LitiGaTION ABOUT Compensatory Discrnanation 1950-1977 EDUCATION

GROUP

SC/ST oBc Both Toras

GOVT. POSTS

Reservationsin Other professional

Note: State schemesin(

OTHER

TOTALS

Initial ~Promo- Group Other Recruit- tions © Memberment ship

Colleges

—2(2) (7) 3(3) 12(12)

ELECTORAL

0 0 0 0

0 0 22) 2(2)

6(2) 3(3) 00) 9/5)

140), NOV"), 0 0 0 0 0 0 14(0)—sE1(O)}, S11),

2365) 10(10) 5(5) 38 (20)

).

* This case (Dora) is listed in group membership column—hence adjustment of totals. >The total here (38) is one less than the total for Supreme Court cases given earlier. In dealing with judicial participation, the two Triloki Nath opinions were treated as two units, but in dealing with litigations that case is counted as a single unit.

Judicial Intervention

507

and the Central Secretariat) for Scheduled Castes and Tribes (5 cases);

(3) problems of group membership raised by challenges to elections to

legislative seats reserved for Scheduled Castes and Tribes (14 cases).

Thus, these three strands of litigation account for 29 of the 38 cases in which the Supreme Court has encountered the compensatory discrimination policy. Considering the varied and diverse government programs of comnsatory discrimination, how can we account for this clustering of itigation around these few problems? Both electoral and medicaladmission situations involve high stakes and claimants with status and resources (including, in the electoral cases, organizational backing). The government service promotion cases involve high stakes for those concerned— government servants in the middle grades—and often common stakes around which group support can be organized. In all of these cases there is at stake a substantial interest in some lifeopportunity for the petitioners. Few, if any, cases are deliberately brought as test cases—that is, for the primary purpose of establishing a principle rather than advancing the interests of an individual litigant.

But these cases do sometimes become a focus of group or communal

struggle.116 A lawyer may develop a reputation as a champion of one particular interest, providing an informal focus for organized effort. In several of these cases we find low-grade government servants supporting expensive litigation that appears to be beyond their individual means,?7 and in fact such litigants are often assisted either by organized associations or by ad hoc informal groups of supporters. Occasionally, organized communal associations have taken the initiative in instituting litigation, but more typically they have intervened in cases which have already been filed by interested litigants. For example, in Kerala, in the Jacob Matthew case, the Nair Service Society provided financial assistance to those challenging reservations, while the Sree

Narayana Dharma Paripalana Yogam, an Ezhuva association, hired a

leading lawyer to argue in favor of the reservations. A somewhat different pattern of intervention is found in Balaji’s case, where several

communal associations formally intervened, attacking the reservation on the grounds that they had been wrongfully excluded from the Backward Classes. Although such formal iitervention is relatively rare in 116. Sometimes there may be struggles within the government about reservation

policy and litigation strategy. The extension of reservations at the promotion stage by the Railway Ministry and the conduct of the subsequent litigation (Rangachari) became a cabinet-level dispute between the Home and Railway Ministries. |

117. The petitioner in Rangachari v. General Manager, A. 1. R. 1961 Mad. 35, rev'd A. I.

R. 1962S. C. 36, was supported by an informal association of his colleagues. The cost of this litigation was estimated by knowledgeable Indian lawyers as Rs. 3,000-4,000.

508

The Judicial Process

Indian litigation, it seems to be slightly less rare in these compensatory discrimination cases.118 None of this intervention has been by “universalistic” associations remote from the immediate parties or by disinterested amici curiae who seek the establishment of an abstract principle of law. There are, as we saw above, instances of participation by pre-existing and multi-purpose communal organizations, built around traditional ties. There are also instance of participation by organizations of a more “modern” type (e.g., the associations of government servants of a particular grade). An intermediate instance is provided by composite communal interest groups which are bound together specifically by an interest in the

reservation or other schemes of preference—for example, the Central

Government Scheduled Castes and Scheduled Tribes Employees Welfare Association, which formally intervened in the C. A. Rajendran case; or the state-wide association of Backward Classes, which took the ini-

tiative in attempting to restore the Andhra Pradesh reservations by litigation; or the Kerala Harijan Samskarika Kshema Samity, at whose urging the Kerala government issued the extensions that led to the Thomas litigation.119 Some associations of beneficiaries of preferences are little more than

paper organizations or mere personal organs. But others, especially

associations of Scheduled Castes and Tribes government employees,

are enduring organizations with a wide membership base—although the centtal government has adamantly refused to recognize them.120 118. From figures supplied by Dhavan (1977: 34, 107) it appears that intervenors were allowed by the Supreme Court in only 301 reported cases from 1950 to 1970, only 14.3 per year and 7.4% of the 4,098 reported cases during that period. In contrast, of the 39 reported Supreme Court cases about compensatory discrimination there were formal interventions in 6 (15.3%). 119. A. I. R. 1976 S. C. 490 at 494. There are also organizations specifically for the purpose of opposing reservations. These include such ad hoc groups as the “‘antireservationist” organizations that were formed in Bihar in 1978, but also more enduring and specialized groups such as associations of government employees other than Scheduled Castes and Tribes, who have agitated against reservations in the services. 120. In 1975 the Commissioner for Scheduled‘Castes and Scheduled Tribes (1973-

74:129) noted that the government refusal to recognize employces associations based on caste, tribe, or religion was misapplied to “service associations of Scheduled Castes and Scheduled Tribes associations as these are not actually based orf any particular caste,

tribe or religion.” He held that the claims of these associations should be judged on their merits, which were considerable, since Scheduled Castes and Scheduled Tribes emp-

loyees had distinctive interests in the working of the special safeguards and concessions in their behalf. These interests were not adequately protected by the general employees associations: “The Commissioner. . . has not come across even a single case where a general employees’ union took up the case of implementation of the orders i reservations and concessions for Scheduled Castes and Scheduled Tribes.” Neverthe.

less, a “High Power Committee” on reservations in the Central Services chaired by the

Judicial Intervention

509

Such associations have been active in making representations in indi-

vidual grievances and pressing collective demands by lobbying and demonstrations, but have only infrequently resorted to litigation. For example, in early 1978 the All-India Scheduled Castes and Scheduled

Tribes Railway Employees Association was supporting a case to dis-

qualify a non-Scheduled Castes person who had obtained a railway job with the help of a false caste certificate.121 Although they had a host of grievances about the situation of Scheduled Caste and Tribe employees—ranging from administration of reservations, interviews, transfers, and quarters, 10 non-recognition of their organization to the excessive complexity and ambiguity of the orders— limitations of financial resources and legal imagination left them unable to visualize the courts as a forum for securing redress. Although some of the participating associations have strong political

links ands some of the contesting groups are reputed to be linked with

contending political parties, it is difficult to discern any general political lines in the fights over reservations. Reservations have been challenged by Communists and defended by Swatantra leaders and vice versa. Oc-

casionally, the battles do take on a party tinge, but many more seem to

be intra-Congress fights. The propensity to litigate these questions depends on a number of factors. One is clearly the feeling of being deprived ofa scarce opportunity. Another is the question of expense: the cost of obtaining that opportunity by litigation must appear to be less than getting it by some other means. It is estimated that in Mysore, where the alternative toa

state medical or engineering college is a private college with a high

capitation fee, litigation was regarded as a much less expensive means of obtaining admission than payment of the fees in a private college.

The tendency to have very large numbers of writ petitions whenever a reservation is challenged can be explained in part by decreasing marginal cost. Once a writ petition is filed, additional petitioners can usually arrange to be parties at much lower cost. Lawyers typically reduce their fees when there are a number of writ petitioners and arrange to process their petitions on a mass basis. Also, junior lawyers will file writ petitions along the line of one already filed by a senior, for a very much

Prime Minister herself later that year decided against such recognition on the ground that recognition of separate associations would “give rise to internal friction and would not only prevent their absorption in the mainstream but would also lead to unhealthy repercussions amongst other employees who might also start clamouring for recognition of other similar groups or associations” (RCSCST, 1974-75: 81). 121. Interview, New Delhi, Feb. 1978. Several cases in which there were complaints

about use of false certificates in connection with government posts were pressed by Scheduled Caste employees’ associations (RCSCST 1974-75: 261).

510

The Judicial Process

lower fee. Again, the publicity attendant upon such litigation tends to draw more petitioners in, and the success of one petition encourages

others to try this mode of obtaining resources.

For the propensity to litigate depends, too, on the expectation of success. Let us take the negative case. One of the striking things about the nationwide pattern of litigation is that after the 1951 amendment of the Constitution there was no reported litigation about reservations in Madras until 1967. Sizable reservations for the Backward Classes in

both employment and profgssional-college admissions existed in Madras;

the Backward Class lists Were formulated on the “caste basis” and thus

were arguably open to attack on the basis of the Balaji case. Yet no litigants came forward to challenge the Madras scheme. In part this may be due to the fact that the Madras reservationsare more modest than those that formerly obtained in Mysore, and thus the “advanced communities” do have substantial opportunities to enjoy the resources involved. Lack of contest in regard to government employment might be attributed in part to the shift in ambition among Madras Brahmins away

from government

service toward

the private sector. But this

would not explain the absence of contention about reservations in professional and technical colleges, which are the prime preparation for advancement in non-governmental employment.12? It appears, then, that the communities not on the backward list regarded the reservations as good as it was possible to get under the circumstances. There was a feeling that litigation would not be effective to prevent some kind of preference for the “Backward Classes.” As one lawyer put it, “[iJf they get rid of it in one way, the majority communities would spare no efforts to introduce it in another way” —a remark that must be placed in the context of the existence of selection committees*23 as well as of reservations. That is, it was felt that the courts could not provide ultimately effective relief. This might explain the lack of organized backing, since no widespread and permanent benefit could be anticipated, but it does not explain the lack of individual litigants, since these might gain some tangible advantage even though the excluded communities did not secure any lasting benefit.1* Litigation, it seems, requires not only parties who would benefit but an example or model of success and an incident to serve as a catalyst.

122. This may be due in part to the growth of alternative private medical colleges in which a capitation fee might be paid. There were more of these in Mysore than in Madras, and even with this safety valve, there was far more litigation in Mysore. 123. Cf. the discussion of selection committees in chap. 13, §A, above. 124. Apparently, there had been a few admission writs in the preceding years and the

Backward Classes argument had been raised but never decided (letter trom G.

Ramaswamy of the Madras bar, 30 Nov. 1968).

Judicial Intervention

511

That is, present discomforts, possible gains, and acceptable costs and risks of litigation must be made vivid and palpable.125 In Madras the model was provided by Balaji and the spate of similar litigation in other southern states. The catalyst was proved in 1967 by the peculiar circumstance that the marks list in the medical school competition had been destroyed. This was widely publicized, and as one of the lawyers described it, a large number of disappointed applicants, hoping that the entire selection might be struck down, filed petitions. The paucity of “affirmative” litigation by the beneficiaries of preferences may be explained in part by lack of resources. But it also reflects the absence of such “models of success”— there has been virtually no successful affirmative

litigation126— and of “catalytic events” to trigger the attempt.127 Summing up, we may surmise that the gross effect of litigation on the compensatory discrimination policy has been to curtail and confine it. Those who have attacked compensatory discrimination schemes in court have compiled

a remarkable

record of success,128 while those

seeking to extend compensatory discrimination have been less successful. One way of expressing this gross effect is to restate table 26 by comparing the level of prefereatial treatment before and after judicial intervention.

Thus, we see that after litigation the level of preference was about the

same as that previously propounded in 49 of the 80 cases; it was less in 27 scoring probably understates the Courts, reluctant to strike down arrangements while warning of the

or administered by the government cases and more in only4 cases. This curtailing effect of the judiciary. government schemes, will uphold necessity of correcting unconstitu-

125. In Mysore, the catalytic event was provided in 1958 by the issuance ofa new Backward Classes Order after the state’s 1956 reorganization. The new Order imPinged on groups outside old Mysore who were not previously covered. In Andhra, the event was the conjunction of Balaji’s case with the state government's expansion of preferences. In Kerala and elsewhere it was the decision in Balaji’s case with its promise of successful challenge of existing policy. . 126. Raghuramula v. State of Andhra Pradesh, A. 1. R. 1958 A.P. 129, established the important principle that reservations could not be used to limit the admission of backward classes and has been much cited. Guntur Medical College v. Mohan Rao, A.1. R.

1976 S. C. 1904, established a group membership principle of deep theoretical import, but one that few others would be in a position to invoke. The other two successes (Ashok

v. Dean, 69 Bom. L. Reporter 603, and Susila v. State of Madras, A.1.R. 1970 Mad. 399)

secured technical corrections in administration which benefitted the backward class applicants but did not significantly broaden the scope of the preferences or establish any general principle. 127. The two affirmative cases with a broader thrust were triggered by “catalytic

events” —the withdrawal (in Dasa Rayudu v. Andhra Pradesh Public Service Commission,

A.I.R. 1967 A.P. 353) or curtailment (in C. A. Rajendran v. Union of India, A.1.R. 1968S. C. 507) of previously existing reservations.

128. See n. 112, above.

512

The Judicial Process

tional practices, thus coaching the government to introduce changes favorable to the petitioners. Looking back over the pattern of litigation, we see that the central focus of the judicial encounter with the compensatory discrimination policy has been, not the general outline and performance of that policy, but rather the way in which government’s attempts at redistributing certain highly valued scarce gpportunities have impinged on the chances of other competitors with the resources and incentive to raise such challenges. The initiative has lain largely with these competitors. Thus the beneficiaries of compensatory discrimination have been at a disadvantage in the development of the law; the competitors have occupied the strategic heights in the litigation battle. The working out of principles is inevitably shaped by the contingencies of the factual situations which formed the context in which they are formulated. It is

the most expansive schemes and the most flagrant abuses which have

come to court. The first round of litigation (the 195] cases) came up

from the old Madras system of communal! quotas rather than from a scheme specifically designed to advance the backward classes. Subsequently, it was the Mysore situation, which was little removed from communal quotas, that produced litigation and Supreme Court law. Debate on backward classes policy has been strongly affected by the fact that the crucial and determinative litigation has come out of Mysore, where reservations for Backward Classes were the most flagrant and expansive. Had some state with a more modest scheme of preferences been the source of the important litigation, the law might

have taken a somewhat different shape.

More generally, it should be noted that litigation (and public debate)

has

tended

to center almost

exclusively around

schemes

for Other

Backward Classes rather than the more extensive schemes for Scheduled Castes and Scheduled Tribes. In only two areas has there been any significant litigation about the latter: problems of group membership and problems of extent, particularly in regard to promotions. That is, the litigation in regard to Scheduled Castes and Scheduled Tribes has concerned the boundaries of the preference policy. All of the other litigation has concerned provisions for the Backward Classes, and here the thrust of the litigation has been to question the basic policy of compensatory discrimination as well as its borderline applications. Thus the law of Articles 15(4) and 16(4) is being made largely in reference to situations involving the Backward Classes. The courts have

erected restrictions on these articles to counter excessive and unwarranted preferences for the Backward Classes. But these redound to restrict what the government may do for the Scheduled Castes and Scheduled

Tribes.

Similarly,

in public debate and scholarly com-

Judicial Intervention

513

mentary, the well-publicized abuses of the Backward Class category easily attach to the resentment against the preferences for the Scheduled Castes—a confusion compounded by the widespread inability of the public to distinguish these groups (i.e., to distinguish the inclusive and narrower senses of “[b/B]ackward [c/C]lasses”. Thus Scheduled Castes and Tribes may be indirectly “‘misrepresented” by spokesmen for the Other Backward Classes. More generally, the configuration of parties in this litigation may fail to make it a forum in which all the relevant interests can be heard. Chitralekha provides an example. In the High Court proceeding, Judge Hegde upheld the state scheme, while finding serious deficiencies in its failure to do enough for those who should have been beneficiaries. The appeal was taken by the petitioners, who wanted the state order struck down—i.e., their criticisms were precisely in the opposite direction from Judge Hegde’s. The state on other hand wanted to maintain its Order. So both parties were

hostile to his observations about the deficiencies of the state’s scheme:

there was no one to press the interests of those he felt wrongly exclu-

ded.129 The absence of adversary control may be responsible for the

conceptual muddle that emerged.1° Although this is an extreme case,

it is not difficult to believe that there may be a similar tendency in cases where the contending parties are the state and non-baekward objectors

to the state’s schemes for the backward. We have seen how the normal leeways of a case-law system are

amplified and structural factors limit the ability of the courts to achieve coherence. In such a setting, what courts do depends in large measure

on the kinds of claims that are brought to them and on the kind of arguments that lawyers make. D. LAWYERS Between courts and litigants there stand the lawyers; it is the lawyers

who translate the grievances of litigants into the claims presented to courts. Notwithstanding recurring complaints that India’s legal system is unsuitable to her conditions, courts and lawyers are a familiar and much-utilized part of Indian life from great public issues to village disputes.131 India has a numerous legal profession— both absolutely 129. Havanur 1965: 119.

130. See chap. 7, §A, above. Similarly, delicate problems of group membership with subtle implications for the integrity of beneficiary groups have arisen in election cases where there may be no effective participation by these groups and where argument may

be dictated by considerations of expediency for the candidate. Meshram is a striking example. See chap. 9, §E, above. 131. See Galanter 1972a.

514

The Judicial Process

and relatively.132 A recent estimate places the number of lawyers at 228,000—approximately 336 for each million people, a ratio far larger than in other Third World countries.433 In spite of differences of region, language, caste, and religion, these lawyers share a common legal culture, which they put into the service of a wide variety of interests. As middlemen between official policy and local interests, they disseminate official norms, rephrase local concerns in acceptable legal garb, and devise new organizational forms for forwarding these interests. In modern India, where local concerns and interests increas-

ingly find expression by representation at centers of power rather than by enjoying a sphere of relative autonomy, the lawyers are crucial agents for the expression of diverse local and parochial interests at the same time that they rephrase (and perhaps re-form) these in terms of official norms. Thus in cases involving preferences, it is the lawyers who rephrase the claims of caste and stratum into constitutional categories. Almost without exception, all of the lawyers involved in these cases,

like most Indian lawyers, have been trained in India. None that I encountered had joined the bar in England, but some had post— graduate legal training or teaching experience, including a few who had some exposure to American legal education. The basic educational background is that of the Indian law college. Except for some desultory apprenticeship requirements, formal legal training in India takes place within universities as a two-year course (recently become

three-year) after the bachelor’s degree. Admission to law colleges is more or less unrestricted. The law colleges are crowded with students who have neither a professional nor a scholarly interest in the subject; only a small proportion go on to practice. There are some imaginative and innovative teachers, but for the most part teaching proceeds by following a rigid syllabus and concentrates on the rote learning of rules. The syllabus is not geared to current problems, nor to legal subjects appropriate to an expanding welfare state or a developing economy, but

is with rare exceptions confined to the standard subjects of half a century ago. Most law colleges, especially in the South, are part-time operations, staffed by practicing lawyers who serve as part-time teachers. There are some full-time law colleges, some of better quality. But except for some very recent innovations, these too are characterized by overcrowding, 132. The rough portrait of the legal profession here is drawn from personal observation combined with Rowe, Bastedo, Morrison, and Schmitthener and the Conference proceedings reported in Galanter and Kidder 1968-69, and the papers by Morrison (1972), 133. Trust, 134.

Khare (1972), Kidder (1974). Personal communication from Dr. N.R. Madhava Menon, Bar Council of India March, 1982. On legal education, see Bastedo 1968-69; Sunshine and Berney 1970.

Judicial Intervention

515

obsolete curriculum, rote learning, inadequate facilities, heavy teaching loads, and an atmosphere which is not congenial to creative re-

search or scholarship. The full-time teachers in these colleges have little contact with the practicing bar. They are, with few exceptions, not held

in high esteem by the profession, nor regarded as repositories of special expertise in their fields, nor, with the rarest exceptions, used as con-

sultants or specialists in professional work. The lawyer learns his trade, not at the law college, but by associating

with the “court community” at the courts. The profession is over-

crowded and competition is stiff. Work is unevenly distributed: “leading lawyers” are extremely busy and hurried while a large number of aspirants have very little to occupy them. Building a successful practice usually requires sufficient resources to wait out the early, lean years. Many with legal training take to other tasks after an unsuccessful fling at the profession. The risk and insecurity ofa career as practicing lawyer have combined with increasing opportunities elsewhere to make law

less attractive than it once was to the ambitious. There has been a

general decline in the prestige of the legal profession in recent years and a drop in the quality of talent that is drawn into the profession. 35 The leading role of lawyers in resisting Emergency Rule during 1975-77 may have boosted their position in public regard, at least temporarily. The profession is a unified one, with a single set of qualifications for

practicing in any court or engaging in any aspect of legal practice.36 Yet both the public and the profession concur in visualizing the lawyer primarily in the role of courtroom advocate, rather than business advi-

sor or negotiator, much less social planner. Although there is no formal solicitor-barrister distinction437 most lawyers see themselves as barristers whose principal function is arguing in court rather than advising their clients. Typically, clients appear at the lawyer’s door at a relatively late stage in the dispute, when they are already intent upon litigation. Ties with clients tend to be episodic and ad hoc, not enduring. Rarely do lawyers keep extensive records that would be helpful in later

cases.

The lawyers tend to be oriented to the courts, rather than to other legal settings. This orientation to courts is vividly displayed in spatial 135. On the decline in the prestige of the bar and the quality of entrants, see Schmitthener 1968-69 and Bastedo 1968-69; on the ambivalence of lawyers to their profession, see Kidder 1974. 136. On the older distinctions, see All-India Bar Committee

(1953); Schmitthener

(1968-69). 137, With the limited exception of the High Courts of Bombay and Calcutta where the solicitor-barrister distinction is observed on the “Original Side” jurisdiction over major cases arising in those cities.

516

The Judicial Process

terms—lawyers

are literally to be found at the court when it is in

session. The office, rarely separate from the home, does not serve as a

staging area for operating in various arenas. The lawyer’s business is usually at a court. He will not move with his client’s case from court to court, but will more often pass the matter on to a lawyer attached to the appropriate court. Typically, the lawyer spends his working life at a particular level of the system. But in spite of the stratification that this entails, the profession is relatively undifferentiated in character. The profusion of appeals, the tendency to have little narrowing-down of the issues in higher courts, and the latter’s broad original jurisdiction leads to courts at all levels dealing with the same kind of problems. Although those at the higher levels enjoy more prestige, lawyers at each level do

much the same sort of thing. Within each level lawyers are stratified by

skill, influence, prestige, and wealth, but there is little division of labor by specialization (beyond civil-criminal) and little coordination in the

form of partnerships and firms (though established and successful lawyers may have one or more juniors, as well as clerical help). Basi-

cally, all lawyers offer the same kind of service under conditions of chronic oversupply.

The

emphasis

on

litigation and

the courtroom

role reinforces

lawyers’ rule-mindedness. Where the lawyer’s task is to win disconnected battles, rather than to pattern relationships, there is little to

induce the practicing lawyer to go beyond the kind of conceptualism that is characteristic of most Indian legal scholarship and that pervades legal education. In his court-centered world, the lawyer works in a

realm of rules with little attention to either the social policy underlying them or the problems of implementing them. In preparing his cases, the lawyer has available a plethora of treatises, mostly mechanical compilations by professional hacks in the pay of legal publishers. There is a smaller body of work prepared by law teachers and practitioners, some of it of excellent quality, but almost invariably confined to scholarship of a traditional type, with heavy

emphasis on close textual analysis and rather less consideration of

either underlying policy or the actual working of legal regulations. Although expressed in an idiom of deference, and in spite of a widespread feeling that judges are sensitive to criticism, outspoken critical

assessment of the courts is not uncommon.

The law regarding compensatory discrimination is not seen to consti-

tute a field or a subject in its own right. In addition to passages in the

standard works on the Constitution, there is a small body of literature produced by academic lawyers (and a little by practitioners who have been involved in preference cases) but much ofit is in publications seen

by few members of the bar. Occasionally, lawyers in preference cases

Judicial Intervention

517

have ransacked government reports and standard ethnographic works

tor evidentiary points, but there has been no attempt to marshall the

empirical data that might be relevant. For practitioners, and critics as

well, the attempt to grapple with the basic principles and practical

effects has been hampered by the emphasis on textual criticism and by the absence of any tradition of assessing law in empirical terms. All of the lawyers involved in cases concerning preferences are lawyers whose practices tenter around the High Courts. There are no specialists in litigation about preferences as such. Although most High Court lawyers do some writ petition work, there are some specialists in writ petitions, whose practice may be made up almost completely of cases of this type. Some of these writ specialists have done much of the

litigation in the preference field. There are a few instances in which a

lawyer has become identified as a champion of one or the other side of the preference issue in a series of cases, most notably one writ specialist who has become famous as a persistent and redoubtable opponent of reservations in Mysore. On the other hand, a somewhat larger number of eminent lawyers have appeared on both sides of these issues. Thus the lawyer who successfully obtained a writ striking down the Andhra reservations subsequently argued a number of writs petitioning for their restoration. Where, as commonly happens, there are a number of writ petitions considered together, leading lawyers may handle several or even a large number of such. petitions. There is at least one instance of a lawyer handling so many admissions cases that he had cyclostyled copies of the

petition into which he filled the names of students. When a petition challenging some arrangement which affects a large number of persons,

such as a college admission, is filed by a senior lawyer, other lawyers

often file similar or identical petitions, and these are consolidated for a hearing. At the hearing, the most “senior” lawyer—that is, the one

most esteemed at the local bar—will ordinarily give the principal

presentation of the arguments, and the other lawyers will add any

additional points if they wish.

Lawyer’s fees are steeply graded according to their eminence at the bar. A leading lawyer at the Supreme Court may charge as much as Rs. 4,000 for handling a writ petition; his counterpart at the High Court in Karnataka may charge Rs. 1,500 to 2,000 (some High Courts are repu-

ted to be more expensive, some less). The average medical-admission

writ in Karnataka might involve lawyers’ fees of Rs. 250. Where there are a number of identical writ petitions, fees are ordinarily reduced for the multiple petitioners. Junior lawyers, riding on the coattails of a senior lawyer’s petition, may charge a little as Rs. 100 to their clients. In these cases, Government is ordinarily represented by its Advocate

518

The Judicial Process

General or one of his assistants. These government lawyers, it should be

noted, are not specialists in this kind of litigation. The administrative departments concerned do not have their own legal staffs; nor do the

general legal offices of the government have any specialists with expertise in the administration or enforcement of these compensatory

discrimination policies.8 In important cases, the government sometimes retains eminent members of the bar, especially for appeals. In some cases, too, private parties are joined to the government as respondents and appear with their own counsel; some arrangement is then

made between the Advocate General and the private lawyers as to the division of labor in the argument.

Most lawyers come from the higher castes—as do most members of the educated classes. Therefore it is not surprising that most of the lawyers in these preference cases are of high caste origin. Many come from families with a long legal tradition; but several prominent in these cases are first-generation lawyers. In a number of cases the caste origin

of the lawyers corresponds with a personal or a communal interest of their clients. But in a larger number of cases there is no such correspondence. There are occasional complaints that untouchables sometimes have difficulty in engaging leading lawyers, but the untouchable

side has been represented by leading members of the bar in several

cases. It is often the case that lawyers on both side are Brahmins. In these preference cases, some lawyers have become engaged

through communal or political or kinship networks. Sometimes such

considerations play a part in selecting a lawyer who is not so connected—a state may prefer an out-of-state lawyer in the Supreme Court in a touchy case or may engage an eminent Brahmin lawyer

from the state to defend its preferences which are detrimental to Brahmins. On the whole, professional standing seems to play a larger role than any other factor in the selection of counsel. Only a few lawyers

have made a “cause” of this kind of case. Some are quite partisan on

these questions, but many are generous-minded, understanding, and capable of considerable detachment on these issues. On the whole there is much less vehemence and bitterness among the lawyers than there is

among other participants.

The lawyers’ method of work places a great deal of emphasis on oral

skills. There is little reliance on written arguments. No briefs are submitted; there is only a skeletal memorandum. Oral argument, how-

ever, is elaborate and often extended. Oral argument in Balaji’s case 138. This follows the general pattern of lack of specialization and initiative in

enforcement tasks in government legal work. E.g., the Law Ministry of the central government has a drafting department and a legal advice department, but no staff

specialising in the enforcement of laws in specific policy areas.

Judicial Intervention

519

took up two full days. the course of the development of the law on

preferences has been determined in part by lawyers’ considerations of

strategy. This strategy has been in the main dictated by the considerations of advantage to the litigants rather than any attempt to establish abstract principles. Thus in Meshram’s cas¢, arguments were advanced which would have compromised the whole Buddhist movement, for the purpose of securing a client his seat in the legislative assembly. On

the other hand, responsiveness to clients may sometimes encourage

counsel to omit a stronger line of argument. For example, in Balaji, the

state took a hard line and attempted to defend the total reservation

although Delhi counsel had strongly advised conceding that the state’s

expanded Backward Class list (including Lingayats) was invalid in the hope that they could save the bulk of the list and the principle of reservation on a caste basis. Questions of reservations in services or admissions to educational

institutions present some special tactical difficulties. Basically, the writ

jurisdiction permits two kinds of challenges: first, one may assert that the individual’s rights are violated and seek a remedy for that violation; second, one may attempt to strike down the whole government order. In some cases only the first kind of relief has been requested and not the

second. Where the appointments or admissions of others would be adversely affected by the decree, it is necessary to join all of the parties who have benefitted by the order that is being challenged. This makes government employment cases even more difficult than admissions cases, and it is what accounts for the concentration of government employ-

ment litigation around posts like munsifs where a small number are involved. Such a case is much easier to handle than a case in which there

is a very large number of persons who are affected. The alternative is to forego immediate relief for the petitioners, securing instead a favorable ruling which will apply to subsequent government action. Thus in

P. Rajendran, the Supreme Court allowed the proceeding without join-

der of the 1,100 admitted medical students, since the petitioners dis-

claimed any desire to disturb the current year’s competition. 144

139. In another respect this may have been one of the rare instances in which the pations of counse! shunted even the interests of the client into the background.

See chap. 9, §E, above. 140. Cf. Ramesh Chander Garg v. State of Punjab, A. 1. R. 1966 Punj. 476, where

petitioner was ordered to be admitted in the subsequent year.

141. P. Rajendran, with its prospective relief, and Guntur Medical College v. Mohan Rao,

where respondent was assured of his place no matter what the outcome of the state’s

appeal, come close to being “test cases” in the American sense. But in each instance the case was initiated and pursued by an individual seeking (and achieving) a tangible and

individual benefit.

520

The Judicial Process

Timing may be a crucial element. Especially in the case of college admissions, there is a danger that if an appeal is necessary, the whole matter may be moot before there is a decision. Another tactical consid-

eration is whether to fight the case in the High Court or to go directly to the Supreme Court. The latter course requires a much greater financial

outlay and, usually, a greater degree of organizational support to provide it. Lawyers have gone to the Supreme Court directly when they anticipate a less favorable response from their High Court (as in Madras?) or find the High Court is unable to control the state government (as in Jammu and Kashmir). In Mysore, the High Court had given several decisions that would have led one to expect a favorable reponse to an attack on the government's reservation policy. However, counsel was convinced that a change in the bench that was admitting

writ petitions lessened his chances of success, so he took Balaji directly to the Supreme Court.

Lawyers have succeeded in translating into contestable claims the

individual and group grievances of those discomfited by compensatory discrimination and in utilizing the writ proceeding to curtail and limit it. Fhey have shown less evidence of creativity in bringing into court the inchoate claims of the lowly for enlargement of preferences orprompt, vigorous, and sympathetic implementation. The profession has } provi-

ded an agency for addressing the sins of commission—the“

thing” and “too much” criticisms—but has had less to offer to those

who are aggrieved by the sins of omission. services in India are supplied on a fee-for-services basis, usually by single practitioners to individual clients already embroiled in a dispute. It is a system that emphasizes litigation rather than counselling, negotiating, or preventive lawyering. And even this limited range of services is available only to those who can pay for it. There has been virtually no “legal aid” in practice. The long-standing but unfilled commitment"? to provide legal aid to the poor was brought to center

state during the 1975-77 Emergency Rule: indeed the Forty-second Amendment introduced a new Directive Principle that the State shall 142. On the pre-history of legal aid, see Koppell 1966; Murphy ef al. 1973: 281 ff., and the 1974 report of the Expert Committee on Legal Aid (Justice Krishna Iyer, chairman), devastatingly assessed by Baxi (1975). The most extensive governmentsponsored legal aid until now has been various state schemes for Scheduled Castes and Scheduled Tribes. The desultory ifnot illusory quality of this aid is suggested by the fact that eight states for which information is available provided Rs. 136,700 for legal aid for

Scheduled Castes (including Scheduled Tribes in one of the states) for 1970-71. Since these states had a combined total Scheduled Caste population of 24.7 million, this works

out to a provision of about one two-hundredth of a rupee per head. Nevertheless, only

58% of the total provided was actually expended (Expert Committee on Legal Aid 1974: 110).

Judicial Intervention

521

“provide free legal aid . . . to ensure that opportunities for securing justice are not denied to any citizen... .”!3 A National Committee on Juridicare reported in 1978 and recommended a massive program of legal aid, including “‘public interest litigation and class actions” and in particular “activist,” “militant legal aid programmes” for Scheduled Castes and Tribes. Such activist programs, were they to come about, under government or other auspices, might stimulate lawyers to prosecute a whole range of claims on behalf of the ostensible beneficiaries of compensatory discrimination and put the courts in a very different posture toward these policies than they have stood previously. 143. Art. 39a. 144. Ministry of Law, Justice and Company Affairs 1978, § 12.01, § 11.06.

I5

The Uses of Judicial Action A. THE IMPACT ON POLICY AND ADMINISTRATION Tne

vecrsions

of courts

do

not

automatically

bring

about the state of affairs that they prescribe. A growing body of research

has indicated that many factors may influence the effect that a judicial decision has on its environment.’ Or perhaps we should say environments, since there are many environments and many effects—on the parties in the case, on others in more or less similar situations, on

interest groups of various kinds, on government administrators and

policy-makers, on other courts, and finally on various sections of the public. The impact of a decision on any of these may vary indepen-

dently of its impact on the others.

There is usually not a great deal of difficulty in assuring relief to the individual petitioner in litigation about compensatory discrimination. Sometimes merely filing a writ petition suffices to obtain relief for him—a petitioner challenging a government order may be given his

admission or job and the petition withdrawn.? Such settlements bet-

ween the government and the petitioner do not become part of the reported case law. However, they sometimes induce the government to revamp its practice in the future in order to avoid similar petitions. Or they may simply provide a way for the government to buy out a complainant and avoid the necessity of shifting its policy—a tactic that can work only so long as the complaint is one that cannot readily be emulated by others. If the case goes to judgment, the court ordinarily enters an order in regard to the admission, promotion, etc., of the petitioners involved.

Since the higher court enters its own order rather than remanding to lower courts, there is no problem of evasion or distortion of higher court

1. The literature on the impact of the decisions of the United States Supreme Court is reviewed in Wasby 1970; Becker and Feeley 1973. 2. Thus in Meera Bai (unreported, High Court of Madras, 20 Sept. 1967), a series of

writ petitions was withdrawn upon the government's discovering that some additional seats were available and agrecing to reassess the qualifications of the petitioners.

Uses of Judicial Action 523 decrees by lower courts.3 The form of the remedy ordered varies: sometimes it is just to reconsider the petitioner’s case, at other times it is to admit him.* Sometimes the court goes beyond this and not only orders relief for the parties before it but strikes down the whole regulation. Since courts are reluctant to upset an entire selection when the successful competitors are not before it as parties, they have often resorted to “prospective overruling”— the decree affects future competitions but the present selection is left undisturbed.5 This has been extended to protect petitioners who were admitted consequent to securing the striking down of a state order by the High Court. When the Supreme Court reversed, restoring the state order, the admissions pursuant to the High Court decision were left intact.® In several instances the courts have gone beyond simply striking down an unconstitutional scheme, by decreeing remedies that actively guided the government toward the construction ofa scheme that passes constitutional muster.” In Periakaruppan v. State of Tamil Nadu, the Supreme Court ordered that state to constitute a new selection committee of specified composition to complete the year’s selection. In an elaboration of “prospective overruling,” the Kerala High Court in Hariharan Pillai, finding the state’s scheme defective and in need of a fresh investigation,® decreed that the present scheme would not be 3. Discussion of this phenomenon in the American setting is found in Murphy 1959; Harvard Law Review 1954. 4. Thus the petitioner in Dilip Kumar v. Govt. of U.P., A. 1. R. 1973 All. 592, 596, was ordered admitted to a medical college four years after he was wrongfully rejected. In Sukhvinder Kaur v. State, A. 1. R. 1974 H.P. 35, 41, the court ordered the medical college to

add a seat to accommodate the petitioner.

5. P. Rajendran v. State of Madras, A. 1. R. 1968 S. C. 1012. The prospective overruling aspects of Rajendran are discussed by Blackshield 1968: 200ff. The problem of devising relief that does not disturb admissions

already made

is surveyed

in

Surendrakumar v. State of Rajasthan, A. 1. R. 1969 Raj. 182. In the employment area, cf.

Venkatramana v. State of Madras, A. 1. R. 1951 S. C. 229. And note the Court’s effort to avoid striking down the selection in B. N. Tewari v. Union of India, A. 1. R. 1965 S.C.

1430. 6. State of Andhra Pradesh 0. Balaram, A. 1. R. 1972 S. C. 1375, 1400. Cf. Guntur Medical College v. Mohan Rao, A. 1. R. 1976S. C. 1904, where the state, appealing from an adverse

decision below, guaranteed the candidate a seat whatever the outcome.

7. Cf. the Constitution’s provision that the Supreme Court may “pass such decrec or order as | and for doing complete justice in any cause or matter pending before it” (Art 142 [1]). 8. Periakaruppan v, State of Tamil Nadu A. 1. R. 1971 S. C. 2303 at 2311. See also Meera Bai v. Director of Medical Education (High Court of Judicature at Madras, 20 Sept.

1967), where as part of a settlement agreement the state agreed to constitute a “High Power Committee” to assess the claims of the petitioners and others on the waiting list in the light of principles embodied in Supreme Court decisions. 9. Hariharan Pillai v. State of Kerala, A.1. R. 1968 Ker. 42 at 51.

524

The Judicial Process

valid after a certain date. And in Triloki Nath, the Supreme Court

directed the High Court to carry out an investigation and report back.?°

The government, which is the respondent in a writ petition proceeding, is ordinarily responsive to the order of the court. Almost invariably, it removes those specific features of its scheme which the court has found objectionable or carries out specific directions of the court. Thus in Kerala the government established an investigatory commission as ordered by the court and implemented the committee’s recommendations. In Mysore, the government successively modified its Backward Classes order to conform to a succession of High Court judgments and then to the judgment of the Supreme Court in Balaji. However, compliance does not necessarily imply a shift in underlying policy. In Mysore relatively prosperous and politically dominant groups had been using the provisions for Backward Classes for their own advantage. Although the state in 1963 dropped its explicit provisions for these groups, it adopted two new practices which produced much the same result without violating the commands of the courts. First, they adopted the income/occupation test of backwardness, which gave considerable advantage to these groups. Second, they adopted a system of interview marks, which permitted a tincture of communal

considerations to be

introduced while ostensibly adhering to secular and non-ascriptive criteria. Thus policies favoring these groups became less visible, but it is not clear that they were less effective, or that the state adhered more

closely to the Court’s admonition that the state direct its aid to the “really backward.”11

Thete have been instances of sustained resistance, the clearest in

Jammu and Kashmir.!2 The Constitution’s Fundamental Rights were made applicable to that state in 1954. In 1958 the High Court held “void and illegal” a cabinet order authorizing appointments “to re-

move the communal disparity,” on the ground that this did not fall

within the “reservations of posts” permitted by Article 16.4 Nevertheless, the state maintained its system of communal quotas in distributing government posts.1* In 1966 a challenge of this arrange10. Triloki Nath Tiku v. State of J. 8 K., A.I.R. 1967 S.C. 1283. 11. On the effects of the income test, see chap. 8, §G, above. Of course, the requirement that preferential treatment be confined to the “really backward” was relaxed ifnot abandoned in Chanchala v. State of Mysore, A. 1. R. 1971 S. C. 1762. See discussion in chap. 12, §D, above. 12. A less dramatic instance of state inaction is found in the wake of Nanda Kishore

Sharma v. State of Bihar , A. 1. R. 1965 Pat. 372. A commission to revise the state’s list of backward classes was established seven years after the decision holding the list unconstitutional (Blair 1980). 13. State of J. & K. v. Jagar Nath, A. 1. R. 1958J. & K. 14. “14. The conflict over medical admissions in Jammu and Kashmir took a somewhat

different course. In early 1966 the High Court struck down a scheme of communal

Uses of Judicial Action

525

ment was taken directly to the Supreme Court by two teachers seeking to quash promotions to posts of headmasters. After laying down the requirements for reservations under Article 16(4), the Supreme Court directed the High Court to gather necessary material on the population and backwardness of the several communities, to ascertain the criteria

being used by the state, and to submit a report within two months.15 Informed by the material gathered by the High Court, the Supreme Court in April 1968 struck down the state scheme as outside the scope of Article 16(4): In effect the State policy ... was a policy not of reservations of some appointments or posts; it was a scheme of distribution of all the posts community-wise.16

The state was invited to devise a scheme for reservations for Backward Classes as allowed by Article 16(4). But, as the Court later recounted,

“no such scheme as was envisaged was ever devised.”’'” Instead, the state officials “thought of an ingenious device of giving ostensible effect to the decision of this court. Those teachers . . . whose promotions became illegal in view of the pronouncement of ... [the Supreme Clourt . .were ‘allowed to work against the posts they were holding

prior to their reversion’ on temporary basis. . . . A new nomenclature was evolved for the post of Head Masters. They were called Teachers-

in-charge. . . . [Each of the reverted teachers thus retained] his emolu-

ments and his posting Other teachers who all promoted to the decision in complete The Supreme Court

as Head of an institution. . . .””18 had not been parties to the earlier petition “were gazetted cadre ...subsequent to the previous defiance of the law laid down by this Court.”19 dismissed as “wholly untenable and miscon-

ceived” the contention that those who were not parties to the earlier

decision were not governed by that decision.

The Judgment which was delivered did not merely declare the promotions granted to the respondents in the writ petition filed at the previous stage as. unconstitutional but also laid down in clear and unequivocal terms that the quotas (Lalita Shuri Tikku v. State of J. @ K., A. 1. R. 1966 J. & K. 101). The state instituted a system of selections by a selection committee with broad discretion to give marks for the interview, but the selection was set aside because the committee’s failure

to take academic merit into consideration was “patently unreasonable, arbitrary and subjective” (Jalali v. Principal, A. 1. R. 1967J. & K. 106). The operation of the selection committee was scrutinized by the High Court again in 1969 and this time it passed muster (Sardool Singh v. Principal, A. 1. R. 1970]. & K. 45).

15. Trilok Nath Tiku v. State of J. & K. [I], A.LR. 1967 S.C. 1283.

of Jammu and Kashmir (11), A. 1. R. 1969S. C. 1, at 4, 16. Triloki Nath Tiku v. State 17. Makhan Lal Waza v. State of J. & K., A. 1. R. 1971S. C. 2206, at 2208. 18. Id., at 2208-9. 19. Id, at 2209.

526

The Judicial Process

distribution of appointments, posts or promotions made in implementation of the communal policy was contrary to the constitutional guarantee of Article 16. The law so declared by this court was binding on the respondent

State and its officers and they were bound to follow it whether a majority of the present respondents were parties or not to the previous petition.20

Consequently, all of the promotions made pursuant to the state’s communal policy were “illegal and unconstitutional” and ‘“‘would have to be revised and reconsidered and appropriate orders must be passed by” the state.21 Almost a year after the first Triloki Nath decision, the state set up a one-man Commission of Inquiry. composed of P. B. Gajendragadkar, the author of the Balaji judgment, who had retired as Chief Justice of India in 1966. The second Triloki Nath decision was handed down while

the Commission was sitting. In its report, submitted in November 1969, the Commission recommended employment of multiple criteria of backwardness and urged that the state appoint a high-powered committee to draw up a list of Backward Classes. Accordingly, a Backward Classes Committee chaired by the retired chief justice of the High Court was appointed in February 1969 and reported in November of that year. While the committee was sitting, the state had essayed the evasive maneuver in the case of the headmasters that came before the Supreme Court in Waza. By the time this was struck down in February 1971 the state had issued new reservations rules and had used an elaborate

interview procedure to fill the now vacated headmaster’s posts—

arriving at appointments bearing marked resemblance to the previous selection.22 These were challenged in Janki Prasad, which upheld the designation of Backward Classes, with minor modifications, but found

the interview procedure entirely unsatisfactory, “‘a travesty of selec-

tion,” and set it aside as unconstitutional.23 The state was ordered to

undertake another selection in which work records as well as interviews would be taken into account. It was seven years since the teachers had first filed their petition and fifteen years after the “communal quota” was struck down in Jagar Nath. But it is clear that Jammu and Kashmir had moved, slowly and grudgingly, toward alignment with the national norms on reservations. 20. Id., at 2209. 21. Id., at 2210. In June 1971 the Prime Minister encountered a deputation protesting “the reversion of 350 odd teachers and others following a Supreme Court judgment” (Hindu, 30 June 1971, p. 8). 22. “[W]hen previously the appointments were made on the communal basis, 178

posts had gone to Muslims and 134 to Jammu Hindus. Now after selection, 177 posts go

to Muslims and 134 to Hindus” (Jankt Prasad v. State of J. & K., A.1.R. 1973 S.C. 930 at 934). 23. Id., at 936,

Uses of Judicial Action

527

Communal quotas couldn’t be arranged openly, and it was increasingly difficult to arrive there by tacit means. But the courts do not always have the last word. The Supreme Court’s decision in Dora, permitting both seats in a double-member constituency to be captured by contenders for the reserved seats, “‘administered the major shock” leading to Parliament’s abolition of the

double-member constituencies.% And there was, of course, the early and dramatic instance of a constitutional amendment to reverse a judicial decision: Article 15(4) was added to counter the Dorairajan

decision and legitimate preferential treatment outside the government employment area.”* But even there the status quo ante was not restored; the post-amendment State policies were more constricted than those that prevailed earlier.

On the other hand, there have been instances in which a state used a

court decision as the occasion for a striking change in policy that went far beyond the order of the court. Thus in Andhra when the High Court

struck down the list of Backward Classes for educational reservations

the state government used the occasion to abolish its reservations in

admissions to educational institutions and in government employment

as well.26

In many cases, litigation transforms a settled policy into an open issue. Thus in Mysore, Andhra, and Kerala, the most striking result of

litigation was the formation of commissions to formulate a new state policy in regard to reservations and a period of intense political concern with this issue. It should be noted that litigation had this effect in Kerala, in spite of judicial upholding of the existing reservations. The results of litigation may be felt in other states as well as in the place of its origin. Although government officials who deal with these matters are often uninformed or misinformed about what the courts have done,

there are some who are more aware and who scrutinize

judgments for guidance— usually with an eye to maintaining existing

policy and avoiding unconstitutional features. The effects seem more pronounced on the side of the prospective petitioners: in several instances a judgment concerning another state served to catalyze local

discontent into litigation against existing preferences. Thus, for example, Balaji stimulated petitions in Andhra and in Kerala. But the delay

24. Dushkin 1972: 192. 25. See the debates over this amendment, discussed in chap. 6, §B, and chap. 11, §A,

above. To some extent the Dora case contributed to the abolition of double-member constituencies (see chap. 3, §A, and chap. 13, §B, above), although this is an instance of avoiding the result, rather than over-turning it. 26. Discussed in chap. 7, §A, above. Note that the state later reinstated reservations along the lines of those abandoned.

528

The Judicial Process

in Madras suggests that such models may be insufficient to overcome features of the local scene that militate against litigation.?7 Indirectly, litigation may alter the constellation of groups which support or oppose a preference policy. The striking down of a scheme may disarm groups that enjoyed a reservation. Since these groups sometimes have little to unify them other than the reservation itself, the loss of the reservation may dissolve them into a series of disparate elements with conflicting views on what the new scheme ought to be. On the other hand, the expectation of success in litigation may bring together groups to oppose the existing scheme. A new policy of reservations mav reflect (or generate) a new convergence of interests. Consider,

for example, the situation in Mysore where, before Balaji, the thrust for preference for Other Backward Classes came from the populous and politically powerful middle castes, acting in conjunction with (or partially on behalf of) the great multitude of less articulate and less politically forceful communities, mostly quite poor. It was mainly the Brahmin community which was the target, and it was mainly Brahmins who opposed the policy through litigation. The post-Balayt shift to the economic test tacitly shifted the alignment of beneficiaries.28 Not only the politically dominant middle castes but Brahmins were now major beneficiaries of this policy; the former had managed in effect to co-opt their most determined and resourceful opponents and shift the burden

(here, of exclusion from medical colleges, etc.) to those least likely to

mount a campaign against it. And, of course, the new standard had the

added virtue of satisfying courts, publicists, and the nationwide educated audience that it was secular and non-communal. (Here, as in other instances, the flow of tangible benefits shifts from the have-nots to the haves, but under the banner of secularism and progress.)29 Even-

tually, political mobilization of lower groups and the resentment of some advanced groups against the extensive benefits enjoyed by Lingayats combined with a campaign of intellectual criticism to undermine this system. We have been speaking in terms of judicial initiative and government response, but Balaji should remind us that the relation of courts and government may be more complex. There the Supreme Court not only acted as critic of the state government's policy; it also acted as an agent disseminating the central government’s policy. The state governments had continued to ignore the Centre’s recommendations as to the makeup of the Backward Classes. Private litigants in effect used the Court to

27. See discussion in chap. 14,§C, above.

28. See chap. 8, §G, above, especially table 21.

29. Generally, on the relation between political symbolism and the distribution of

benefits, see Edelman 1967.

Uses of Judicial Action 5,29 force the state to adopt a policy along the lines of that approved by the central government. The Supreme Court served as a conduit for central

policy.

The flow of influence in this case is somewhat obscured. Later, central government spokesmen attributed the government'’s position to

the influence of the Supreme Court:

following the judgment of the Supreme Court [in'Balaji] the Central Government had accepted economic factors and not caste as the basis for de-

termining the backwardness of people.3°

Partisans of the new policy concurred in this view: thus the Times of India editorialized that “the Government has not been slow in acting on this

admirable judgment.”3! There are two significant transformations here. First, the central government had taken a firm and explicit position against the use of caste units in enumerating the backward classes at least as early as a year before the case was decided.32 Second, the Balaji judgment, as we have seen, did not rule out the use of caste and

communal units; instead it invalidated caste status as an exclusive test

of their backwardness while indicating a strong preference for purely economic tests. It was not the technical holding of the judgment that served to reinforce central policy, but the conflation of the holding with the Court’s policy preferences, flowing from the Court’s failure to disentangle the dual meanings of caste. The government was thus provi-

ded with rhetorical ammunition in its previously launched campaign against the use of caste units. It was relieved of having to debate on its

merits the question of using caste units, for it could now claim that it was constrained by the Supreme Court judgment (and the Constitu-

tion itself). This result was perhaps not unwelcomed or uninvited by

the Court in view of its warm commendation of economic tests applied directly to individuals.

The Court provided ammunition and reinforcement not only to gov-

ernment but also to other opponents of caste units: intellectuals and newspaper editorialists joined some government officials in embracing an undifferentiated and oversimplified version of Balaji as interdicting thé use of caste units in conferring preferences. In part, this is due to imperfect transmission, but it also involves an element of perceptual

distortion to fit the decision into the pre-existing preferences of these

groups.33

30. R. M. Hajarnavis, LSD (3rd Series), Vol. 48, No. 16, cols. 3973-76 (25 Nov.

1965).

31. Times of India, 30 Nov. 1965. 32. See chap. 6, §C above. 33. See chap. 7, above. Cf. the discussion of the perceptual distortion attending the

United States Supreme Court’s school prayer decision in Dolbeare and Hammond

530

The Judicial Process

Thus the Supreme Court not only reinforced central government policy and served as a conduit for disseminating and legitimating it: it also served as an instrument for bringing the states into conformity with it. Earlier the Centre had urged its powerlessness to change the state lists of Backward Classes. Now it claimed to have adopted “the path of persuasion” and in 1965 it claimed that eight states complied with the centrally-favored economic criteria. But inspection of the government’s

list of compliant

states reveals that the impact was less than it

claimed.*4 The list included two states (Gujarat and Maharashtra)

that had retained caste units outside the scholarship area; one state (Punjab) that not only retained the caste test outside the scholarship

area, but employed a list of communities in the scholarship area as well,

albeit in conjunction with an income test; one state (Orissa) that had no schemes of its own for Other Backward Classes, and two states (Assam

and West Bengal) that had none other than scholarships. Only two states with a high level of benefits to OBC adopted non-communal tests: Mysore and Andhra Pradesh. The latter reverted to caste units less than a year later—as Mysore did more than a decade later.

States with low levels of benefits for Other Backward Classes tended to go along with the Centre’s shift to economic tests. These tended to be the states where there had been no litigation. In the states with higher levels of benefits for OBC—and

these were the states where there had

been litigation—the effect was not to stimulate a changeover to an economic basis but to stimulate the pruning and rationalization of the community-wise

list. For

the most

part,

central

pronouncements

against communal units were given effect where the evils deplored were

least present and the quieter qualifications permitting communal units

with safeguards were given effect where extensive benefits to communal groups were at issue. (The exception of course was Mysore, where the economic test remained for more than a decade.)

The impact of court decisions, as of other governmental policies, is limited by decentralization. Even where, as in Balaji, the Supreme Court supports central policy, there is little the Centre can do administratively to bring around recalcitrant states—especially in areas where it has neither legislative power to exercise*5 nor financial inducements

(1971), where the decision was perceived so as to strain out unwelcome elements and make it compatible with prevailing practices. (There, as here, there was also mi: tion of prevailing practices.) In Balaji, the process works to magnify favored aspects of the decision. Actually, the newspaper version closely anticipated Chitralekha, but no notice was taken when that case effectively dropped out of sight. 34. See chap. 6 at n. 124, above. 35. “Education, including technical education, medical education and universi-

ties ....”” was a state subject until 1976 when it was moved from the state list to the

Uses of Judicial Action 531 to offer.36 (Of course, if the state ministry is politically dependent on the central government, there may be political inducements—but these may run in the other direction as well.) There are no central enforcement agencies for carrying out such policy. The function of implementing and enforcing central policies is left to the states. So while the system is very centralized in declarations of policy, including judicial policy, it is very decentralized in implementing these declarations. The

problem of implementation is compounded by the lack of reliable and

systematic data about what is going on. The Centre has no informationgathering apparatus of its own but obtains all of its information about the implementation of policy from the states, usually from those very officials in charge of the implementation. Provisions for the Backward Classes is an area in which the states have not been very forthcoming with information. Thus the effect of a court decision is not necessarily as widespread or as automatic as we might tend to assume from a formal chart of Indian government. The effect of a decision may be as much through cultural

and political channels as through structures of command and administration, and there may be a dissociation of symbolic changes from administrative ones.37 Even within the hierarchy of courts, the effect of a decision on the decisions in subsequent cases*® is not so automatic as our discussion of the doctrine of precedent in India might suggest. concurrent list by the 42nd Amendment. {t was subsequently restored to the state list by the Constitution (43rd Amendment) Act, 1977. 36. The Centre had applied its new views in the administration of the postmatriculation scholarship scheme. This is not an area in which the Centre felt strongly enough to provide other inducements or sanctions. For a criticism of this, see Ghurye 1969: 428H. 37. A consideration of the layers of symbolic transformation and distortion in this instance will surely induce modesty in attempting any estimate of effects. The Balaji court misperceived/misrepresented the policy of the Maharashtra government; the central government misperceived/misrepresented the holding of Balaji; the press and commentators misperceived/misrepresented the relation between the judgment and the central government's position; the central government mispercei

represented the state government response to central urging, etc., etc. We have here a

symbolic reality, a tissue of illusions built on illusions. This kind of dissociation of symbolic from administrative reality is not necessarily typical of policy- making in India, but I suspect it is not unique, especially in carrying out policies which are redistributive to low-status, inarticulate, and unorganized beneficiaries.

38. The interpretation of the constitutional provisions regarding compensatory dis-

crimination forms, for the most part, a rather isolated, self-contained area of Indian

constitutional law. Decisions here do not have wide ramifications in other areas. A notable exception is the Rangachani case, which in the course of bringing promotions

under Article 16 (4) firmly established that the equality guaranteed by Article 16(1) and

(2) applied to promotions, a point which gave rise to much litigation by government servants.

532

The Judicial Process

Outright evasion of prior judgments is not so formidable a problem in India as it has sometimes been in the United States. The system is more centralized, and the checks, formal and informal, against innovation and

initiative by lower courts are stronger. (And this is especially true in compensatory discrimination litigation, where the matter is entirely

within the higher judiciary.)39 Yet the Indian judiciary should not be viewed as a machine for the automatic and relentless application of previously established rules. The body of doctrinal learning propounded by the judiciary is constantly exposed to two opposing kinds of erosion. On the one hand, a subsequent court may’ screen out subtleties and minor themes and vulgarize a delicate calculus into a rigid and mechanical black-letter rule. On the other hand, receptivity to the subtleties and minor themes

ofa complex judgment may succeed in dissolving away the thrust of the judgment, a process often aided by the normative overload of the original decision. Balaji again provides an example of a judgment that has undergone both kinds of erosion: formalization (rigidification) and deformalization (diffusion). Thus Balayi’s holding about the permissible extent of reservations (which we might paraphrase as “reasonable under the circumstances, but never more than 50%”) was transformed

by subsequent courts into a flat 50% rule.*° On the other hand, Balayi’s teaching about caste was so overloaded— (1) disapproving caste units while (2) permitting them but (3) striking down caste standing as an

exclusive

test—that

subsequent

courts,

administrators,

and

com-

mentators could (and did) find support there for almost any position. The dangers of distortion are greater if we consider that judicial doctrines in this area have to be applied not only by courts, but by a host of administrators and clerks with less education and less legal sophistication than judges. The more subtle and pragmatic (in the sense of taking into account all the complexities of the given case) the rules that the courts lay down, the greater the distortion that will attend their application—as routine administration by harried functionaries generates pressures toward either formalism (rigid rules) or broad dis39. Transmission of rules within the system is attended by friction and slippage. Thus High Court judges may misunderstand the Supreme Court’s pronouncements; or they may believe that with their better information about local conditions, they can make the Supreme Court’s rule more suitable to those conditions. See, e.g., Viswanath v. Govern-

ment of Mysore, A. 1. R. 1964 Mys. 132, as an emendation on Balayi.

40. This may illustrate Friedman’s observation (1967: 815) that in recurrent crisis situations courts are likely “to develop a rule that can be delegated to other authorities for administration.” The courts have gravitated toward selecting those elements in the Balaji rule that can be quantitatively expressed, relieving themselves of imponderable judgments.

Uses of Judicial Action

533

cretion.*! Thus the courts are in something of a dilemma. The limits of their own energy and resources require that they make rules that can be administered by others. Yet the more sophisticated, subtle, and pragmatic the rules they propound, the less translatable into routine administration and the more likely to invite further recourse to the courts they become.? One solution would be to have more sophisticated and sensitive administrators and clerks, with greater resources for investiga-

tion and more time for deliberation. But this entails costs of its own.

And the diversion of resources to such a reform lies outside the scope of

judicial remedies.

B. JUDICIAL REVIEW: THE DOCTRINAL ACCOMPLISHMENT The Constitution confronts both government and courts with the

problem of reconciling the conflicting principles of equal treatment and compensatory discrimination. The sweeping language of Articles 15(4) and 16(4) suggests that the framers of the Constitution relied primarily on the discretion of the politicians and administrators of the future, rather than on the courts, to effect such a reconciliation. But while these provisions give the executive and the legislatures broad discretion in their application, judicial review is not entirely excluded. The initial

leverage for judicial intervention was the placement of these provisions as exceptions to the judicially enforceable Fundamental Rights. These

rights can only be vindicated to the extent that the courts undertake to see that the government has used its powers within the limit of the

constitutional authorization. Thus the courts may scrutinize the government’s designation of backward classes to see that the beneficiaries are indeed the backward classes. Similarly, the courts may examine the government’s schemes to see that they work in favor of the intended beneficiaries and not to their detriment and that the extent or method of operation of the schemes does not unduly impair the rights of others. Such review would seem essential not only to vindicate the rights of the non-backward but to effectuate the policy of these provisions.

41. Even though discretion implies increased power, those who exercise it may find it too costly and may prefer to divest themselves of it. Thus it was reported, that in Mysore, the practice of scoring the interview was dropped by the interview committee.

See chap. 13, n. 52, above.

42. Friedman 1967: 818.

43. This dilemma may characterize all judicial control over redistributive schemes

that have to be administered on a day-to-day basis by an army of clerks. On the limits of

judicial capacity to control the administration of complex welfare programs in the United States, see Handler 1966.

534

The Judicial Process

The main lines of policy-formation— the use, extent, method, reci-

pients, duration, and administration of preferences—lie with the government. But courts may still serve to keep their use consistent with constitutional boundaries. If the most important determinants of the success of compensatory discrimination lie beyond the power of the judiciary, there is still much that the judiciary can contribute to their working. In our survey of the law as developed in the courts, we encountered a number of substantial accomplishments: 1. The courts have decisively prevented the use of these provisions as justification for a regime of communal quotas.

2. The courts have inhibited runaway expansion of the category of

Other Backward Classes. (In restraining expansion they have pro-

tected not only the non-backward but also the intended beneficiaries of

these policies, preventing diversion and dilution of benefits.) 3. The courts have firmly established that gross over-inclusion will

not be tolerated; the Backward Classes must be confined to those who are demonstratively backward by some tangible criteria; that the appli-

cation of these criteria must be supported by adequate data, recent enough to be relied upon as reflective of current conditions. These requirements have not only institutionalized periodic reassessment and encouraged collection of information which may facilitate better administration, but they have clearly contributed to making discourse on the subject of Backward Classes more precise and cogent.4# 4. The courts have encouraged flexibility and scope in compensatory

discrimination policy by helping to free the government from the notion that the beneficiaries must be selected by some single pre-ordained criterion; by pointing out a variety of ways in which the backward might be selected; by dispelling the notion that the same data and standards could be used indefinitely. 5. By refusing to accept reservations as compartments or maximum quotas, and by refusing to confine special treatment to quantitative representation, the courts have insured that special provisions are not

used to curtail the opportunities of the beneficiaries.

6. The courts have succeeded in confining the quantum of preferences, thus assuring that other important interests will not be ignored. In developing these responses, the courts have proceeded sometimes by narrow technical construction of the constitutional provisions, sometimes by undertaking to construe these provisions with an eye to the 44. As is evident from a comparison of the more recent commission reports—

Government

of Gujarat

1976, Government

of Kamataka

1975, Andhra

Pradesh

Backward Classes Commission 1970; Kerala (Kumara Pillai Commission) 1966—with

¢.g., the repcrts of the Mysore Backward Classes Commission 1961 or the [Central]

Backward Classes Commission 1955.

Uses of Judicial Action

535

various broader social interests affected by preference policies. They have, in effect, read the words “reasonable” into Articles 15(4) and

16(4)—an approach that permits them to weigh the issues in terms of

the broad lines of policy set forth in the Constitution. The courts have served policy-makers and the nation well by keeping open options in pursuing its commitment to compensatory discrimination. They have rejected both the “modernist” view that communal units are impermissible and the “historical” view that only communal units may be used to designate Backward Classes, in favor of the “elastic’”’ view that enables the government to utilize a variety of ways of selecting Backward Classs.‘S Again, the courts have refused to reduce compensatory discrimination to compassionate adjustments appended to a regime of formal equality. And they have for the most part held ata

distance the view that the Constitution provides a blueprint for a regime of substantive equality and mandates the general use of compensatory classification for equalizing purposes.** By refusing to accept a single view of either the goals or the means of compensatory discrimination, the courts have been loyal to the original ambivalence and have kept open options for the future.

Having provided both critical standards and flexibility, the judicial

achievement is marked by a number of conceptual problems which could well impede the effectuation of the compensatory discrimination licy. wr irst, there is the abiding confusion surrounding the use of (1) castes as the units or classes that are deemed backward, and (2) the use of caste rank or standing as a measure of backwardness.47 Notwithstanding their general disapproval of caste, the courts have given little guidance to the states as to the limitations on the use of communal units in selecting Backward Classes. They have failed to discuss the kind of units that can be used or the kind of tests that can be used to select them. While they have put government on notice that caste standing is not byitself an acceptable test of backwardness, they have left unclear just how it may be measured and just what role it may play in the determination of backwardness. The failure of the courts to make clear the distinction (between castes as units and castes as ranks) has made it difficult to solve either of these problems. It has also led to considerable confusion and, in particular, it has obscured and diffused the original commitment to overcome the heritage of caste distinctions.

Second, while encouraging movement away from exclusive reliance

on communal

factors, the courts have portrayed

45. See chap. 7, §B, above.

46. Cf. chap. 11, §D, and chap. 12, §D, above.

Indian society as

47. See chap. 7, §A, above.

536

The Judicial Process

comprising closed and mutually exclusive compartments, arranged in

an unambiguous rank order: for example, by formal construction of the

schedules, by permitting religious tests for membership in Scheduled Castes, and by intruding notions of caste rank into the determination of tribal status.*8 The courts not only have allowed religious discrimination and community ranking to be obtruded into the law of preferences,

but have given currency to views of Indian society that are neither empirically persuasive nor congenial to the principles of the Constitution. Third, while promoting non-communal

tests of backwardness, the

courts have not indicated how these can be assessed by the high standards they have used to criticize communal tests.49 They have failed to reconcile economic tests with the doctrines set up for judging: the acceptability of communal units. Nor have they shown how economic tests may be directed to the purpose of overcoming accumulated inequalities. Fourth, although imposing some quantitative limits on preferences, the courts have not penetrated beyond the superficial question of the

specified amount of preference to address the real quantitative question of the net effects of preferences—and the way in which this is entwined with the question of allowable methods of administering reserva-

tions.5°

Fifth, in their concern with the abuses and excesses of some schemes

of preference, the courts have come close to propagating the notion that the government must designate some single group of “backward” classes who must be treated uniformly. This tendency has appeared in the judicial insistence that backward classes be both socially and educationally backward.5! It has appeared again in the judicial hostility to compartmental reservations and “layers” of preference.5? In spite of their recognition that backwardness comes in kinds and degrees, the

courts appear to be imposing on the government limitations which not only have little constitutional warrant, but could hamper the flexibility of governmental schemes for the backward. The State should not be confined to addressing its efforts to an undifferentiated group called “backward classes” but should be free to select backward classes by

criteria appropriate to the particular scheme. Such judicial restrictions would make difficult a transition to more differentiated criteria of backwardness. Again, the courts should not foreclose efforts to see that the smaller and most underprivileged groups are not overlooked while 48. See chap. 9, §§ A, D, E, above.

49. See chap. 8, §G. above. 51. See chap. 8, §C, above.

50. See chap. 12, §B, and chap. 13, §B, above. 52. See chap. 13, §C, above.

Uses of Judicial Action

537

the lion’s share of the benefits go to the most advanced and assertive sections of the backward. Sixth, while indicating some.of the interests that have to be balanced against the thrust of compensatory preference, the courts have not developed any standards for assessing the weightiness of such interests.53 They have, at times, slipped into a facile equation of the national interest with that of the advanced classes or of the non-backward applicants and students before them.54 Of course, such assessments are not susceptible of exact quantitative statement but rest upon judgments that proceed from value commitments as well as from estimates of fact. But the coherence and consistency of such judgments may be increased, not only by more incisive analysis, but by the presence of data that is relevant, comprehensive, and of high quality. Courts have required that in designating Backward Classes, government act on the basis of adequate data rather than mere assertion or casual opinion.$5 No comparable standards have yet developed for the kind of factual basis that must serve as the foundation for assertions about efficiency, morale, effective use of talent, etc.

Seventh, while developing justifications for compensatory treatment as consistent with the Constitution’s equality provision, the courts have left unclear the relation between compensatory discrimination for historically disadvantaged sections of society and the measures to remedy various kinds of personal and circumstantial deprivation.5® By diffusing the special authorization, they have obscured the commitment to overcome the heritage of historical differences in a hierarchic society. Nor have they convincingly assessed the difference between the policy of a favorable or equalizing tilt in favor of the Other Backward Classes and the commitment to intensive forced draft measures to promote the inclusion of the Scheduled Castes and Scheduled Tribes. Finally and perhaps foremost, there is a problem not of questionable doctrine but of the absence of doctrine. The courts have not equipped themselves with any doctrine by which they might reach the “‘affirmative” problems of compensatory discrimination policy—that is, by which they might assure that the deserving are included among the beneficiaries, that preferences are of sufficient scope and amount, that

they are implemented in a timely and effective fashion, etc. In part this flows from the basic constitutional posture— Fundamental Rights are justiciable, Directive Principles are not—which casts the courts as a 53. Thus, e.g., in C. A. Rajendran v. Union of India A. 1. R. 1968 S. C. 507, 514, the Court

simply accepts the government's satisfaction as sufficient evidence that efficiency would be adversely affected by extending promotion reservations to Classes I and II. 54. See, e.g., the discussion of Deoadasan in chap. 12, §B, above.

55. See chap. 8, §F, above.

56. See chap. 11, §D, and chap. 12, above.

538

The Judicial Process

brake and a baffle, rather than as stimulant and energizer of the compensatory discrimination policy. The decline of the older notion that Directive Principles were subordinate to Fundamental Rights has not really changed that posture. Although the Directive Principles have been elevated in constitutional dignity,57 and the powers of the legislature have thereby been amplified, the Directive Principles still lack self-executing legal force. They do not create any justiciable rights in favor of any party,58 nor in current

view are courts competent to compel the government to carry out any of

these directives.59 Thus the Directive Principles do not supply the courts with a lever for increased responsiveness to the beneficiaries of compensatory discrimination. But judicial unresponsiveness to beneficiary claims is as much the cause as the result of the lack of doctrine to justify an affirmative judicial role. Judicial unresponsiveness in turn reflects basic ideological presuppositions about the role of courts, but also the limited capacities of the

courts as institutions (e.g., to gather information or monitor implemen-

tation). But even if courts were not resistant, such an affirmative role is not something that can be produced by judges. It is the joint product of

courts and of competent and resourceful parties and lawyers who can identify grievances, devise workable remedies, erect legal theories to justify those remedies, and press their demands in a sustained and persuasive manner in courts and other fora. The limits of the court are compounded by the weaknesses of the other elements of legal activism. In this area, as we saw, group participation has been feeble and intermittent: there are few recurrent organizational actors with the resources or legal imagination to mount a persistent campaign to use the courts affirmatively. Nor is any significant continuity supplied by the lawyers. Although a few writ specialists have handled a number of such claims, cases involving compensatory discrimination do not form

the mainstay of the practice of a single one of India’s 228,000 lawyers. There is no occasion to develop any professional expertise regarding the administration of preferences. Nor has such expertise developed in the realm of legal scholarship. Judges, lawyers, academics, and potential 57. The Constitution (25th Amendment) Act, 1971, added Art. 31C, which provided that no law giving effect to the directives in Art. 39(b) and (c) should be deemed void as inconsistent with Arts. 14, 19, or 31. The Constitution (42nd Amendment) Act,

1976,§4, enlarged this to cover all of the Directive Principles. The pre-1976 law

regarding the relation of fundamental rights to Arts. 15, 16, and 20 is left intact.

58. Kesavananda v. State of Kerala, A. 1. R. 1973 S. C. 1467 (para. 134, 139, 1714); In re

Thomas, A. 1. R. 1952 Mad. 21.

59. Kesavananda, op cit., above; In re Kerala Education Bill, A. 1. R. 1958 S. C. 956.

Uses of Judicial Action

539

parties share a view of courts as critics and restraints on paternalistic government, but not as prods or energizers to make the government live up toits commitments. . The setting and task of the courts expose them to opposing dangers of, on the one hand, incoherence (ambivalence, inconsistency, overload) and, on the other hand, of formalism (conceptualism, false preci-

sion). Not surprisingly, both are found in our examination of the work of the courts. Courts have to respond to obvious abuses and to address intractable problems and imponderable fact situations. It is not surpirsing that they often erect a general rule in response to the formal characteristics of the situation rather than a rule calibrated to distinctions of fact and purpose. (Thus the “both socially and educationally” rule and the “no compartments” doctrine.) Or they substitute rule-manipulation for complex factual determinations (as in reading the lists or dealing with. converts). As a result, the courts promulgate rules that might well undermine the effectuation of the policies they are enunciating. In an area of law founded on the constitutional embrace of conflicting principles, it should not be expected that courts would provide an enduring synthesis that transcends and encompasses them and settles disputed issues with finality. Rather, we would expect—if the courts are at all representative of the larger society—some ambiguity and vacillation. And the courts have vacillated, sometimes emphasizing compensatory discrimination; other times, formal equality; sometimes,

rectification of communal disparities; other times, non-recognition of

communal units. We have seen that a number of factors may operate to relax the pressure for consistency and economy and to facilitate the

admission into the corpus of authority ofa variety of doctrines, not all of

them

reconcilable, which reflect the tensions between

antagonistic

principles. There are powerful forces operating to deflect the thrust toward consistency into a tolerance for ad hoc elaboration: heavy work pressures, high turnover on courts, a multiplicity of small benches of changing composition. Judges address the exigencies of different factual situations and have to apply doctrine which they have encountered infrequently and without the help of expertise on the part of counsel. Rajeev Dhavan attributes judicial tolerance for conflicting strands of precedent to “‘ineptness ur counsel and the lack of interest in the judges in trying to evolve a comprehensive and complete picture of the disputed points of law on a particular subject.” He goes on to cite the decisions about the caste basis for designating Backward Classes as 60. R. Dhavan 1977: 451.

540

The Judicial Process

exemplifying the observation that in India ‘‘new ideas are added without old ideas being discarded.”’61 Thus the courts generate a kind of doctrinal overload. In a setting of weak hierarchic structure, the presence of inconsistent doctrinal strands leads to the ambivalence and vacillation we noted earlier. As Dhavan remarks of the Supreme Court, “law does not evolve in the Supreme Court, it meanders. . . .”’62 Thus both formalism and incoherence are built into the structure of the situation. A conceptualistic intellectual tradition combines with imponderable questions and limited fact-finding capacities to produce overgeneral rules. Heavy work loads, sitting in small and shifting benches, lack of expertise on the part of advocates, and a lack ofa drive

for synthesis lead to the elaboration of various inconsistent strands of authority. The two tendencies are mutually reinforcing: the plurality of rules adds to overload and ambiguity, which is temporarily resolved by further rule elaboration. In the area of compensatory discrimination, both incoherence and formalism have yet another source: they flow from, and are accentuated by, the failure of the courts to come to grips with the immense variation and complexity of the group structure of Indian society. In examining the judicial handling of compensatory discrimination we can discern not only divergent proposals for combining and implementing the principles of the Indian Constitution, but also competing (ifinchoate) views of what Indian society is like. The judiciary are inevitably engaged in the delicate task of mediating between social actualities and the farreaching reconstruction of society envisaged by the Constitution. They are not only authoritative interpreters of these goals but assessors of the changing actuality in which these are to be realized. These are, of course, not wholly separate undertakings, for the perception of what is tends to be informed by one’s goals, and the interpretation of goals and values takes on content and color from one’s picture of what is. What one imagines to be desirable and attainable and what one imagines to be real are interdependent. The contribution of the judiciary to social reconstruction, then, is affected by the clarity, scope, and penetration of its view of Indian society. By virtue of their training, experience, and positions, judges

hold views of society that are to some extent partial and distorted.6

61. Dhavan 1977: 452. 62. Dhavan (1977: 452) submits that this is “India’s version of the ‘happy mean’. a

broad margin of tolerance, which is a variety within the common-law pattern and not to be judged by the standards of highly integrated and consolidated commonwealth countries.” 63. See chap. 9 generally. A recent example from a judge extremely sympathetic to the lowly, suggestsa kind of curious foreshortening in the judicial view of lower strata.

Uses of Judicial Action

541

This is of course not peculiar to Indian but is common to at least all large and diverse societies. And it is certainly not peculiar to judges. Each participant in society can see it only from his own vantage point and can apprehend it only in terms of available concepts and symbols. Along with other men of affairs, judges may have more elaborate and intricate images‘of social reality, but this makes them even more dependent on the quality of the concepts and data available to them. And along with other men of words, they come to cherish the symbols not only as instruments but as an independent source of grievance or satisfaction. A society is a symbolic system as well as a structure of social relations. For those charged with designing and effectuating policies to change fundamental aspects of social structure, deficiencies in the quality of these images can be a special disability. We have suggested earlier that the learning available to the courts may be an insufficient guide to the complexities of social structure and social process in India. Compared with other decision-makers, judges have the decided advantage of dealing with concrete factual situations, but they do so from the perspective of a relatively closed system of normative learning about that society, one whose fixed categories, formal classifications, and emphasis on rules may make it difficult for them

to develop the kind of understanding of society appropriate to guide this transformation. There is little in their milieu to assist them in overcoming the deficiencies of their training and the limits of their experience. It remains to be seen whether Indian scholarship can generate and pro-

vide the information (e.g., about the actual effects of judicial decisions)

and concepts that will enable the constitutional goals to serve as guides and monitors for (rather than as symbolic substitutes for) social

transformation.

C. THE LIMITS OF JUDICIAL ACTION We have seen several reasons that contribute to giving the courts a central role in the working out of the compensatory discrimination policy. There has been an absence of other agencies with specific responsibilities to oversee the implementation of these policies (like, e.g., equal opportunity commissions, attorneys general with enforcement staffs); there are no actors in the legal arena with detailed Thus Justice Krishna lyer thinks that the Kerala provision for clerks in the Thomas case may contribute to “an equalitarian break through” because “lower division clerks are

likely to be drawn from the lowest levels of harijan humanity” (A. I. R. 1976S. C. 490 at 531-32). 64. The Commissioner for Scheduled Castes and Scheduled Tribes is an information-

542

The Judicial Process

policies, expert staffs, and institutional momentum. Indeed, the government in implementing these policies avoids using the legal arena— the courts are not used affirmatively as an instrument for implementing government policy.65 They are brought in almost invariably by someone objecting to government action— government is always the defendant in compensatory discrimination cases. The courts encounter the policy in its incidence on aggrieved individuals. The higher judiciary enter directly through writ petitions, not by the indirect filtering route of appeals from lower courts or from an agency with expertise in implementing policy in this area. The higher judiciary, then, gets directly involved, not on its own initiative or that of the government as part ofa plan to implement these policies or as part of a process of monitoring their implementation, but in reaction to private grievances arising out of their administration. Thus the judicial encounter with compensatory discrimination is selective in several ways: First, those who

invoke the courts are those who not only have

something substantial at stake, but also have financial or organizational resources to defend or pursue it—e.g., excluded medical students, unpromoted

middle-level

government

employees,

defeated

can-

didates.© The student who fails to get his stipend on time or the disappointed aspirant for a reserved government post are not likely to litigate— much less the hutment dweller with no decent water supply. Second,

as we

have seen, preferential

treatment

tends to be con-

centrated in practice and tends to impinge heavily on relatively few individuals. It is from these areas of heavy concentration and these

targets of heaviest impingement that the cases come and the courts get their picture of compensatory discrimination policy. The doctrinal gathering agency with no enforcement powers. (On. the limitations of this office, see chap. 3, §C). Since 1969 it has been augmented by a Joint Parliamentary Committee on the Welfare of Scheduled Castes and Scheduled Tribes which has investigative powers but no enforcement or policy responsibilities. In July 1978 the five-member Scheduled Castes and Scheduled Tribes Commission, was established. See chap. 3, §C, at n. 126,

above. 65. This is not suprising, for the targets.of the compensatory discrimination policy are almost entirely within the government itself. Government does not use courts to compel

its agencies to carry out its policies. Presumably, they are reachable through the chain of

command (which may, however, be somewhat illusory). 66. That is, these are parties who lost the first round. Of course, those who are most amply provided with wealth and influence may find it possible to secure admissions (or

other prizes) through other channels and thus have no need to pursue these goals through

the courts. Cf. Dolbeare’s observation (1967) that plaintiffs in an American trial court

tended to be drawn from the prosperous, but those who had failed to prevail in more political forums.

Uses of Judicial Action

543

overload (i.e., the plurality of irreconcilable principles) in this area inclines courts to adopt a “balancing approach.” Naturally, in these cases they see their job as one of balancing the claims of the backward (as asserted by the government) against the claims of those immediately affected (e.g., government servants in posts subject to heavy reservations). Balancing or deciding what is “reasonable” becomes a weighing

of these limited interests, not all of those involved. The courts encounter only fragments of the total picture. They have no information about (and no power to act in regard to) those who escape entirely from the incidence of compensatory discrimination. Nowhere in the balance do we find the government officer in a post effectively immune from reservations or persons in private employment (where there are no reservations at all). Courts thus never get a chance to look at the whole but see only the fragment before them.” Third, the setting and the configuration of parties make it difficult for courts to give great weight to the long-term interests embodied in the compensatory discrimination principle. The claims that come to court focus on the distribution of scarce opportunities in the short run. The battle before them is about allocation this year, with its vivid and palpable effects. The courts naturally incline to visualize the preference policy as an attempt to be fair among competing claimants in the short run rather than in the light of long-term historical recompense. It is difficult for the courts to assess the long-term interests that are invoked on both sides. Although courts talk in terms of interests like efficiency, morale, utilization of talent, etc., as well as fairness to the parties, they

typically have little to go on in judging these wider effects. Little data on such effects have been generated by Indian social science; lawyers are not accustomed to locating such data or-arguing from them. In dealing with such interests the courts tend to fall back on an admixture of commonsense impressions and aprioristic surmise—which may be more or less acute and current.® The long-run interests of the nation in equalization are less tangible and immediate than interests in morale, etc., and even more difficult to assess. And the configuration of ad67. On

the distortion built into the forum’s reliance on individual complaints,

compare Mayhew’s observation (1968: 159) that the pattern of complaints to an anti-discrimination agency “did not correspond to the structure of discrimination. By and large, the complaints did not attack the major bastions of discriminations. Rather, the complaints reflected the current structure of Negro employment; they tended to be directed toward areas where racial barriers had already fallen.” On the deficiencies of reliance on individual complaints, see Lockard 1968: 98. 68. See. e.g., the remarks in Balaji (A.1.R. 1963 S.C. 659 at 662) about the inevitable

fall in standards. This apriorism is aggravated by the writ petition proceeding with its absence of opportunity to adduce and challenge evidence.

544

The Judicial Process

versaries may not bring before the courts those parties who could best assert these long-term and imponderable interests. The interests of the disappointed competitor do not necessarily correspond with those of the public; the interests of the government, eager to defend its existing schemes, do not necessarily correspond with those of the backward classes. While providing a forum responsive to what I have called the “‘too much” and “wrong thing” criticisms of compensatory discrimination policy, the courts have not been responsive to the “not enough” or “‘not the right thing”

criticisms. They

have done

little directly to offer

remedies for the deficiencies of implementation of existing schemes— to see that reservations are filled, scholarships are distributed on time, programs are well suited to circumstances and pursued with sufficient

energy, resources, and attention. In part, this is due to the posture of the Constitution, which provides no explicit authorization for affirmative

judicial action. Claims under Directive Principles are not justiciable—

so there is no justiciable claim that preferences be awarded. But courts,

abetted by lack of inventiveness at the bar, have extended this passivity to claims that those preferences which are awarded be implemented effectively. If courts could provide more in the way of judicial remedies for

abuses, maladministration, and omissions which hurt beneficiaries of

preference, they are not necessarily well suited to addressing the larger problems of pursuing the long-term goal of equalization. The limitations of the courts can be appreciated if we visualize compensatory discrimination as a special irrigation scheme, designed to bring lifegiving waters to lands which have long been parched and sterile as a result of neglect and oppression by the owners of the more prosperous lands among which they are situated. The system is instituted on the assumption that without it the available water would inevitably flow to the fields that are already fertile. The capabilities of the courts as keepers of the sluice gates are very different for the different groups involved. There is little that they can do for the parched lands— they can close down illicit diversions and stop the state from placing illegitimate limits on the flow in existing channels. But they cannot dig new

channels or deepen the existing ones. They cannot insure that the channels will be maintained in good repair and that all steps will be

taken to effectuate delivery. They cannot insure that the special water is

in addition to the trickle that might reach them through ordinary channels. And, they can do nothing to increase the flow of water into the system. But for the prosperous neighbors, they can stem the flow ifit is so excessive that it depletes the supply of water available for the rich land; they can see that water doesn’t flow to the wrong fields or through

Uses of Judicial Action

545

channels that damage adjacent interests. The sluice gates that the courts use to protect these neighboring interests may impede the legiti-

mate flow.

Ofcourse, the courts do not tend all the sluice gates themselves—only those major ones which have been the subject of complaint by the owners of the richer lands. The vast network of channels is tended by

keepers with less education and sophistication, fewer resources for

fact-finding, less time for deliberation. The more subtle and complex the rules that the courts establish at their gates, the more difficult it will be for the attendants along the channels to apply them. But the courts cannot deploy resources to upgrade and equip the attendants. They can act only in those cases which are brought to them. Thus the courts are a check on the system, not an expediter of it except in some marginal cases. They cannot put water into the system; they cannot supply any moving power; they cannot improve the channels or their attendants. These things can be done by politicians and administrators— including the occupants of reserved seats (and reserved posts). It would be wrong to visualize the courts as enemies or even as inadvertent wreckers of compensatory discrimination policy, somehow responsible for its deficiencies and shortcomings. These are not due mainly to courts. If courts have m a few instances played a restrictive role, this has been far outweighed by their positive contributions. Probably the greatest of these is to give compensatory discrimination legitimacy. By giving remedies in cases where others are hurt by it, the courts have made crucial publics feel, ifnot enthusiasm, that at least

it is not out of control and is somehow compatible with the welfare of all. By containing and curtailing the system, the courts have helped to maintain and preserve it, though at a level of performance below that of paper commitments. But the shortcomings are not in what the courts did, but in what other agencies of government (and outside government) did not do. The question is to what extent might the courts be instrumental in stimulating and energizing a higher level of performance by these agencies. In part, the inability of courts to institutionalize compensatory discrimination at a higher level of performance flows from the basic

architecture of the common-law court. Courts as we know them lack

independent fact-finding apparatus; they lack capacity for systematic and continuing monitoring and surveillance; they take up “cases” rather than address whole problems; they are limited in the nature of the remedies they can deploy (e.g., they have no scope to address problems by devising new administrative machinery). These structural 69. See Galanter 1974.

546

The Judicial Process -

‘

constraints are overlaid and accentuated by a number of features of the Indian legal mili¢u— narrow recruitment and high turnover of judges, an intellectual tradition of conceptualism and positivism, weak connection to social scientific or other sources of data. These institutional limitations of the courts are articulated to other features of the legal process. On the level of rules, there is the absence of doctrinal grounding for justiciability and for class actions. The bar has no specialists with enduring relationships to this kind of party, and these parties themselves are poorly organized, lack expertise, have no facilities for systematic gathering of information, etc. The absence of pressure from the courts for a higher level of performance of compensatory discrimination is the

result ofa complex of interlocking factors. But there is no reason to think

that the legal system is incapable of being mobilized to give greater support to these policies. One cannot say in advance what mix of organization of parties, supply of legal services, doctrinal inventiveness, and changes in institutional incumbents it would take to get new affirmative input from the legal system. But once such an affirmative role is undertaken, it may very well be self-sustaining: court success might stimulate organization of parties, the development of expertise, genera-

tion of data, the stimulation of further doctrinal innovation, etc. There

is no reason inherent in the legal process that its elements (doctrine, courts, lawyers, and parties) could not be joined in a pattern supporting increased effectiveness, just as they are at present joined in a low-

performance equilibrium.

CONCLUSION

16

“The Little Done, the Vast Undone” A. ACOSTLY SUCCESS Ir EXAMINING the response of the courts to compensatory discrimination policies teaches us something about Indian courts, what does it teach us about the policies themselves? Have they “worked”? What results have they produced? and at what cost? Our earlier tabulation of alleged costs and benefits suggested the complexity hidden in these apparently simple questions.’ Performance is difficult to measure: effects ramify in complex interaction with other factors. Compensatory policies are designed to pursue a multiplicity of incommensurable goals in unspecified mixtures that vary from program to program, from time to time, and from proponent to proponent. Evaluation of a specific scheme for a specific group during a specific period is itself a daunting undertaking. Does our inquiry tell us anything about the thrust and meaning of this protean cluster of measures and models? What has the commitment to compensatory discrimination done to the shape of Indian society and of lives lived within it? Employing some of the rubrics from our earlier list of alleged costs and benefits, I shall draw a very crude sketch of the policy’s effects in their largest outline. The limited clarity of such a sketch is dimmed

by the necessity to distinguish between

compensatory discrimination for the Scheduled Castes and Tribes on the one hand and for the Other Backward Classes on the other. The

following summary focusses on programs for Scheduled Castes (SC) and Scheduled Tribes (ST) and adds some qualifications in the light of experience with schemes for the Other Backward Classes (OBC). Undeniably, compensatory discrimination policies have produced substantial redistributive effects. Reserved seats provide a substantial 1, See the table of alleged costs and benefits in chap. 3, §D, above, and the

specification of that list to the job reservation area in chap. 4, above.

548

Conclusion

legislative presence and swell the flow of patronage, attention, and favorable policy to SC and ST. The reservation of jobs has given to a sizable portion of the beneficiary groups earnings, and the security, information, patronage, and prestige that goes with government emp-

loyment. At the cost of enormous wastage, there has been a major redistribution of educational opportunities to these groups. (Of course not all of this redistribution can be credited to preferential policies, for some fraction would presumably have occurred without them.)

Such redistribution is not spread evenly throughout the beneficiary groups. There.is evidence for substantial clustering in the utilization of these opportunities. The clustering appears to reflect structural factors. (e.g., the greater urbanization of some groups) more than deliberate group agerandizement, as is often charged.? The better situated among the beneficiaries enjoy a disproportionate share of program benefits.3 This tendency, inherent in all government programs— quite independently of compensatory discrimination— is aggravated here by passive administration and by the concentration on higher-echelon benefits. Where the list of beneficiaries spans groups. of very disparate condition—as with the most expansive lists of OBC — the “creaming” effect is probably even more pronounced. Reserved

seats afford a measure

of representation

in legislative

settings, though the use of joint electorates deliberately muffles the assertiveness and single-mindedness of that representation. The presence of SC and ST in legislative settings locks in place the other programs for their benefit and assures that their concerns are not dismissed or ignored. Job reservations promote their presence in other influential roles, and educational preferences provide the basis for such

participation. Of course, these positions are used to promote narrower

interests—although we should not assume automatically that those they displace would bestow the benefits of their influence more broadly. If, for example, reserved-seat legislators are disproportionately attentive to the concerns of their fellows who already have something, it is not clear that this is more the case with them than legislators in general seats. Preference programs are integrative in several ways. Reserved legislative seats are occupied by members of national political parties. They must aggregate broad multi-group support in order to get elected and, once elected, must participate in multi-group coalitions in order to be effective. In the office setting, too, there are relations of reciprocity and interdependence. The broad participation afforded by reserved seats and reserved jobs is for many others a source of pride and a warrant of security. 2. Shah and Patel 1977: 149 ff.

3. Malik 1979: 158.

“The Little Done, the Vast Undone”

549

If the separate and special treatment entailed by preferential pro-

grams wounds and alienates the members of beneficiary groups, this is

amplified by the hostility experienced on being identified as a recipient. As sources of alienation, these experiences must be placed against the background of more devastating manifestations of hostility, such as the much publicized assaults and atrocities perpetrated on Scheduled Castes. At the policy-making level, reserved seats have secured the acceptance of SC and ST as groups whose interests and views must be taken into account. In every legislative setting they are present in sufficient numbers so that issues affecting these groups remain on the agenda. Anything less than respectful attention to their problems, even if only lip service, is virtually unknown. Overt hostility to these groups is taboo in legislative and many other public forums. But there is evidence that SC and ST are not accepted politically. Very few members of these groups are nominated for non-reserved seats, and only a tiny number are elected. There is massive withdrawal by voters from participation in election for reserved seats in the legislative assemblies. Apparently, large numbers of people do not feel represented by these legislators and do not care to participate in choosing them.* In the long run, education and jobs help weaken the stigmatizing association of SC and ST with ignorance and incompetence, but in the short run they experience rejection in the offices, hostels, and other settings into which -they are introduced by preferential treatment.5 Resentment of preferences may magnify hostility to these groups, but rejection of them obviously exists independently of compensatory

programs.

Compensatory programs provide the basis for personal achievement and enlarge the beneficiaries’ capacity to shape their own lives. But in other ways the programs curtail their autonomy. The design of the legislative reservations—the dependence on outside parties for funds and organizations and the need to appeal to constituencies made up overwhelmingly of others—tends to produce compliant and accommodating leaders rather than forceful articulators of the interests of these groups. The promise of good positions offers a powerful incentive for individual effort. But reservations in: government service—and educational programs designed to provide the requisite qualifications— deflect the most able into paths of individual mobility that remove them from leadership roles in the community. Constraints are intruded into central issues of personal identity by eligibility requirements that 4. This evidence is presented in Galanter 1979.

5. Cf. Malik’s finding (1979: 50) that middle-class Schedisled Castes experience more

exclusion than do their less-educated fellows.

550

Conclusion

penalize those who would solve the problem of degraded identity by conversion to a non-Hindu religion. Although preferential treatment has kept the beneficiary groups and

their problems visible to the educated public, it has not stimulated widespread concern to provide for their inclusion, apart from what is mandated by government policy. (This lack of concern is manifest in the record of private-sector employment—as it was in publicundertaking employment before the introduction of reservations.) Against a long history of such lack of concern, it is difficult to attribute its current absence to compensatory discrimination policy. But this

policy has encouraged a tendency to absolve others of any responsibility for their betterment on the ground that it is a responsibility of the

government.® The pervasive overestimation of the amount and effec-

tiveness of preferential treatment reinforces the notion that enough (or too much) is already being done and nothing more is called for.

Compensatory preference involves a delicate combination of selfliquidating and self-perpetuating features. Reservations of upperechelon positions should become redundant as preferential treatment at earlier stages enables more beneficiaries to compete successfully, thus decreasing the net effect of the reservations. A similar reduction of net effect is produced by the extension to others of benefits previously enjoyed on a preferential basis (e.g., free schooling). Judicial requirements of more refined and relevant selection of beneficiaries (and of

periodic reassessment) and growing use of income cut-offs provide opportunities to restrict the number of beneficiaries. Reserved seats in legislatures are self-perpetuating in the literal sense that their holders can help to produce their extension. But extension requires support from others. The periodic necessity of renewal provides an occasion for assessment and curtailment. Programs for SC and

ST are for a delimited minoriy and pose no danger that the compensatory principle will expand into a comprehensive and self-perpetuating system of communal quotas. Although restrained by the courts, the provisions for OBC are open-ended: a majority may be beneficiaries, and the dangers of self-perpetuation cannot be dismissed. The diversion of resources by compensatory discrimination programs entails costs in the failure to develop and utilize other talents.

The exact extent of this is unclear. It seems mistaken, for example, to

consider compensatory discrimination a major factor in the lowering of standards that has accompanied the vast expansion of educational facilities since Independence. The pattern in education has been less 6. Cf. Lelah Dushkin’s observation (1979:666): “In the course of my visits to India over

two decades I have noticed an erosion and virtual disappearance of a liberal-minded public opinion supporting private efforts to improve opportunities for the S.C.”

“The Little Done, the Vast Undone”

one of excluding extending them the effectiveness shadowed by a

551

others than one of diluting educational services while nominally to all. Similarly, the effect of SC and ST on of a much enlarged government bureaucracy is overgeneral lowering of standards combined with the as-

sumption of a wide array of new and more complex tasks. The most disturbing costs of preferential programs may flow not from

their exclusion of others but from their impact on the beneficiaries. What do the programs do to the morale and initiative of those they purport to help? The numbers who fall by the educational wayside are legion.

How

rewarding is the educational experience of those who

survive? Compensatory discrimination policies are not the source of the deficiencies of Indian education that impinge with special force on the beneficiaries. As a forced draft program of inclusion of SC and ST within national life, compensatory discrimination has been a partial and costly success. Although few direct benefits have reached the vast mass of landless

laborers in the villages, compensatory discrimination has undeniably succeeded in accelerating the growth.of a middle class within these groups— urban, educated, largely in government service. Members of these groups have been brought into central roles in the society to an

extent unimaginable a few decades ago. There has been a significant redistribution of educational and employment opportunities to them; there is a sizable section of these groups who can utilize these opportunities and confer advaitages on their children; their concerns are firmly placed on the political agenda and cannot readily be dislodged. But if compensatory discrimination can be credited with producing this self-sustaining dynamic of inclusion, there is at the same time a lesser counter-dynamic of resentment, rejection, manipulation, and low selfesteem. And these gains are an island of hope in a vast sea of neglect and oppression. This mixed pattern of inclusion and _ rejection, characteristic of urban India and of the “organized” sector, is echoed in

the villages by a pattern of increasing assertion and increasing repression. Since Independence, India had undergone what might crudely be summarized as development at the upper end and stagnation at the bottom. With the boost given by compensatory discrimination, a section of the SC and ST has secured entry into the modern class manning the organized sector. What does this portend for the bulk of untouchables and tribals who remain excluded and oppressed? Are they better or worse off by virtue of the fact that some members of their descent

groups have a share in the benefits of modern India? The meaning of these achievements ultimately depends on how one visualizes the emergent Indian society and the role of descent groups in it.

552

Conclusion

Even this kind of crude characterization of the overall impact of policies.is not possible in dealing with measures for Other Backward Classes. Policies diverge from state to state, and very different groups of people are involved. In some states the OBC category is used to address

the problems ofa stratum of lowly groups who are roughly comparable in circumstance to the SC and ST. In other places this category has been used to tilt the distribution of government benefits in favor of a major section of the politically dominant middle castes. The latter doubtless produce substantial redistributive effects, if less in the way of including the most deprived. But these expansive preferences for OBC are of immense consequence for the SC and ST. They borrow legitimacy from the national commitment to ameliorate the condition of the lowest. At the same time they undermine that commitment by broad-

casting a picture of unrestrained preference for those who are not

distinctly worse off than non-beneficiaries, which attaches indiscrimi-

nately to all preferential treatment. And because the OBC categories are less bounded and are determined at the state rather than at the

Centre, they carry the threat of expanding into a general regime of

communal allotments.

B. FAIRNESS AND HISTORY Arguments

about the utility of compensatory

preference entwine

with arguments about its fairness. Apart from the other costs, is compensatory preference so tainted by unfairness that it is illegitimate to promote the general welfare by this means? We have encountered several arguments about the fairness and unfairness of these programs.

We shall examine four here: (1) Is it unfair to depart from judgments on “Sndividual merit” to favor the beneficiaries over other contenders for valued resources? (2) Is it unfair to compensate members of some groups for injustice perpetrated on their ancestors? (3) Is it unfair to compensate some victims and not others? (4) Is it unfair that some should bear more of the burden of compensation than others? Before taking up these fairness arguments it may be helpful to recall the justifications that may be advanced to the compensatory discrimination policy. Although they are often entwined in practice, we can separate out three sorts of justifications for these measures, which I label the non-discrimination, the general welfare, and reparations themes. The non-discrimination theme Compensatory discrimination may be viewed as an extension of the

norms of equal treatment, an extension invited by our awareness that even when

invidious discriminatory standards are abandoned

there

“The Little Done, the Vast Undone”

553

remain subtle and tenacious forms of discrimination and structural

factors which limit the application of new norms of equality. Aspiring

members of previously victimized groups encounter biased expectations, misperceptions of their performance, and cultural bias in selec-

tion devices; they suffer from the absence of informal networks to guide them to opportunities, entrenched systems of seniority crystallize and perpetuate the results of earlier discriminatory selections. Thus norms

of non-discrimination in present distributions are insufficient to erase or

dislodge the cumulative effects of past discrimination. Compensatory preference operates to counter the residues of discrimination and to overcome structural arrangements which perpetuate the effects of past selections in which invidious discrimination was a major determinant. In this view compensatory preferences serves to assure personal fairness to each individual applicant. Group membership is taken into account to identify those individuals who require special protection in order to vindicate their claim for selection on “merit” grounds. (The justification for much American affirmative action is often cast in these terms— as

an

extension

principles.)

of

classical

individualistic

non-discrimination

The general welfare theme One the other hand, compensatory discrimination may be advocated,

not as a device to assure fairness to individuals, but as a means to produce

- desired social outcomes—e.g., to reduce group disparities, afford representation, encourage the development of talent, and so forth. Arrange-

ments for reservations in British India were justified on such “functional” grounds as are the various preference for “Oriental” Jews in Israel today.7 Americans are familiar with the “balanced ticket” and other arrangements by which shares are apportioned among

various constituencies in the expectation

that abrasive dis-

parities are kept in bounds, participation is spread out, representation is secured, and responsiveness is assured. The units in such functional “welfare” calculations are groups rather than individuals. The chances of individuals are affected by the rearrangement of the chances of groups. But the purpose is not to rectify discriminatory selection among

individuals, but to introduce a standard quite apart from personal desert.

The contrast of non-discrimination and welfare themes is displayed

in imaginary alternative proposals for admitting more members of Group X to medical colleges. The non-discrimination proposal might argue that selection procedures be revised to eliminate subtle bias

7. On these programs, sce Smooha 1978; Alder et al. 1975.

554

Conclusion

which impinged on individual Xs—e.g., culturally biased tests or

differences in networks for acquiring recommendations. The general welfare proposal might argue that more Xs should be admitted in order to equalize the distribution of medical services or for the represenation of X views in making health policy or to afford non-X doctors the experience of fellowship with Xs. The goal is not a non-discriminatory selection of Xs among individual applicants, but a selection that optimizes social goals. Such a selection might diverge from that which

would be dictated merely by the prospects of individual performance on the job, because it defines the job to include the symbolic, representational, and educational aspects that may not be included in the job description. The reparations theme In some cases, compensatory policies have another root—that a history of invidious treatment has resulted in accumulated disabilities which are carried by certain groups. No matter how fair and unbiased the measures now employed for distributing benefits, the victims of past

injustice will not fare well in terms of current performance. To distribute

benefits by neutral standards will perpetuate and amplify unjust exactions and exclusions in the past. Fairness, then, demands that present

distributions

be arranged

perpetuate them.®

to undo

and

offset old biases, not to

Like the non-discrimination theme this is a fairness argument, rather than a welfare argument. But it emphasizes groups as the carriers of historic rights rather than as indicators of individual victimization. And

it looks to a very different time frame. Welfare arguments are prospec-

tive; non-discrimination looks at the present situation and seeks to refine out lingering inequalities. The reparations theme sees the present as an occasion to reckon accounts for past injustice. Do preferential programs unfairly confer benefits on grounds that

depart from evenhandedness, merit, etc., that should govern the dis-

tribution of opportunities and resources? Along the lines referred to 8. The notion of restitution for collective misdeeds is familiar in the indemnity and

reparation payments exacted from defeated nations (see Angell 1934). Something closer to the compensatory discrimination situation is found in the reparations payments paid

by Germany to Israel as the representative of Jews victimized by the Nazis (see K. Grossman 1954). In these instances the lapse of time between misdced and reparation is relatively short. But cf. the occasional claims for “three hundred years of back pay” raised in the American setting (Lecky and Wright 1969; Schuchter 1970; Bittker 1973). But a scheme of preferences may lack this theme of justification entirely, as in the Israeli case (n. 7, above), where the privileged and deprived groups had little previous contact.

“The Little Done, the Vast Undone” 555 above as the non-discrimination theme, proponents might respond that some of the preference accorded is not departure from evenhandedness but its extension in substance rather than form to individual members from the beneficiary groups. In this view, compensatory discrimination arrangements counter subtle discriminations and overcome the struc-

tural arrangements which entrench the results of past selections from which the beneficiaries were excluded. In the Indian setting, few would argue that compensatory discrimination seeks only to protect merit against subtle or structural bias.

Preferential treatment is accepted as a departure from merit selection in

order to promote such goals as redistribution, integration, and represen-

tation. Is it unfair to combine these with merit as a basis for distributing benefits? Let us take merit to mean performance on tests (examinations, in-

terviews, character references, or whatever) thought to be related to

performance relevant to the position (or other opportunity) in question and commonly used as a measure of qualification for that position. (In every case it is an empirical question whether the test performance is

actually a good predictor of performance in the position, much less of

subsequent positions for which it is a preparation.) Performance on these tests is presumably a composite of native ability, situational

advantages (stimulation in the family setting, good schools, sufficient wealth to avoid malnutrition or exhausting work, etc.), and individual effort. The latter may be regarded as evidence of moral desert, but

neither native ability nor situational advantages would seem to be. The common forms of selection by merit do not purport to measure the moral desert dimension of performance. Unless one is willing to assume that such virtue is directly proportionate to the total performance, the argument for merit selection cannot rest on the moral deservingness of individual candidates. Instead, it rests upon the supposed consequences: those with more merit will be more efficient or productive; awarding them society’s scarce resources will produce more indirect benefits for their fellows. A regime of rewarding merit will maximize incentive to cultivate talents; the demoralizing effects of departing

from merit outweigh supposed advantages, based on calculations of imponderables. The argument for merit is an argument for production of more social well-being. Many sorts of effects flow from any allocation of resources. Benefits are multiple; they include not only tangible production, but symbolic affirmations and the creation of competences. The allocation of educa-

tion, government jobs, or medical careers arguably has consequences for the distribution of incentives, levels of participation, disparities in

the delivery of services. Which dimensions of benefit are to be taken into

556

Conclusion

account in designing a given selection? In settings where there has been a broad consensus that a single-minded test of performance is appropriate, what is the argument for a shifting to a broader, promotional basis of selection? Compensatory discrimination schemes involve enlargement of the basis of selection to include other criteria along with

the productivity presumably measured by “merit” —representation, integration, stimulation, and so forth. This enlargement is justified on the ground that without it society would be deprived of the various benefits thought to flow from the enhanced participation of specified groups in key sectors of social life. The argument is that the combination maximizes the production of good results. Of course, there is always the empirical question of whether the promised results are indeed produced, but the supplementation of merit with other instru-

mental bases of selection hardly seems unfair in principle. Pursuit of

other worthy results can be balanced against merit, as one resultoriented justification for unequal allocations against another. Compensatory discrimination is both more and less than a reformulation of selection criteria. It is less because typically merit (in narrow

performance terms) is left intact for the main part of the selection. The

criteria are modified

to require the inclusion of certain groups, an

inclusion thought to produce a wide spectrum of beneficial results. But the new mixed standards are not applied across the board to the whole selection. So compensatory discrimination involves something more: the demarcation of those groups in whose behalf the broader promo-

tional standards should be employed.

To prefer one individual over another on grounds of caste, religion, or

other ascriptive criteria is specifically branded as unfair by the antidiscrimination provisions of the Indian Constitution. The ban on the use of these criteria is, as we have seen, qualified to allow preferential treatment of a certain range of groups, whose history and condition seemed distinctive. There was agreement that some groups were burdened by a heritage of invidious discrimination and exclusion (and/ or isolation), which made their condition distinct from that of their

fellow citizens; the deprivations of their past and present members were thought to justify a special effort for their improvement and inclusion. Spokesman for backward classes sonietimes call for measures specifically to remedy the wrongs of the past. Ifone thinks of the blighted lives,

the thwarted hopes, the dwarfing of the human spirit inflicted on generations of untouchables, or of the oppression and exploitation of

tribal peoples, the argument for measured vindication of these historic wrongs has an initial appeal. But there are many kinds and grades of victimization; deprivations are incommensurable. Perpetrators and

“The Little Done, the Vast Undone” 557 victims sometimes stand out in stark clarity, but infirm and incomplete

data often leave unclear precisely who were brutually exploitative, who willing or reluctant collaborators, who inadvertent beneficiaries of what ‘we now see as systems of oppression. These arrangements interact with many

other

factors—climate,

invasions,

technology—in

their in-

fluence on the present distribution of advantages and disadvantages. The web of responsibility is tangled, and as we try to trace it across generations, only the boldest outlines are visible. Without minimizing its horrors, we may concede that the past provides a shaky and indistinct guide for policy. It is beyond the capacity of present policy to remedy these wrongs: in the literal sense these injustices remain irremediable. But if our perception of past injustice does not provide a usable map

for distributing reparative entitlements, it can inform our vision of the present, sensitizing us to the traces and ramifications of historic wrongs. The current scene includes groups which are closely linked to past

victims and which seem to suffer today from the accumulated results of that victimization. In a world in which only some needs can be met, the

inevitable assignment of priorities may take some guidance from our sense of past injustice—thus providing the basis for a metaphoric restitution. All remedies involve new distinctions and thus bring in their train new (and it is hoped lesser) forms of unfairness. Singling out these historically deprived groups for remedial attention introduces a distinction among all of the undeserved inflictions and unfairnesses of the

world. One batch of troubles, but not others, are picked out for com-

prehensive remedy using extraordinary means. Those afflicted by other handicaps and misfortunes are left to the succor and aid that future policy-makers find feasible and appropriate within the framework of competing commitments (including commitments to equal treatment). But drastic and otherwise outlawed remedies were authorized for victims of what was seen as a fundamental flaw in the social structure. The special quality of the commitment to correct this flaw is dramatized by the Constitution’s simultaneous rejection of group criteria for any other pu The line of distinct history and condition that justifies compensatory discrimination is of course less sharp in practice than theory. There are border lines, grey areas, gradual transitions. There is disagreement about just where the line should be drawn. And once it is drawn, the

categories established are rough and imperfect summations of need and

desert;

there are inevitable

inclusion.

“errors”

of under-inclusion

and over-

558

Conclusion

We arrive, then, at an ironic tension that lies at the heart of the

compensatory discrimination policy. Since the conditions that invite compensatory treatment are matters of degree, special treatment generates plausible claims to extend coverage to more groups. The range of variation among beneficiaries invites gradation to make benefits proportionate to need. These preferential policies create new discontinuities, and it is inviting to smooth them out by a continuous modulated system of preferences articulated to the entire range of need and/or desert. But to do so is to establish a general system of group allotments. Compensatory discrimination replaces the arbitrariness of formal equality with the arbitrariness of a line between formal equality and compensatory

treatment. The

principles that justify the preference

policy counsel flexibility and modulation. We may shave away the

arbitrary features of the policy in many ways. But we may dissolve the

arbitrary line separating formal equality and preferential treatment only at the risk of abandoning the preference policy for something very different. If there is to be preferential treatment fora distinct set of historically victimized groups, who is to bear the cost? Whose resources and life chances should be diminished to increase those of the beneficiaries of this policy? In some cases, the costs are spread widely among the taxpayers, for example, or among consumers of a “diluted” public service. But in some cases major costs impinge on specific individuals, like the applicant who is bumped to fill a reservation. Differences in public acceptance may reflect this distinction. Indians have been broadly supportive of preferential programs—e.g., the granting of educational facilities and the sharing of political power—where the “cost” of inclusion is diffused broadly. Resentment has been focussed on settings where the life chances of specific others are diminished in a palpable way, as in reservations of jobs and medical college places. There is no reason to suppose that those who are bumped from valued opportunities are more responsible for past invidious deprivations than are those whose well-being is undisturbed—or that they were disproportionately benefitted by invidious discrimination in the past. Reserved seats or posts may thus be seen as the conscription of an arbitrarily selected group of citizens to discharge an obligation from which equally culpable debtors are excused. The incidence of reserva-

tions and the effectiveness with which they are implemented tends to

vary from one setting to another. Reservations impinge heavily on some careers and leave others virtually untouched. The administration of compensatory discrimination measures seems to involve considerable unfairness of this kind. If some concentration of benefits is required by

“The Little Done, the Vast Undone”

559

the aims of the preference policy,? it seems clear that more could be

done to distribute the burden among non-beneficiaries more widely and more evenly.

C. SECULARISM AND CONTINUITY Fairness apart, to many Indian intellectuals compensatory discrimi-

nation policies seems to undermine progress toward the crucial national goal ofa secular society. Secularism in this setting implies more than the separation of religion and State (religious freedom, the autonomy of religious groups, withdrawal of State sanction for religious norms, and so forth).

It refers to the elimination

(or minimization)

of caste and

religious groups as categories of public policy and as actors in public

life.1° In the 1950s and 1960s this was frequently expressed as pursuit

of a “‘casteless” society. Proponents of such a transformation were not always clear whether they meant the disestablishment of social hierarchy or the actual dissolution of caste units. But at the minimum, what was referred to was a severe reduction in the salience of caste in all spheres of life. The Constitution envisages a new order as to the place of caste in Indian life. There is a clear commitment to eliminate inequality of status and invidious treatment and to have a sociey in which govern-

ment takes minimal account of ascriptive ties. But beyond this the posture of the legal system toward caste is not as single-minded as the notion of a casteless society might imply. If the law discourages some assertions of caste precedence and caste solidarity, in other respects the prerogatives previously enjoyed by the caste group remain unimpaired. The law befriends castes by giving recognition and protection to the new social forms through which caste concerns can be expressed (caste associations, educational societies, political parties, religious sects).1

If the legal order’s posture toward caste is ambivalent, public denuciation of caste has universal appeal. For lower castes it provides an opportunity to attack claims of superiority by those above them; for the highest castes it is a way to deplore the increasing influence of previously subordinate groups, either the populous middle castes that have risen to power with adult suffrage or the lowest castes whose inclusion is 9. Cost spreading may itself entail costs in terms of the benefits delivered to the beneficiaries. For example, rotation of reserved legislative seats among constituencies might cure unfair clustering of costs, but it would vitiate the value of the reserved seats,

rendering impossible the accumulation of experience, seniority, and political strength. 10. On the aspiration to secularism, see the literature cited in chap. 9, n. 85, above.

11. See Galanter 1968.

560

Conclusion

mandated by compensatory discrimination programs. Looking up, the call for castelessness is an attack on the advantages retained by those who rank high in traditional terms; looking down, it denies legitimacy

to the redistributive claims of inferiors and insists on evenhanded applica-

tion of individual merit standards. The use of caste groups to identify the beneficiaries of compensatory discrimination has been blamed for perpetuating the caste system, accentuating caste consciousness, injecting caste into politics, and generally impeding the development of a secular society in which communal

affiliation is ignored in public life.12 This indictment should be re-

garded with some skepticism. Caste ties and caste-based political mobilization are not exclusive to the backward classes. The political life within these groups is not necessarily more intensely communal in

orientation;!> nor are the caste politics of greatest political impact

found among these groups. Communal considerations are not confined to settings that are subject to compensatory discrimination policies, but flourish even where they are eschewed. Although it has to some extent legitimated and encouraged caste politics, it is not clear that the use of caste to designate beneficiaries has played a preponderant role in the marriage of caste and politics. Surely it is greatly overshadowed by the franchise itself, with its invitation to mobilize support by appeal to existing loyalties. But the avowed and official recognition of caste in compensatory discrimination policy combines with the overestimation of its effects to provide a convenient target for those offended and dismayed by the continuing salience of caste in Indian life. The amount of preference afforded the Scheduled Castes and Tribes is widely overestimated. The widespread perception of ubiquitous and unrestrained preferment for these groups derives from several sources. First, there is the chronic overstatement of the effects of reservation:

large portions of reservations (especially for cherished higher positions)

are not filled; of those that are filled, some would have been gained on

merit; diversion of benefits to by a much larger number. The popularly perceived. Second, attention combine to blur the duled Castes and Tribes and

a few may be perceived as a deprivation net effect is often considerably less than is ambiguous nomenclature and public indistinction between measures for Schethose for Other Backward Classes. The

12. See chap. 3, §D, above.

13. Consider, for example, the fascinating finding of Eldersveld and Ahmed (1978:

205), studying political participation, that “the overwhelming majority of . . . activists are politically conscious of caste. They know how the caste leader voted. Among upper

castes the ratio is 6 to 1 that activists are aware of caste, in the middle castes it is better

than 8 to 1, and for the lower castes and Harijans it is 3 to | that activists are caste

conscious.”

.

“The Little Done, the Vast Undone”

561

resentment and dismay engendered by use of the OBC category to stake out massive claims on behalf of peasant-middle groups (particularly in some southern states) are readily transferred to discredit the more

modest measures for SC and ST.

If caste has displayed unforeseen durability, it has not remained unchanged. Relations between castes are increasingly independent and competitive, less interdependent and cooperative. “Horizontal” solidarity and organization within caste groups have grown at the expense of “vertical” integration among the castes of a region. The concerns of the local endogamous units are transformed as they are linked in wider networks and expressed through other forms of organization—caste associations, educational societies, unions, political parties, religious

societies.

If secularism is defined in terms of the elimination of India’s com-

partmental group structure in favor of a compact and unitary society,

then the compensatory discrimination policy may indeed have impeded secularism. But one may instead visualize not the disappearance

of communal groups but their transformation into components ofa pluralistic society in which invidious hierarchy is discarded while diversity is accommodated. In this view compensatory discrimination policy contributes to secularism by reducing group disparities and blunting

hierarchic distinctions.

The development of a secular society in which the hierarchic order-

ing of groups is not recognized and confirmed in the public realm is a departure from older Indian patterns. The compensatory discrimination policy is a major component in the disestablishment of a central part of the traditional way of ordering the society. But this break with the

past is itself conducted in a familiar cultural and institutional style. The

administration of preference programs reflects older patterns in the

fecund proliferation of overlapping schemes, the fragmentation of res-

ponsibility, and the broad decentralization of authority under the aegis of unifying symbols. When these policies encounter the judiciary, what purports to be a pyramidal hierarchy establishing fixed doctrine turns out to be a loose collegium presiding over an open textured body of

learning within which conflicting tendencies can be accommodated and elaborated.

The‘ compensatory principle of substantive equality is added to the constitutional scheme of formal equality, but it does not displace it. This juxtaposition of conflicting principles is an instance of what Glanville Austin admiringly describes as one of India’s original contributions to constitution-making, [that is] acoommoda-

tion . . . the ability to reconcile, to harmonize, and to make work without

562

Conclusion

changing their content, apparently incompatible concepts—at least concepts that appear conflicting to the non-Indian, and especially to the European or American observer. Indians can accommodate such apparently conflicting principles by seeing them at different levels of value, or, if you will, in compartments not watertight, but sufficiently separate so that a concept can

operate freely within its own sphere and not conflict with another operating in a separate sphere . ...With accommodation, concepts and viewpoints, although seemingly incompatible, stand intact. They are not whittled away by compromise but are worked simultaneously.’

The expectation that these principles could co-exist has been fulfilled.

The compensatory principle has been implemented, but it has not been

allowed to overshadow or swallow up opposing commitments to merit and to formal equality.

The compensatory discrimination policy is not to be judged only for

its instrumental qualities. It is also expressive: through it Indians tell themselves what kind of people they are and what kind of nation. These policies express a sense of connection and shared destiny. The groups that occupy the stage today are the repositories and transmitters of older patterns. Advantaged and disadvantaged are indissolubly bound to one another. There is a continuity between past and future that allows past injustices to be rectified. Independence and nationhood are an epochal event in Indian civilization which makes possible a controlled transformation of central social and cultural arrangements. Compensatory discrimination embodies the brave hopes of India reborn

that animated the freedom movement and was crystallized in the Constitution. If the reality has disappointed many fond hopes, the turn away from the older hierarchic model to a pluralistic participatory

society has proved vigorous and enduring.

D. EX ORIENT LUX? The Indian experience has obvious resonances with the contemporary American encounter with “affirmative action” and/or “reverse discrimination.” India and the United States are two of a number of democratic countries which employ policies of formal preference for members of disadvantaged groups for the purpose of erasing group disparities.15 I have tried to portray the Indian policies in the setting of Indian society and in its distinct political and legal culture. It would be 14, Austin 1966: 317-18. 15. Others are programs for Oriental Jews in Israel (see n. 7 above) and programs for the Burakumin in Japan (see Hah and Lapp 1978; DeVos and Wagatsuma 1966.) Cf.

“The Little Done, the Vast Undone”

563

inappropriate to assume that patterns found there would accompany cognate measures in very different settings. But Indian policies are more than expressions of unique and peculiar Indian conditions. The government that conducts them shares basic architectural features that

are found elsewhere (constitutionalism, the rule of law, a “neutral” civil service, an independent judiciary, etc.,); these policies are oriented to values that are widely shared. India and the United States are both large, complex, democratic

societies with heterogeneous populations in which sizable descent groups were in the past systematically suppressed, deliberately deprived of resources, skills, and opportunities, and branded as socially inferior. Each is embarked on an attempt to offset this heritage of invidious distinction by policies of preferential treatment for these historically disadvantaged groups. In both of these countries preferential policies are framed by a commitment to formally equal treatment of individuals. In India as in the United States courts occupy an extraordinarily prominent position as forums for individual grievances, monitors of governmental policy, and producers of political symbols. In both societies courts are called upon to play a key role in resolving the tensions between commitments to formal equality and to compensatory preference. Notwithstanding the colossal differences in history, culture,

and wealth that separate the two countries, there are enough parallels and affinities to suggest that India’s encounter with the perplexities of these policies may illuminate our own. At the least the Indian experience should make us wary of hasty and

tendentious inferences about the necessary characteristics of compensatory discrimination policies. It displays a range of options that are open. If it alerts us to some general tendencies in such policies, it suggests ways of tempering or counteracting these features. It cautions us about the difficulties of assessing such programs. Perhaps the most important lesson is that there is no single big lesson. That is, there is no large general consequence that flows inexorably from embrace ofa principle of compensatory preference. Compensatory discrimination does not necessarily extinguish commitments to merit

and evenhandedness, and it does not necessarily metastasize into a

comprehensive system of communal quotas. On the other hand, it does

not automatically produce the sought-after redistribution and it is not

costless.

The Indian experience demonstrates that the compensatory commit-

Malaysia’s extension of similar preferential devices into a comprehensive scheme of communal quotas favoring its politically dominant but economically backward Malay majority (Suffian 1972; MacDougall 1968; Ratnam 1965).

564

Conclusion

ment can co-exist with commitments to formal equality. Tensions at the level of principle can be contained; conflicting principles can be accommodated ina relatively stable pattern. The effects of the compensatory policy cannot be predicted from analysis of the principles that animate it. Nor can the performance of the policy be judged by the legal doc-

trine or public discourse that accompanies it. The symbolic dimensions of the policy are important in their own right, but they do not necessarily reflect crucial successes and failures in the design and implementation of pr .

The Indian experience indicates that recognition of group identity in

public life does not inevitably erode evenhandedness, generate invidious exclusions, or justify oppression. This instance ofa relatively benign jurisprudence of group identity should remind us that South Africa and Nazi Germany are not the most relevant examples of the use of group membership as a category for policy. But it should also be clear that the development of authoritative official learning on the question of group identity is an unavoidable concomitant of compensatory preference policies. However commonsensical the categories may appear at first blush, programs which turn on descent or ethnic categories inevitably encounter questions of who is included within these categories. Where descent, affiliation, avowal, and public regard coincide, such questions are easy to answer; where they diverge, we are drawn into a small forest

of less tractable puzzles. These puzzles reflect the overlap and inde-

terminacy of social identities; their malleability is stimulated by the

incentive offered by such programs to align identities with official categories. These tendencies appear in India, where these policies could address bounded corporate groups; surely they are no less present in the United States, where identities such as Black or Hispanic are less bounded and more diffuse. The development of such a jurisprudence reflects in miniature the irony of employing group identity in the service of reducing disparities among groups. To the extent that the policy aims to dissolve or diffuse

the boundaries between groups, application of rigid categories based on “objective” criteria should be eschewed in favor of determinations that recognize movement, overlap, and voluntarism. It requires more resources to administer flexible standards than mechanical ones; and the

administration of flexible standards broadcasts the problematic quality of group identity and points to the relative character of the selection of beneficiaries. Questions about which individuals are within the beneficiary groups flow into the problem of selection—why these groups and not others?—and thus into the justification of the programs. If justification is logically prior to group selection, in practice it is likely to be the other

“The Little Done, the Vast Undone”

565,

way around. Some group is widely acknowledged to require or deserve special treatment. The emergent response to that paradigm case— “untouchables” in India or Blacks in the United States—is summarized and codified in a set of justifications which present both barriers and precedents for other claimants. Once some group histories or conditions are accepted as justifying preferential treatment, there are inevitable questions of comparison. Are the deprivations of forced migration or peonage or religious persecution comparable to those of slavery? How should incidents of personal misfortune— family handicap, disease, accident, disaster, victimization—compare with those

inflicted on whole groups? Should preference be triggered by the need of the victims? or by the accountability of others for their deprivations? or by the potential for promotion of general well-being? Is there to be a single line separating those deserving of preferential treatment? Or should it be modulated by degrees? Once deserving beneficiaries are selected and levels of benefits set, the

system is attended by other ironies that can be displayed with a simple numerical example. Let us take as a paradigm case a selection for some coveted benefit in limited supply —e.g., medical college admission. Let us suppose that there is an uncontroverted way of measuring ability to perform as a medical student and that there is general agreement that it is good to choose medical students on the basis of this test. Suppose that there are twenty available places and that two of them are reserved for members of “backward class” X. Hundreds of students take the test,

including dozens of Xs. Suppose that the highest ranking X is number 6 and the next highest is number 46. In accordance with the reservation, number 20 is deprived of the twentieth seat and it is given to number 46 instead. (This example presents the basic preferential mechanism in the starkest form. The edges may be softened in various ways: there may be less than full agreement about the suitability of the test; the selection may involve a summing of incommensurables; and student performance may not be unambiguously ranked; who the Xs are may be open-ended; the reservation may be merely a guideline; and number

46 might be included without excluding number 20, presumably diluting ever so slightly the training given to each student.) This example points to some of the salient features of many compensatory discrimination schemes. First, there is the tendency of the program to cast a symbolic shadow much longer than the program itself, producing demoralization and resentment far disproportionate to the

benefits delivered. In our example, not only is number 20 displaced, but

numbers 21 to 45 are “‘passed over” and someone with a lower score is

chosen before them. Those below 46 may come away feeling that their

566

Conclusion

failure to gain the prize was due to unfair preference for the Xs. Even the successful non-Xs may resent the additional jeopardy to which their ambitions were exposed. Second, it is likely that all of the Xs are stigmatized as the recipients of preferential treatment, even though in our example 50% of the Xs won their place on “merit.” Third, in contrast to the non-Xs who see ubiquitous and unrestrained preference, to the dozens of Xs who didn’t

make it the amount of preference may seem paltry. Fourth, anyone who

seeks admission to a medical college may be deprived in the select company of his fellow applicants, but ranks high in education and advantages compared with the whole society and especially with the Xs. Ironically, benefits justified by the need and deprivation of the Xs tend to flow to those Xs who are least needful and deprived. Number 46 is elevated because his selection will benefit the Xs— by a percolation of tangible benefits, by representation, by their vicarious participation in his success, or in some other way. In India, where programs deal with

endogamous corporate groups, the diffusion of benefits throughout the

group is problematic; in the United States, where beneficiary classes are less bounded

and

cohesive,

the diffusion

of benefits

is even

more

problematic. Similarly, there is the converse of this problem. In our example, the direct cost of the preference was borne by number 20, who lost his opportunity to attend medical college. To the extent that the compensatory discrimination policy is based upon promotion of social

benefit or recompense for historic wrongs, the transfer from number 20

to number 46 is a payment for a collective good. These costs should be spread to the whole society (or some appropriate segment) rather than being borne by those individuals who happen to be located in the path of the policy. Like all remedial redistributions, compensatory discrimination imposes its own arbitrariness and unfairness. This is amplified to the extent that benefits are large and non-decomposable higher-echelon benefits (i.e., open only to those who have acquired demanding preliminary qualifications). They can be diminished by concentrating on benefits that can be spread more generally (like early educational

opportunities) and the cost of which can be monetized and spread widely. Such widespread costs and benefits impinge less dramatically on individual expectations and are far less provocative of resentment and resistance. While benefits and costs should be spread, they must also be bounded. Compensatory discrimination not only creates resentment of particular redistributions, but arouses fears that, by broadening the

“The Little Done, the Vast Undone”

567

categories and modulating benefits, it will gravitate into a comprehensive system of group allotments. To prevent such a slide it is important

that boundaries be clearly established— for what groups and for what

amount of time—with self-liquidating devices built into preference

programs.

Insight into the cumulative structural character of inequality has taught us that a regime of formal equality tends to perpetuate disparities we find intolerable. But if we find formal equality an insufficient remedy, we shrink from the uncharted plunge into comprehensive individualized substantive justice. Compensatory discrimination offers a

way to leaven our formalism without entirely abandoning its comforts.

The Indian example is instructive: India has managed to pursue a commitment to substantive justice without allowing that commitment

to dissolve competing commitments to-formal equality that make law viable in a diverse society with limited consensus. The Indian experience displays a principled eclecticism that avoids suppressing the altruistic fraternal impulse that animates compensatory policies, but that also avoids being enslaved by it. From afar it reflects to us a tempered legalism—one which we find more congenial in practice than in theory.

Appendix RELEVANT PROVISIONS OF THE INDIAN CONSTITUTION [AS OF 1975 EXCEPT AS NOTED] PREAMBLE WE, THE PEOPLE OF INDIA, having solemnly resolved to constitute India into a SOVEREIGN DEMOCRATIC REPUBLIC and to secure to all its citizens: JUSTICE, social, economic and political; LIBERTY of thought, belief, faith and worship;

them all

EQUALITY of status and of opportunity; and to promote among

FRATERNITY unity of the Nation;

assuring the dignity of the individual and the

IN OUR CONSTITUENT ASSEMBLY this twenty-sixth day of November, 1949, do HEREBY ADOPT, ENACT AND GIVE TO OURSELVES THIS CONSTITUTION. Part III Art. 14. Art.15.

Fundamental Rights The State shall not deny to any person equality before the law or the equal protection of the laws within the territory of India.

(1) The State shall not discriminate against any citizen on grounds only of religion, race, caste, sex, place or birth or any of them. (2) No citizen shall, on grounds only of religion, race, caste, sex, place of birth or any of them, be subject to any disability, liability,

restriction or condition with regard to— (a)

access

to shops,

public restaurants,

public entertainment; or

hotels and places of

(b) the use of wells, tanks, bathing ghats, roads and places of public resort maintained wholly or partly out of State funds or dedicated to the use of the general public. (3)

Nothing in this article shall prevent the State from making

(4)

Nothing in this article or in clause (2) of article 29 shall

any special provision for women and children.

prevent the State from making any special provision for the adv-

ancement of any socially and educationally backward classes of

citizens or for the Scheduled Castes and the Scheduled Tribes."

1. Added by the Constitution (First Amendment) Act, 1951, s.2.

570

Appendix

Art. 16.

q) There shall be equality of opportunity for all citizens in matters relating to employment

(2)

under the State.

or appointment

to any office

No citizen shall, on grounds only of religion, race, caste, sex,

descent, place of birth, residence or any of them, be ineligible

for, or discriminated against in respect of, any employment or

office under the State.

(3) Nothing in this article shall prevent Parliament from making any law prescribing, in regard to a class or classes of employ-

ment or appointment to an office under the Government of, or

any local or other authority within, a State or Union territory,

(4)

any requirement as to residence within that State or Union territory prior to such employment or appointment.

Nothing in this article shall prevent the State from making any

provision for the reservation of appointments or posts in

favour of any backward class of citizens which, in the opinion

of the State, is not adequately represented in the services under

(5)

the State. Nothing in this article shall affect the operation of any law

which provides that the incumbent of an office in connection with the affairs of any religious or denominational institution

or any member of the governing body thereof shall be a person

Art. 17.

Art. 29.

Art. 32.

professing a particular religion or belonging to a particular denomination. “Untouchability” is abolished and its practice in any form is

forbidden.

The

enforcement

of any

disability

arising

out of

“Untouchability” shall be an offence punishable in accordance with law.

(2)

No citizen shall be denied admission into any educational

institution maintained by the State or receiving aid out of

State funds on grounds only of religion, race, caste, language or any of them.

q) The right to move the Supreme Court by appropriate proceed-

ings for the enforcement of the rights conferred by this Part is guaranteed.

(2) The Supreme Court shall have power to issue directions or orders or writs, including writs in the nature of habeas corpus, mandamus, prohibition, quo warranto and certiorari, whichever

may be appropriate, for the enforcement of any of the rights

conferred by this Part.

(3) Without prejudice to the powers conferred on the Supreme

Court by clauses (1) and (2), Parliament may by law empower any other court to exercise within the local limits of its jurisdic-

tion all or any of the powers exercisable by the Supreme Court

(4)

under clause (2). The right guaranteed by this article shall not be suspended

except as otherwise provided for by this Constitution.

Appendix Part IV Art. 37.

Art. 46.

Directive Principles of State Policy The provisions contained in this Part shall not be enforceable by

any court, but the principles therein laid down are nevertheless fundamental in the governance of the country and it shall be the

duty of the State to apply these principles in making laws.

The State shall promote with special care the educational and

economic interests of the weaker sections of the people, and, in particular, of the Scheduled Castes and the Scheduled Tribes, and

shall

protect

exploitation.

Part VI

Art. 226.

571

them

from

social

injustice

and

all forms

of

The States (1) Notwithstanding anything in article 32, every High Court

shall have power, throughout the territories in relation to

which it exercises jurisdiction, to issue to any person or authority, including in appropriate cases any Government, within those territories directions, orders or writs, including

writs in the nature of habeas corpus, mandamus, prohibition, quo

warranto and certiorari, or any of them, for the enforcement of

any of the rights conferred by Part III and for any other purpose.

Part XV Election Art. 325. There shall be one general electoral roll for every territorial con-

stitutency for election to either House of Parliament or to the

House or either House of the Legislature of a State and no person shall be ineligible for inclusion in any such roll or claim to be included in any special electoral roll for any such constitutency on

grounds only of religion, race, caste, sex or any of them.

Part XVI Special Provisions Relating to Certain Clauses Art. 330. (1) Seats shall be reserved in the House of the People for— (a) the Scheduled Castes;

(b) the Scheduled Tribes except the Scheduled Tribes— (i) in the tribal areas of Assam;

(ii) in Nagaland; (iii) in Meghalaya; (iv) in Arunachal Pradesh; and (v) in Mizoram; and (c) the Scheduled Tribes in the autonomous districts of Assam.

(2) The number of seats reserved in any State or Union territory

for the Scheduled Castes or the Scheduled Tribes under clause (1) shall bear, as nearly as may be, the same proportion to the

total number of seats allotted to that State or Union territory in the House of the People as the population of the Scheduled

Castes in the State or Union

territory or of the Scheduled

572

Appendix Tribes in the State or Union territory or part of the State or Union territory, as the case may be, in respect of which seats

are so reserved, bears to the total population of the State or Union territory.

(3) Notwithstanding anything contained in clause (2), the number

of seats reserved in the House of the People for the Scheduled Tribes in the autonomous districts of Assam shall bear to the

total number of seats allotted to that State a proportion not less than the population of the Scheduled Tribes in the said auton-

Art. 332.

onmous districts bears to the total population of the State.

(1) Seats shall be reserved for the Scheduled Castes and the Scheduled Tribes, except the Scheduled Tribes in the tribal areas of Assam, in Nagaland and in Meghalaya, in the Legislative Assembly of every State. (2) Seats shall be reserved also for the autonomous districts in the Legislative Assembly of the State of Assam. (3) The number of seats reserved for the Scheduled Castes or the

Scheduled Tribes in the Legislative Assembly of any State under clause (1) shall bear, as nearly as may be, the same

proportion to the total number of seats in the Assembly as the population of the Scheduled Castes in the State or of the

Scheduled Tribes in the State or part of the State, as the case

may be, in respect of which seats are reserved, bears to the

Art. 334,

total population of the State. Notwithstanding anything in the foregoing provisions of the Part, the provisions of this Constitution relating to— (a)

(b)

the reservation of seats for the Scheduled Castes and the Scheduled Tribes in the House of the People and in the

Legislative Assemblies of the States; and the representation of the Anglo-Indian community in the House of the People and in the Legislative Assemblies of the States by nomination,

shall cease to have effect on the expiration ofa period of forty years?

from the commencement of this Constitution: Provided that nothing in this article shall affect any representation in the House of the People or in the Legislative Assembly ofa State until the dissolution of the then existing House or Assembly,

Art.335.

as the case may be. Theclaims of the members

of the Scheduled

Castes and

the

Scheduled Tribes shall be taken into consideration, consistently

with the maintenance of efficiency of administration, in the making of appointments to services and posts in connection with the affairs

of the Union or ofa State.

2. Originally ten years, but extenaed tor successive ten-year periods by the Constitution (Eighth Amendment) Act, 1959; the Constitution (Twenty-third Amendment) Act, 1969; and the Constitution (Forty-fifth Amendment) Act, 1980.

Appendix Art. 338.

Q) (2)

573

There shall be a Special Officer for the Scheduled Castes and Scheduled Tribes to be appointed by the President. It shall be the duty of the Special Officer to investigate all matters relating to the safeguards provided for the Scheduled Castes and Scheduled Tribes under this Constitution and report to the President upon the working of those safeguards at

such intervals as the President may direct, and the President

Art. 339.

(3)

shall cause all such reports to be laid before eaclf House of Parliament. In this article references to the Scheduled Castes and

(1)

article 340, by order specify and also to the Anglo-Indian community. The President may at any time and shall, at the expiration of

Scheduled Tribes shall be construed as including references to such other backward classes as the President may, on receipt of the report of a Commission appointed under clause (1) of

ten years from the commencement of this Constitution by order appoint a Commission to report on the administration of the Scheduled Areas and the welfare of the Scheduled Tribes

in the States. The order may define the composition, powers and proce-

dure of the Commission and may contain such incidental or ancillary provisions as the President may consider necessary

Art. 340.

or desirable.

(2)

The executive power of the Union shall extend to the giving of

Q)

The President may by order appoint a Commission consisting

directions to a State as to the drawing up and execution of schemes specified in the direction to be essential for the welfare of the Scheduled Tribes in the State.

of such persons as he thinks fit to investigate the conditions of socially and educationally backward classes within the territ-

ory of India and the difficulties under which they labour and to

make recommendations as to the steps that should be taken by the Union or any State to remove such difficulties and to

improve their condition and as to the grants that should be made for the purpose by the Union or any State and the conditions subject to which such grants should be made, and

the order appointing such Commission shall define the procedure to be followed by the Commission. (2) A Commission so appointed shall investigate the matters referred to them and present to the President a report setting out the facts as found by them and making such recommenda-

(3)

tions as they think proper.

The President shall cause a copy of the report so presented together with a memorandum explaining the action taken thereon to be laid before each House of Parliament.

574 Appendix Art. 341. (1) The

President may respect to any State or Union territory, and where it is a State after consultation with the Governor thereof,

by public notification, specify the castes, races or tribes or parts of or groups within castes, races or tribes which shall for the purposes of this Constitution be deemed to be Scheduled

Castes in relation to that State or Union territory, as the case

(2)

may be.

:

Parliament may by law include in or exclude from the list of

Scheduled Castes specified in a notification issued under clause (1) any caste, race or tribe or part of or group within

any caste, race or tribe, but save as aforesaid a notification issued under the said clause shall not be varied by any subse-

Art. 342.

(1)

quent notification.

The President may with respect to any State or Union terri-

tory, and where it is a State, after consultation with the Gover-

nor thereof, by public notification, specify the tribes or tribal communities or parts of or groups within tribes or tribal com-

munities which shall for the purposes of this Constitution be

deemed to be Scheduled Tribes in relation to that State or

Union territory, as the case may be.

Parliament may by law include in or exclude from the list of

Scheduled Tribes specified in a notification issued under clause (1) any tribe or tribal community or part of or group

within any tribe or tribal community, but save as aforesaid a notification issued under the said clause shall not be varied

by any subsequent notification. Part XIX Art. 366.

Miscellaneous [Definitions]

(24)

“Scheduled Castes” means such castes, races or tribes or

(25)

purposes of this Constitution. “Scheduled Tribes” means such tribes or tribal communities

parts of or groups within such castes, races or tribes as are deemed under article 341 to be Scheduled Castes for the or parts of or groups within such tribes or tribal communities

as are deemed under article 342 to be Scheduled Tribes for

the purposes of this Constitution.

Bibliography

Books, MoNnoGRAPHS, ARTICLES Abraham, Amrita 1978 “A Report from Marathwada.” Economic and Political Weekly 13 : 1536-40.

Adam, William 1840 The Law and Custom of Slavery in British India in a Series of Letters to Thomas Powell Buxton, Esq. Boston: Weeks, Jordan & Co. Adler, Chaim, Reuven Kahane, and Amy Avgar 1975 The Education of the Disadvantaged in Israel: Comparisons, Analysis and Proposed Research. Jerusalem: The Hebrew University of Jerusalem

School of Education.

Agarwal, C. B.

1934

— The Harijans in Rebellion. Bombay: Taraporevala Sons & Co.

Agarwal, R. N.

1964 — Inter-Caste Agarwal.

Tensions

Among

Children.

Agra:

Lakshmi

Narain

Aggarwal, PartapC. 1980 Halfway to Equality. Forthcoming. Aggarwal, Partap C., and Mohammed Siddiq Ashraf 1976

Equality through Privilege: A Study of Special Privileges of Scheduled

Castes in Haryana. New. Delhi: Shri Ram Centre for Industrial

Relations and Human Resources. A “Harijan” (pseud.)

c.1946 | Gandhi or Ambedkar. Madras: Gandhi Era Publications. 46 pp. Ahir, D.C. 1962 Babasaheb’s Message. Delhi: Institute of Buddhist Thought. 16 pp. Ahmed, Karuna

1978 — ‘‘Towards Equality: Consequences of Protective Discrimination.” Economic and Political Weekly 13:69-72.

Aiyappan, A. 1965 Social Revolution in a Kerala Village: A Study in Culture Change. New York: Asia Publishing House. Alexandrowicz, Charles Henry 1957 Constitutional Developments in India. Bombay: Oxford University Press.

576

Bibliography

All-India Backward Classes Federation 1954 Memorandum to the Backward Classes Commission. Amravati: 1955

Shivaji Printing Press.

Shri

Report of Working of the All-India Backward Classes Federation from the

Date of Its Formation up to 30th September, 1955. \0 pp. 1955 Report. Amravati: Shri Shivaji Printing Press. 1956-59 Reports. New Delhi: New India Press. 1961 Review. New Delhi: New India Press. All-India Bar Committee 1953 Report. Delhi: Manager of Publications. All-India Congress Committee n.d. Indian National Congress 1930-34: Being the Resolutions Passed by the Congress ... During the Period Between January 1930 to September 1934. ... Allahabad. All-India Federation of Scheduled Castes, Scheduled Tribes and Other Backward Classes 1964 Memorandum for the Implemenation of the Report of the Backward Classes

Commission, Government of India. New Delhi: S. M. Kaushik. 7 pp.

All-India Scheduled Castes Pratinidhi Sabha (held at Nanded) 1965 The Four Immediate Needs of Twelve Crore Supressed Human Beings in

India (Resolutions Passed Unanimously). Hyderabad: B. Sham Sunder.

Ambedkar, Bhimrao Ramji

1943.

1945 1946 1948

Mr. Gandhi and the Emancipation of the Untouchables. Bombay: Thacker & Co. Annihilation of Castes with a Reply to Mahatma Gandhi. 3rd ed. Bombay: Bharat Bhushan Printing Press. What Congress and Gandhi have Done to the Untouchables. 2d ed. Bombay: Thacker & Co. The Untouchables: Who Were They and Why They Became Untouchables?

1964

New Delhi: Amrit Book Co. States and Minorities: What Are Their Rights and How to Secure Them inthe Constitution of Free India. Bombay: Thacker & Co. Ranade, Gandhi and Jinnah (Address Delivered on the 10Ist Birthday

1968

Publications. The Annihilation of Castes with a Reply to Mahatma Gandhi [1945] and

1947

Celebration of M. Castes in India:

G. Ranade). Jullundur

Their Mechanism,

City:

Bheem

Patrika

Genesis and Development [1916].

[Reprint with an introduction by Bhagwan Das] Jullundur City:

Bheem Patrika Publications.

Ambedkar, Bhimrao Ramji, and Rao Bahadur R. Srinivasan

1932

‘Supplementary Memorandum on the Claims of the Depressed Classes for Social Representation.” Appendix 7 in Indian Round Table Conference (Second Session) 7th September, 1931—1st December 1931. Proceedings of Federal Structure Committee and Minorities Committee (Vol.

3). Calcutta: Government of India.

Anandan, P. I.

1911

“The Coronation and the Depressed Classes,” Hindustan Review 24:529-32.

Bibliography

577

Anant, Santokh Singh 1972 The Changing Concept of Caste in India, Delhi: Vikas Publications. 1978 “Changing Caste Hindu Attitudes Toward Harijans—A Follow-

up After Four Years.” In G. R. Gupta, ed., Cohesion and Conflict in Modem India. Durham, N. C.: Carolina Academic Press. Angell, James W.

1934

Anonymous

c. 1913

“Reparations.” In Encyclopaedia of the Social Sciences New York: MacMillan.

13:300-8.

The Depressed Classes: An Enquiry into Their Conditions and Suggestions

Sor Their Uplift. Madras: G. A. Natesan & Co.

1935-37 The Depressed Classes: A Chronological Documentation. Part 1, Ranchi: Rev. Fr. J. Jans, S. J., Catholic Press. Parts I—VII, Kurseong: St. Mary’s College. . Aryamane, R. V. 1964-66 “Backward Classes in the Indian Constitution,” Political Scientist

(Ranchi) 1-2: 27-36.

Association of Indian Universities 1977 Handbook of Medical Education. New Delhi: Association of Indian

Universities.

Austin, Granville 1966 The Indian Constitution: Cornerstone of a Nation. Oxford: C!arendon

Press.

Ayyar, R. S. Vaidyanatha.

1939

After the Reforms, Communal Versus Democratic Psychology. Karur: Maheswari Publishing House. 118pp.

Ayyar, S. V.

1965

A Study of Opinion Regarding the Various Ameliorative Measures to Improve the Conditions of Backward and Scheduled Castes. Annamalainagar: Annamalai University.

Bailey, F. G. 1960 Tribe, Caste and Nation: A Study of Political Activity and Political Change 1961

in Highland Orissa. Manchester: Manchester University Press.

““Tribe’ and ‘Caste’ in India.” In Contributions to Indian Sociology,

No. 5, edited by Louis Dumont and D. Pocock. Paris: Mouton &

Co. Baines, Athelstan

1912

“Ethnography of India.” Grundriss der Indo-Arischen Philologie und

Altertumskunde. Vol. 2 pt. 5. Strassburg: KarlJ. Trubner Verlag. Balasundaram, S. N. 1972 “Higher Education and Mass Politics.” In The Crisis in Higher Education; Proceedings, edited by G. D. Parikh. Bombay: Leslie Sawhney Programme of Training for Democracy. Friedrich-

Naumann Stiftung. Banaji, D. A. c. 1933 Slavery in British India. Bombay: D. B. Taraporewala Sons & Co. 2d ed.,n.d.

578

Bibliography

Banerjee, Anil Chandra

1948

The Making of the Indian Constitution 1939-1947. Vol. 1, Documents. Calcutta: A. Mukherjee & Co.

Banerjee, D. N. 1960

Our Fundamental Rights: Their Nature and Extent (as Judicially Determined). Calcutta: World Press.

Barnabas, A. P., and Subhash C. Mehta.

1965

Caste in Changing India. New Delhi: The Indian Institute of Public Administration.

Barnett, Marguerite Ross

1976

The Politics of Cultural Nationalism in South India. Princeton:

Princetion University Press.

Barrier, Norman Gerald

1968

‘The Punjab Government and Communal Politics, 1870-1908.”” Journal of Asian Studies 27:523-39.

Bastedo, T. G.

1968-69 “‘Law Colleges and Law Students in Bihar.” Law & Society Review 3: 269-94.

Basu, Durga Das

1977

Constitutional Law of India. New Delhi: Prentice-Hall of India.

Baxi, Upendra 1967 “The

Little

Done,

The

Vast

Undone’—Some

Reflections

on

Reading Granville Austin’s The Indian Constitution.” Journal of the

1975

Indian Law Institute 9:323-430.

‘Legal Assistance to the Poor: A Critique of the Expert Committee Report.” Economic and Political Weekly 10 (July 5): 1005-13.

Beals, Alan R.

1962

Gopalpur: A South Indian Winston.

Village. New

York:

Holt, Rinehart

&

Becker, Theodore L., and Malcolm M. Feeley, eds.

1973

The Impact of Supreme Court Decisions: Empirical Studies. 2d ed. New York: Oxford University Press.

Beidelman, Thomas O. 1959

A

Comparative

Augustin.

Analysis

of the Jajmani

System.

New

York:

J. J.

Berreman, Gerald D.

1963

Hindus of the Himalayas. Berkeley: University of California Press.

1965a

“The Future of the Backward Classes: The Competing Demands of Status and Power.” In Perspectives [Supplement to the Indian

Béteille, André

Journal of Public Administration, Vol. XI, no. 1], pp. 1-39. Reprinted 1965b

in India and Ceylon: Unity and Diversity. Edited by Philip Mason. New York: Oxford University Press, 1967. Caste, Class and Power: Changing Patterns of Stratification on a Tanjore

1974

Studies in Agrarian Social Reform. London: Oxford University Press.

Village. Berkeley: University of California Press.

Bibliography

579

Bhandare, R. D.

1965

The New Converts to Buddhism and Their Place in Indian Society: A Note

Presented to the Honorable Prime Minister of India. Bombay.

typescript. 55 pp.

Bound

Bhatt, Anil 1975 Caste, Class and Politics: An Empirical Profile of Social Stratification in Modem India. Delhi: Manohar Book Service. Bittker, Boris 1973 The Case for Black Reparations. New York: Random House. Black, Donald

1973 “The Mobilization of Law,” Journal of Legal Studies 2:125-49. Blackshield, A. R.

1968

“Fundamental Rights and the Economic Viability of the Indian

1980

“Rising Kulaks and Backward Classes in Bihar: Social Change in

Nation. Part III: Prospective Overruling.” Journal of the Indian Law Institute 10:183-240. Blair, Harry W.

the Late 1970s,” Economic and Political Weekly 15 (Jan. 12): 64-74. Blalock, Hubert M., Jr. 1967 Towards a Theory of Minority-Group Relations. New York: John Wiley & Sons. Blaug, Mark, Richard Layard, and Maureen Woodhall 1969 The Causes of Graduate Unemployment in India. London: Allen Lane. Blunt, E. A. H. 1931 The Caste System of Northern India: With Special Reference to the United Provinces of Agra and Oudh. London: Oxford University Press.

Blunt, E. A. H., ed. 1938 Social Service in India: An Introduction to Some Social and Economic Problems of the Indian People. London: His Maijesty’s Stationery Office. Bombay Presidency Social Reform Association 1939 Social Reform Annual, 1939. Bombay: Bombay Vaibhav Press. Bonarji, N. B.

1956 “Untouchability and the Law.” Times of India, July 8. Bopegamage, Albert, and P. V. Veeraraghavan 1967 Status Images in Changing India. Bombay: Manaktalas. Borale, P. T.

1968

Segregation and Desegregation in India: A Socio-Legal Study. Bombay:

Manaktalas. Bose, Nirmal Kumar 1954 “Who are the Backward Classes?” Man in India 34:89-98. Reprinted in Culture and Society in India. Bombay: Asia Publishing House,

1967.

Bose, Nirmal Kumar, ed.

1960

Data on Caste: Orissa. Memoir No. 7. Calcutta: Anthropological Survey of India.

5380 Bibliography Brandt, Richard B., ed.

1962

Social Justice. Englewood Cliffs, N. J.: Prentice-Hall.

1966

Factional Politics in an Indian State: The Congress Party in Uttar Pradesh. Berkeley: University of California Press.

Brass, Paul R. Brown, F. H.

1914

“The Social Dawn in India.” (Twentieth Century) 75:538—48.

The Nineteenth Century and After

Calabresi, Guido, and Phillip Bobbit. 1978 Tragic Choices. New York: W.W. Norton & Co.. Carter, Robert, Dorothy Kenyon, Peter Marcuse, and Loren Miller

1965 Equality. New York: Pantheon Central Government Scheduled Castes Welfare Association 1966 Memorandum on Reservation of Scheduled Castes and the Scheduled

Books. and Scheduled Tribes Employees Posts in Government Services for the Tribes. New Delhi, Feb. 1, 1966

Chakravarti, Satis Chandra, and Sarojendra Nath Ray, eds.

1933

Brahmo Samaj, the Depressed Classes and Untouchability. Calcutta: Sadharan Brahmo Samaj. 87 pp.

Chandidas, R.

1969

“‘How Close to Equality Are Scheduled Castes?” Economic and

Political Weekly4 (June 14):975-79. Chandidas, R., W. Morehouse, L. Clark, and R. Fontera 1968 India Votes (A Sourcebook on Indian Elections). New York: Humanities

Press.

Chatterjee, B. B., S. S. Singh, and D. R. Yadav

1971

Impact of Social Legislation on Social Change. Calcutta: Minerva Associates.

Chatterjee, N.C.

1965

Judicial Review and Fundamental Rights under the Constitution of India. Bangalore: Bharadwaja Publications.

Chaudhary, N. D.

1975

“Educational Development among the Scheduled Tribes in Rajasthan.” Education Quarterly 27 (3): 26-30. Chauhan, Brij Raj 1969 “Scheduled Castes and Scheduled Tribes.” Economic and Political Weekly 4 (Jan. 25):257-63. Chauhan, Brij Raj, G. Narayana, and T. N. Singh 1975 Scheduled Castes and Education. Meerut: Anu Publications. Chauhan, D.N. S., and Chopra, S. L.

1969

“A Comparative Study of the Socio-Economic Background of the Members of the Lok Sabha Elected from the Reserved and the General Constituencies.” Journal of Constitutional and Parliamentary Studies 3 (2):94—100. Chayes, Abram 1976 “The Role of the Judge in Public Law Litigation.” Harvard Law Review 89: 1281-1316.

Bibliography

581

Chidambaram Pillai, P.

1933.

Right of Temple Entry. Nagercoil: Chidambaram Pillai.

1901

Indian Social Reform: In Four Parts, Being a Collection of Essays, Addres-

1974

Literacy and Educational Enrolment Among the Scheduled Castes of Maharashtra. Bombay: Tata Institute of Social Sciences.

Chintamani, C. Yajnesvara

Ses, Speeches, etc., with an Appendix. Madras: Thompson and Co. Chitnis, Suma 1977.

A Long Way to Go

... (Report on a Survey of Scheduled Caste

High School and College Students in Fifteen States of India). Surat: Centre for Social Studies. Mimeographed.

Chowdhury, S. R. 1961 “‘Equality Before the Law in India.” Cambridge Law Journal 19: 223-38.

Cohen, Stephen P. 1969

“The Untouchable Soldier: Caste, Politics and the Indian Army.”

Journal of Asian Studies 28:453-68.

Cohn, Bernard S.

1965

‘Anthropological Notes on Disputes and Law in India.” American Anthropologist 67 (6), Part 2:82—-122. Coupland, R. 1944

The Constitutional Problem in India. (3 volumes in 1: Part I. The Indian

Problem, 1833-1935; Part II. Indian Politics, 1936-1942; Part III.

The Future of India.) London: Oxford University Press.

Dahl, Robert 1961

Who Governs? Democracy and Power in an American City. New Haven:

1967

“The Gandhian View of Caste, and Caste After Gandhi.” In India and Ceylon: Unity and Diversity, edited by Philip Mason. New York:

Yale University Press. Dalal, M.N. 1940 Whither Minorities? Bombay: D. B. Taraporevala Sons and Co. Dalton, Dennis

Oxford University Press. Damaska, Mirjan 1975 “Structures of Authority and Comparative Criminal Procedure.” Yale Law Journal 84:480-544. Das, Amal Kumar, and M. K. Raha

1962

West Bengal Tribes: Through Photographs (Special Series No. 6, Bulletin of the Cultural Research Institute). Calcutta: Temple Press.

Das, Amal Kumar, B. K. R. Choudhury, and M. K. Roha

1966

Handbook on the Scheduled Castes and Scheduled Tribes of West Bengal.

Special Series No. 8, Bulletin of the Cultural Research Institue,

Tribal Welfare Dept. Calcutta: Temple Press.

Das, Bhagwan, comp.

1963

Thus Spoke Ambedkar (Selected Speeches). Vol. 1. Jullundur: Bheem

1969

Thus Spoke Ambedkar: Selected Speeches. Vol. 11. Jullunder: Bheem Patrika Publications.

Patrika Publications.

582

Bibliography

Dave, P. K.

1966

“Promotion and Incentives in Public Services.” Indian Journal of Public Administration 12:533-43.

Davis, David Brion 1966 The Problem of Slavery in Western Culture. Ithaca: Cornell University

Press.

Davis, Kingsley 1951 The Population of India and Pakistan. Princetion: versity Press.

Princeton

Uni-

Derrett,J. Duncan M.

1958

“Statutory Amendments of the Personal Law of the Hindus since Indian Independence.” American Journal of Comparative Law 7:380—

93.

1963

Introduction to Modern Hindu Law. Bombay: Oxford University Press.

1968 Religion, Law and the State. London: Faber & Faber. De Reuck, Anthony, and Julie Knight, eds.

1967

Caste and Race: Comparative Approaches. London: J. & A. Churchill.

1969

“Reservation of Seats in the House of the People and State Assemblies for Scheduled Castes and Scheduled Tribes and Nominations of Anglo-Indians thereto: Case for Extension of the Period.” Journal of Constitutional and Parliamentary Studies 3 (2):8693.

Desai, C. C.

Deshpande, V. V. 1967

“Educational Planning: Some Aspects of Protective Discrimination.” In Educational Planning: Its Legal and Constitutional Implications in India, edited by G. S. Sharma. Bombay: N. M. Tripathi.

Deva, Shanthi, and C. M, Wagh

1965

Dr. Ambedkar and Conversion. Hyderabad: Dr. Ambedkar Publications Society. 46 pp.

De Vos, George, and Hiroshi Wagatsuma

1966

Japan’s Invisible Race: Caste in Culture and Personality. Berkeley: University of California Press. Dharmalingam, A. M. (Dharmu) 1951 Blasted Hopes: or Democracy in India; Bangalore: The Kural Nilayam. Dhavan, Rajeev

1977

— The Supreme Court of India: A Socio-Legal Critique of its Juristic Techniques. Bombay: N. M. Tripathi.

Dhaven, S. S., ed.

1964

Aspects of Indian Constitution. Allahabad: Central Book Depot.

1967

Trial Courts in Urban Politics. New York: John Wiley & Sons.

1971

The School Prayers Decisions: From Court Policy to Local Practice.

Dolbeare, Kenneth M.

.

Dolbeare, Kenneth M., and Phillip Hammond

Chicago: University of Chicago Press. Driver, Edwin D. 1963 Differential Fertility in Central India. Princeton: Princeton University

Press.

Bibliography

583

D'Souza, Victor S. 1969

“Education, Social Structure and Democracy in India.” In S. P.

Dua, Veena

1971

Ruhela, ed., Social Determinants of Educability in India. New Delhi: Jain Brothers. “Elections in a Scheduled Caste Neighbourhood: Town.” Economic and Political Weekly 6:2323-27.

A Punjab

Dubey, Harihar Prasad

1968

A Short History of the Judicial Systems of India and Some Foreign Countries. Bombay: N. M. Tripathi.

Dubey, S. N. 1973 Administration of Social Somaiya Publications.

Welfare

Programmes

in India.

Bombay:

Dumont, Louis

1970

Homo Hierarchicus: The Caste System and Its Implications. London: Weidenfeld and Nicholson.

Dushkin, Lelah 1957 “The Policy of the Indian National Congress Toward the Depres-

sed Classes: An Historical Study.” Master’s thesis, University of

1961 1972 1974 1979

Pennsylvania.

“The Backward Classes: Special Treatment Policy.” Economic Weekly 13: 1665-68, 1695-1705, 1729-38.

“Scheduled Caste Politics.” In The Untouchables in Contemporary India, edited byJ. Michael Mahar. Tucson: University of Arizona Press.

“The Nonbrahman Movementin Princely Mysore.” Ph. D. dissertation, University of Pennsylvania.

“Backward Class Benefits and Social Class in India, 1920-1970.” Economic and Political Weekly 14:661-67. Dutta, Ratna 1969 “The Party Representative in Fourth Lok Sabha.” Economic and Politial Weekly 4:179-89. Edelman, Murray 1967 The Symbolic Uses of Politics. Urbana: University of Illinois Press. Eldersveld, Samuel J., and Bashiruddin Ahmed

1978

Citizens and Politics: Mass Political Behavior in India. Chicago: University of Chicago Press. Eldersveld, Samuel J., V. Jagannadham, and A. P. Barnabas 1968 The Citizen and the Administrator in a Developing Democracy: An Empirical Study in Delhi State. New Administration.

Delhi:

Indian

Institute of Public

Fisher, Fred 1919 India’s Silent Revolution. New York: Macmillan.

Fisher, MargueriteJ. 1970 “Problems of Implementation in India’s Office of the Commis-

sioner for Scheduled Castes.” Western Political Quarterly 23:7 15-32.

.

584 Bibliography Fox, Richard D. 1969

‘Varna Schemes and Ideological Integration in Indian Society.”” Comparative Studies in Society and History 11:27-45.

Friedman, Lawrence M.

1966

‘On Legalistic Reasoning—A Footnote to Weber.”

1967

“Legal Rules and the Process of Social Change.” Stanford Laz Review 19:786-840.

Law Review 1966:148-71.

Wisconsin

Gadbois, George H., Jr.

1968-69 “Indian Supreme Court Judges: A Portrait.” Law & Society Review 1970a

3:317-36. “Indian Judicial Behavior.” Economic and Political Weekly 5:1-11.

1970b = “The Supreme Court of India: A Preliminary Report ofan Empirical Study.” Journal of Constitutional and Parliamentary Studies

1977.

(1):33-54.

4

“The Emergency, Mrs. Gandhi, the Judiciary and the Legal Culture.” Mimeographed.

Gaikwad, B. K., M. P. 1966 Speech delivered on the occasion of the 5th Session of the Republican Party of India held at Delhi on 1 May 1966. 14 pp. Gajendragadkar, P. B. 1951 The Hindu Code Bill. Two lectures delivered at the Karnatak University at Dharwar on 12-13 April 1951. Dharwar: Karnatak University. 52 pp. Galanter, Marc

1965 1968 1971

1972a

1972b 1974 1975 1976 1978 1979

“Secularism East and West.”

History 7: 133-59.

Comparative Studies in Society and

“Changing Legal Conceptions of Caste.” In Structure and Change in Indian Society, edited by Milton Singer and Bernard S. Cohn. Chicago: Aldine. “Hinduism, Secularism and the Indian Judiciary.” Philosophy East and West 21:467-87.

“The Abolition of Disabilities: Untouchability and the Law.” In The Untouchablesin Contemporary India, edited byJ. Michael Mahar. Tucson: University of Arizona Press. An earlier version appeared in 1969, Economic and Political Weekly 4:13 1-70. “The Aborted Restoration of ‘Indigenous’ Law in India.” Comparative Studies in Society and History 14:53-70. “Why the ‘Haves’ Come Out Ahead: Speculations on the Limits of Legal Change.” Law & Society Review 9:95-160. “Afterword: Explaining Litigation.” Law & Society Review 9:347-68 “Delivering Legality: Some Proposals for the Direction of Research.” Law & Society Review 11:226-46.

“Who are the Other Backward Classes? An Introduction to a Constitutional Puzzle.” Economic and Political Weekly \3: 1812-28. “Compensatory Discrimination in Political Representation: A Preliminary Assessment of India’s Thirty-Year Experience with

Bibliography 1983

585

Reserved Seats in Legislatures.” Economic and Political Weekly 14:437-54. ‘Emerging Patterns of Legal Services in India,” Journal of the

Indian Law Institute, forthcoming, 1983. Forthcoming. “Adjudication, Litigation

and

Related Pheno-

mena”. To appear in a handbook of law and social science, to be published under the auspices of the Social Science Research

Council. Galanter, Marc, ed. 1965 Legal Materials for the Study of Modern India. Chicago:

Division, University of Chicago Press.

Syllabus

Galanter, Marc, and Robert Kidder, eds.

1968-69 Lawyers in Developing Countries with Special Reference to India. Law &

Society Review 3, Special Issue. Galanter, Marc, Frank Palen, and John M. Thomas

1979

= ‘The Crusading Judge: Judicial Activism in Urban Trial Courts.” Southern California Law Review 52:699-74 1.

Gandhi, M. K.

1954a

1954b 1965

For Workers Against Untouchability. Anmedabad: Navajivan Publishing House. 34 pp. The Removal of Untouchability. Anmedabad: Navajivan Publishing House.

My Varnashrama Dharma. Edited by Anand T. Hingorani. Bombay: Bharatiya Vidya Bhavan. n.d. Untouchability. Edited by Dewan Ram Parkash. Lahore: Gandhi Publications League. Gauba, K. L. 1973 — Passive Voices: A Penetrating Study of Muslims in India. New Delhi: Sterling Publishers. Ghouse, Mohammad

1969 1973

“Judicial Control of Protective Discrimination.” Journal of the Indian Law Institute 11:371-81. Secularism, Society and Law in India. Delhi: Vikas Publishing House.

Ghurye, G. S. 1932 Caste and Race in India. London: Routledge and Kegan Paul. 1950 Caste and Class in India, 2d ed. Bombay: Popular Book Depot. 1957 Caste and Class in India. 3rd. ed. Bombay: Popular Book Depot. 1961

1963

Caste, Class and Occupation. 4th ed. Bombay: Popular Book Depot.

The Scheduled Tribes. 3d. ed. Bombay: Popular Prakashan.

1969 Caste, Class and Occupation. 5th ed. Bombay: Popular Prakashan. Glazer, Nathan

1975

Affirmative Discrimination: Ethnic Inequality and Public Policy. New York: Basic Books. Gledhill, Alan 1964 “The Expansion of the Judicial Process in Republican India.” In Some Aspects of Indian Law Today: International and Comparative Law Quarterly Supplementary Publication No. 8.

586

Bibliography

Gough, E. Kathleen

1960

“Caste in a Tanjore Village.” In Aspects of Caste in South India, Ceylon and North-west Pakistan, edited by E. R. Leach. Cambridge: Cambridge University Press.

Gould, Harold

1963

“‘The Adaptive Functions of Caste in Contemporary Indian Society.” Asian Survey 3: 427-38. Government of Malaysia 1964 — Citizenship of Malaysia. Kuala Lumpur: Government Press. 136 pp. Grimshaw, Allen D. 1959 “The Anglo-Indian Community: The Integration of a Marginal

Group.” Journal of Asian Studies 18:227-40.

Grossman, Joel, and Richard Wells

1972

Constitutional Law and Judicial Policy Making. New York: John Wiley

1954

Germany’s Moral Debt: The German-Israel Agreement. Washington:

& Sons. Grossman, Kurt Richard

Public Affairs Press.

Groves, Harry E. 1963

“Equal Protection of the Laws in Malaysia and India.”’American

1964

Aspectsof Indian Constitution. 2d ed. Allahabad: Central Book Depot.

Journal of Comparative Law 12:385-96. Gupta, Khadija A. 1971 “General Elections of 1967 in a Small U.P. Town.”’ Economic and Political Weekly 6:1881-86. Gupta, Madan Gopal Gupta, Raj Kumar 1969

‘‘Justice: Unequal but inseparate.”” Journal of the Indian Law Institute 11:57-86. Gurdon, P. R. T.

1914 The Khasis. 2d. ed. London: Macmillan and Co. Hah, Chong-Do and Christopher C. Lapp 1978 “Japanese Politics of Equality in Transition: The Case of the Burakumin.” Astan Survey 18:487-504.

Handler, Joel F. -1966 “Controlling Official Behavior in Welfare Administration.” In The Law of the Poor, edited by Jacobus tenBroek and the Editors of the California Law Review. San Francisco: Chandler Publishing Co. Hardgrave, Robert L., Jr.

1965

1969

The Dravidian Movement. Bombay: Popular Prakashan.

The Nadars of Tamilnad: The Political Culture of a Community in Change. Berkeley: University of California Press.

Harijan Sevak Sangh 1953-1970/71 Annual Report. Delhi: Udyogshala Press. Harrison, Selig S. 1960

India: The Most Dangerous Decades. Princeton: Princeton University Press.

Bibliography 587 Hart, Henry C. 1971 “Parliament and Nation-Building: England and India.” In Modern Parliament: Change or Decline? edited by Gerhard Loewenberg. Chicago: Aldine-Atherton.

Harvard Law Review 1954

Note, “Evasion of Supreme Court Mandates in Cases Remanded

to State Courts Since 1941.” Harvard Law Review 67:1251-59.

Hasan, Saiyid Zafar

1963.

Federal Grants and Public Assistance: A Comparative Study of Policies and Programmes in U.S.A. and India. Allahabad: Kitab Mahal.

Havanur, L. G.

1963. 1965

c1970

“Backward Classes.” Mysore Law Journal 1963(2): 15-19. Specifying the Backward Classes Without the “Caste” Basis. Bangalore: Mysore State Backward Classes Welfare Association. Judicial Blunders: Governmental Sins Against Backward Classes. Speech delivered by L. G. Havanur, Advocate, at the Preliminary Meeting of the SC, ST, and BC held Nov. 8, 1970, in Bangalore.

N.p. 8 pp. Hazari (pseud.) 1951

An Indian Outcaste: The Autobiography of an Untouchable. London: Bannisdale Press.

Heimsath, Charles H.

1964

Indian Nationalism and Hindu Social Reform. Princeton: Princeton University Press. Heller, Celia S. 1969 Structured Social Inequality: A Reader in Comparative Social Stratification.

New York: Macmillan.

Holderness, T. W.

1911 Peoples and Problems of India. London: Williams and Norgate. Horowitz, Donald L.

1977

The Courts and Social Policy. Washington, D. C.: The Brookings

Institution. Huang-Thio, S. M. 1963 “Equal Protection

1963:412-40.

and

Rational

Classification.”

Public

Law

Hutton,J. H. Caste in India: Its Nature, Function, and Origins. 3d ed. Bombay: 1961 Oxford University Press. Imam, Mohammed

1966

‘Reservation of Seats for Backward Classes in Public Services and Educational Institutions.” Journal of the Indian Law Institute 8:44 1-

49. Indian Institute of Public Administration 1959

= Morale in the Public Services: Report of a Conference, January 3-4, 1959. New Delhi: New India Press. Indian Institute of Public Opinion 1964 Monthly Public Opinion Survey. No. 109-10. 1966 Monthly Public Opinion Survey. No. 124.

588

Bibliography

Irschick, Eugene F. 1969 1977

Politics and Social Conflict in South India: The Non-Brakman Movement and Tamil Separatism, 1916-1929. Berkeley: University of California Press. Social Movements and the Indian Nationalist Movement: Tamilnad in the 1930’s. (Chapter I: “The Theory and Practice of Backwardness.””) Mimeographed Draft of Chapter I. ( forthcoming.)

Issacs, Harold R.

1965

India’s Ex-Untouchables. New York: John Day.

Jackson, A. M. T. 1907 “‘A Note on the History of the Caste System.” Journal of the Asiatic Society of Bengal (New Series) 3:509-15. Jagadisan, T. N., and Shyamlal, eds.

1949

§ Thakkar Bapa: Eightieth Birthday Commemoration Volume. Madras [no publisher given] Jain, M. P. 1961

“Reservation of Posts for Scheduled Castes and Article 16(4)—

General Manager, Southern Railway v. K. Rangachari.” Journal of

the Indian Law Institute 3:366-7 1. Jennings, Sir Ivor 1953

Jha, M. L. 1973

Some Characteristic of the Indian Constitution. London: Oxford University Press. Untouchability and Education. Meerut: Namita Publications.

Jones, Kenneth W.

1968

“Communalism in the Punjab: The Arya Samaj Contribution.”

Journal of Asian Studies 28:39-54.

Joshi, Barbara 1978 ‘Scheduled Caste Political Reservations: Who Benefits?” Paper presented to the Annual Meeting of the Association for Asian

Studies, Chicago.

Kannan, C. T.

1963

Intercaste and Inter-community Marriages in India. Bombay:

Publishers.

Allied

Karve, Irawati

1961

1968

Hindu Society: An Interpretation. Poona: Deccan College. Maharashtra:

Land and Its People. Maharashtra

State Gazetteers,

General Series. Bombay: Directorate of Government Printing. Karve, Irawati, and Y. B. Damle 1963 Group Relations in Village Community. Deccan College Monograph

Series, no. 24. Poona: Deccan College Post-Graduate and Research Institute.

Katzenstein, Mary F.

1979

Ethnicity and Equality: the Shiv Sena Party and Preferential Policies in

Bombay. Ithaca: Cornell University Press.

Keer, Dhananjay 1954

Dr. Ambedkar: Life and Mission. Bombay: A. V. Keer

Bibliography 589

1962 1964

Dr. Ambedkar: Life and Mission. 2d ed. Bombay: Popular Prakashan. Mahatma Jotirao Phooley: Father of Our Social Revolution. Bombay: Popular Prakashan. Kennedy, Duncan 1973 “Legal Formality.” Journal of Legal Studies 2:35 1-98, 1976 = “Form and Substance in Private Law Adjudication” Harvard Law Review 89: 1685-1778. Ketkar, Shridhar V.

1909 1911

= The History of Caste in India. Ithaca: Taylor and Carpenter.

An Essay on Hinduism: Its Formation and Future. London: Luzac and

Co. Khare, R. S. 1972

“Indigenous Culture and Lawyer’s Law in India.” Comparative

Studies in Society and History 14:71-96.

Khobrogade, B. D.

1966

Report of the General Secretary, Republican Party of India. New Delhi.7

Khosla, G. Rp

1949 Our Judicial System. Allahabad: University Book Agency. Kidder, Robert L. 1973 “Courts and Conflict in an Indian City: A Study in Legal Impact.” Journal of Commonwealth Political Studies 11:121-39. 1974 ‘Formal Litigation and Professional Insecurity: Legal Entrepreneurship in South India.” Law & Society Review 9: 11-37. Kirpal, Viney 1978 ‘Higher Education for the Scheduled Castes and Scheduled Tribes.” Economic and Political Weekly 13: 165-69. Kirpal, Viney, and Govind S. Kelkar

1976

‘Scheduled Castes and Tribes: Are They Brahmins’?” Illustrated Weekly of India 97 (16):4-9.

‘Government

Kochanek, Stanley A.

1968a

1968b

The Congress Party of India: The Dynamics of One-party Democracy. Princeton: Princeton University Press.

“The Relation between Social Background and Attitudes of In-

dian Legislators.” Journal of Commonwealth Political Studies 6:34-53. Koppell, Oliver 1966 Note, “Legal Aid in India.” Journal of the Indian Law Institute 8:224— 51. Krishna, Gopal

1966

“One Party Dominance—Develepment and Trend.” Perspectives. Supplement to the Jndian Journal of Public Administration 12. 1967 “Electoral Participation and Political Integration.” Economic and Political Weekly 2:179-90. Krishna Shetty, K. P. 1969 Fundamental Rights and Socio-Economic Justice in the Indian Constitution. Allahabad: Chaitanya Publishing House.

590

Bibliography

Kumar, Dharma 1965 Land and Caste in South India: Agricultural Labour in Madras Presidency in the Nineteenth Century. Cambridge: Cambridge University Press.

Kuppuswamy, B. 1978

Lajpat 1914 Lakoff, 1964

Backward Class Movement in Kamataka. Bangalore: Bangalore University. Rai, Lala The Arya Samaj. London; n.p. (1967 rev. ed. by Sri Ram Sharma, A History of the Arya Samaj. Bombay: Orient Longman). Sanford A. Equality in Political Philosophy. Cambridge: Harvard University Press.

Lal, Sheo Kumar, and Umed Raj Nahar

1978

Higher Education: Scheduled Castes and Scheduled Tribes. Jodhpur:

Jainsons.

Lander, Stephen

1965

“Backward Classes’ in the Indian Constitution.” Mimeographed. Chicago: University of Chicago.

Latthe, A. B. 1924 Memoirs of His Highness Shri Shaku Chhatrapati Maharaja of Kolhapur. 2 vols. Bombay: Times Press.

Leach, E. R., ed. 1960

Aspects of Caste in South India, Ceylon and North-west’ Pakistan. Cambridge Papers in Social Anthropology, no. 2. Cambridge for the Dept. of Archaeology and Anthropology: University Press.

Lecky, Robert S., and H. Elliott Wright, eds.

1968

Black Manifesto—Religion, Racism and Reparations. New York: Sheed and Ward.

Lewis, Oscar

1958

Village Life in Northern India: Studies in a Dethi Village. New York: Vintage Books. Lockard, Duane

1968

Toward Equal Opportunity: A Study of State and Local Antidiscrimination

Laws. New York: Macmillan. Lohia, Ram Manohar 1964 The Caste System. Hyderabad: Navahind: Luthera, Ved Prakash

1964

The Concept of the Secular State and India. London: Oxford University Press. Lynch, Owen M.

1969

The Politics of Untouchability: Social Mobility and Social Change ina City

of India. New York: Columbia University Press. McCormack, William 1966 Caste and the British Administration of Hindu law. Journal of Asian

andi African Studies |:25-32.

MacDougall, John Arthur

1968

— Shared Burdens: A Study of Communal Discrimination by the Political

Bibliography

591

Parties of Malaysia and Singupore. Cambridge, Mass.:Harvard University Press.

Maddison, Angus

1971

Class Structure and Economic Growth: India and Pakistan Since the Moghuls. New York: W. W. Norton.

Madras, Henry

1909 ‘The Out-Castes of India.” The Nineteenth Century 66: 1022-31. Majumdar, Nabendu Dutta 1955 “The Backward Classes Commission and Its Work.” In Social

Welfare in India (issued on behalf of the Planning Commission). New Delhi: Government of India. Malik, Hafeez

1963

Moslem Nationalism in India and Pakistan. Washington. D. C.: Public

1979

Social Integration Publications.

Affairs Press. Malik, Suneila

of Scheduled

Castes.

New

Delhi:

Adhinav

Mandelbaum, David G.

1970 — Society in India. 2 Vols. Berkeley: University of California Press. 1974 Human Fertility in India: Social Components and Policy Perspectives. Berkeley: University of California Press.

Marriott, McKim

1959 1960 1976

= “Interactional and Attributional Theories of Caste Ranking.” Man in India 39:92-104. Caste Ranking and Community Structure in Five Regions of India and Pakistan. Deccan College Monograph Series, no. 23. Poona: Deccan College Postgraduate and Research Institute. ‘Hindu Transactions: Diversity Without Dualism.” In Transaction and Meaning: Directions in the Anthropology of Exchange and Symbolic Behavior, edited by B. Kapferer. Philadelphia: Institute for the Study of Human Issues.

Marriott, McKim, ed.

1955

Village India: Studies in the Little Community. Chicago: University of

Chicago Press. Marriott, McKim, and Ronald Inden 1974 “Caste Systems.” Encyclopaedia Britannica. Macropaedia 3:982-91.

15th ed. Chicago:William Benton.

Mathur, P. C. 1966 “Politics and Anti-Casteism in India.” Political Science Review 5 (1):

82-93.

Mayakrishnan, V., and N. Radhakrishnan

1965

“A Critical Note on State of Kerala v. R. Jacob” Madras Law Journal 1965 (1): 23-28. Mayer, Adrian C. 1960 Caste and Kinship in Central India: A Village and Its Region. Berkeley: University of California Press.

592

Bibliography

Mayhew, Leon H.

1968

1975

Law and Equal Opportunity: A Study of the Massachusetts Commission Against Discrimination. Cambridge: Harvard University Press. “Institutions of Representation: Civil Justice and the Public.” Law & Society Review 9:40 1-29.

Mehta, Ashoka 1965a The Politics of Poverty. Feroze Memorial Lecture. New Delhi: Plan1965b

ning Commission.

Freedom and Justice in a Changing Society. New Delhi: Government of India. 9 pp. Mehta, Subhash Chandra 1963 “Persistence of the Caste System: Vested Interest in Backward-

ness.” Quest no. 36, pp. 20-27. Merillat, H.C. L. 1967 “The Soundproof Room: A Matter of Interpretation.” Journal of the Indian Law Institute 9:521-46. Miller, RobertJ. 1966

“Button, Button

... Great Tradition,

Little Tradition, Whose

Tradition.” Anthropology Quarterly 36:26-42. Mittal,J. K. 1964 “Special Representation in Legislative Bodies in India.” Indian 1965a 1965b 1965c

Advocate 4:23-48.

:

‘Educational Equality and the Supreme Court Advocate 5( 1-2) : 31-39. | “‘Equality of Opportunity in Matters of Public the Indian Supreme Court.” Malaya Law Review “‘Right to Equality and the Indian Supreme Journal of Comparative Law 14:422-58.

of India.” Indian . Employment and 7:113-26. Court.” American

Mohapatra, Manindra Kumar 1973. “The Scheduled Caste and the Tribal Political Elites: An Empiri-

1974

cal Exploration of the Attributes, Ascent and Role Perceptions of ‘Token Elites’ in an Indian State.” International Behavioural Scientist 5 (4):1-23. “College Youth and Legislators’ Role in a Modernizing Legislature: Preliminary Report of an Empirical Study.” Paper prepared for the American

1978

Chicago.

Political Science

Association’s

70th meeting,

“‘The Scheduled Caste Members of the Indian Lok Sabha in 1977:

An Empirical Study of Developmental Change in Legislative Recruitment.” Paper presented to the Seventh Annual Conference on South Asia, University of Wisconsin, Madison, November 3-4,

1978.

Mootham, Sir Orby

1963

“Constitutional Writs in India.” In Changing Law in Developing Countries, edited by J. N. D. Anderson. London: Allen and Unwin. 1966-67 “The Indian High Courts Before and After Independence.” Jn South Asia Seminar, Contributions to the Study of Indian Law and

Bibliography

593

Society. Philadelphia: South Asia Regional Studies, University of Pennsylvania.

Morkhandikar, R. S.

1978

‘‘Marathwada Riots: The Background.” Economic and Political Weekly 13:146 1-2. Morris-Jones, W. H. 1957 Parliament in India. Philadelphia: University of Pennsylvania Press. Morrison, Charles.

1972 1974

.

“Kinship in Professional Relations: A Study of North Indian District Lawyers.” Comparative Studies in Society and History 14:10025. “Clerks and Clients: Paraprofessional Roles and Cultural Identities in Indian Litigation.” Law & Society Review 9:39-61.

Mukherjee, B. K. 1967 ‘Parliamentary and Constitutional Developments: Government's Snap Defeat.” Journal of Constitutinal and Parliamentary Studies | (4):55-56. Mukherji, Panchanandas 1920 = The Indian Constitution ... and all Relevant Documents Relating to the Indian Constitutional Reforms of 1919. Calcutta: Thacker, Spink & Co. Murdoch, J., ed. 1896 Caste: Its Supposed Origin; Its History; Its Effects; the Duty of Govern-

ment, Hindus and Christians with Respect to It; and Its Prospects. Com-

piled from Muir, Max Weber, Sherring, Wilson, Monier Williams, Cornish and Others. Madras: The Christian Literature Society for Murphy,

India. 62 pp.

Arthur, B. Errabi, N. R. M. Menon, Narasimhaswamy Mysore,

and B. Sivaramayya 1973 Law and Poverty: Cases and Materials. Bombay: N. M. Tripathi. Murphy, Walter 1959 “‘Lower-Court Checks on Supreme Court Power.” American Political Science Review 53:1017-31. Murthy, B. S.

c. 1972 Depressed and Oppressed (Forever in Agony). New Delhi: S. Chand & Co. Nair, Kusum

1962

Blossoms in the Dust: The Human Factor in Indian Development. New York: Frederick A, Praeger.

Namboodiripad, E. M.S.

1979

“Caste

Conflicts vs. Growing

Unity

of Popular

Forces.” Economic and Political Weekly 14:329-36.

Democratic

Naoroji, Dadabhai n.d. Speeches and Writings of Dadabhai Naoroji. 2d ed. Madras: G. A. Natesan & Co. Narayana, G. 1974 “The Political Elite among the Scheduled Castes.” Journal of ConStitutional and Parliamentary Studies 8:214-19. 1978

“Social Background of Scheduled Caste Lok Sabha Members, 1962-71.” Economic and Political Weekly 13:1603-8.

594

Bibliography

1980

‘Rule Making for Scheduled Castes: Analysis of Lok Sabha Debates, 1962-71" Economic and Political Weekly 15:433-40. Narayana Rao, K. V. 1970

= State and the Harijan in Andhra Pradesh. Ministry of Home Affairs, Research and Policy Division. New Delhi. Natarajan, D.

1972

Extracts from the All-India Census Reports on Literacy (Census Centenary Monograph No. 9). New Delhi: Office of the Registrar Gen-

eral, Ministry of Home Affairs. Natarajan, Swaminath 1959 —_A Century of Social Reform in India. London: Asia Publishing House. National Convention

of Christian

Scheduled Caste Origin 1978

Leaders on the Plight of Christians of

The Story of 30 Years Struggle for Justice: A Consolidated Account of 30

Years’ Effort in the Cause of Christians of Scheduled Caste Origin. Bangalore: National Centre (NBCLC)

National Council of Applied Economic Research 1970

Survey of Backward Districts of Andhra Pradesh. New Delhi: National

1971

Educational and Economic Condition, and Employment Position of Eighteen

Council of Applied Economic Research. National Council of Educational Research and Training Nehni, S. S. 1932

Tribes. New Delhi: Delhi Press.

Caste and Credit in the Rural Area, A Survey.

Green & Co.

Calcutta:

Longmans,

Newell, William H.

1961

“The Comparative Study of Caste in India and Japan.” Asian Survey 1:3-10.

Nigam, S.S. 1960 ‘Equality

and

the Representation

of the Scheduled

Parliament.” Journal of the Indian Law Institute 2:297-320.

Classes in

O'Malley, L. S. S.

1932

Indian Caste Customs. Cambridge: University Press.

1941

Modern India and the West: A Study of the Interactions of their Civiliza-

1976

Cultural Revolt in a Colonial Society: The Non-Brahman Movement in

O'Malley, L.S. S., ed.

tions. London: Oxford University Press. Omvedt, Gail

Western India. 1873 to 1930. Bombay: Scientific Socialist Education Trust.

O'Neil, Robert M. 1975 Discriminating Against Discrimination: Preferential Admissions and the DeFunis Case. Bloomington: University of Indiana Press. Oommen, T. K.

1968 1972

“Strategy for Special Change: A Study of Untouchability.” Economic and Political Weekly 3:933-36. Charisma, Stability and Change: An Analysis of the Bhoodan- Gramdan Movement. New Delhi: Thomson Press.

Bibliography

595

Ouwerkerk, Louise 1945 The Untouchables of India. London: Oxford University Press. 47 pp. Pai Panandiker, V. A.

1964

“Classification of Services in the Central Government.” Indian

Journal of Public Administration 10:647-6 1. Palkhivala, N. A. ed. 1973 A Judiciary Made to Measure, Bombay: M. R. Pai.

Pandya, B.V. 1959 Striving for Economic Equality. Bombay: Popular Book Depot. Panikkar, K. M. 1933 Caste and Democracy. London: Hogarth Press. 39 pp.

1961

Hindu Society at Cross Roads. 3d ed. New York: Asia Publishing

1971

Politics and Religion: A Study of Historical Interaction between Socio-

House. Parvathamma, C.

political Relationships in a Mysore Village. New Delhi: Sterling Publishers. Parvathamma, C., and R. T. Jangam 1969 = “India’s Scheduled Caste MPs: A Socio-Economic Profile.” Jour-

nal of Karnatak University—Social Sciences 5: 153-66.

Patankar, Bharat, and Gail Omvedt 1979 “The Dalit Liberation Movement in Colonial Period.” Economic

and Political Weekly 14:409-24.

Patawardhan, Sunanda 1973 Change Among India’s Harijans: Maharashtra—A

Delhi: Orient Longman.

Case Study. New

Pennock,J, Roland, and John W. Chapman

1967

Equality. New York: Atherton Press.

Phadnis, Urmila

1969

“‘Neo-Buddhists in India and Ceylon.” Economic and Political Weekly 4: 1897-98. Pickett,J. Waskon 1933

Christian Mass Movements in India: A Study with Recommendations. New

York: Abingdon Press. Ponnarangam, V. A. 1958 The Discovery of India’s Plan. Madras: n.p. Popkin, William D. 1962

“Prematurity

and Obiter Dictum in Indian Judicial Thought.”

Journal of the Indian Law Institute 4:23 1-60.

Pradhan, M. C.

1966

The Political System of the Jats of Northern India. Bombay: Oxford

University Press. Prakash, Kailash 1967 “The Annual Assessment

and

Promotion

Systems

in Indian

Bureaucracy.” Indian Journal of Public Administration 13:268-82.

Pyarelal (Nair)

1932

The Epic Fast. Ahmedabad: Navajivan Publishing House.

596

Bibliography

Rachiah, N.

1962

Rachiah’s Speeches. Bangalore: Government Press.

Radhakrishnan, N. 1964 “Reservation to the Backward Classes.” Indian Yearbook of International Affairs 13:293-345. 1965

1966

“Units of Social, Economic and Educational Backwardness: Caste

and Individual.” Journal of the Indian Law Institute 7:262-72.

“‘Chitralekha v. State of Mysore:Art. 15(4)—Meaning of Socially

and Educationally Backward Classes.” Jaipur Law Journal 6:113— 21. Rajagopala Aiyangar, N. 1937

The Government of India Act, 1935, with a Commentary, Critical and Explanatory. Madras: Madras Law Journal Office.

Rajagopalachari, C. 1933 Plighted Word. Being an Account of the History of the Untouchability Abolition and Temple Entry Bills. Delhi: Servants of Untouchables Society. 34 pp. 1946 Ambedkar Refuted. Bombay: Hind Kitabs. 39 pp. Rajah, M. C. 1925 The Oppressed Hindus. Madras: Huxley Press. Rama Rao, T. S.

1967

“Protective Discrimination and Educational Planning.” In Educational Planning: Its Legal and Constitutional Implications in India, edited

by G. S. Sharma. Bombay: N. M. Tripathi. Ramanathan, G.

1965

Indian Babu (A Study in Social Psychology). New Publications.

Delhi: Sudha

Ramaswamy, K. V.

1934

India’s Untouchable Saints. Madras: G. A. Natesan & Co. 47 pp.

1978

“The Belief System of the Non-Brahmin Movement in India: The Andhra Case.” Asian Survey 18:290-300.

Ramaswamy, Uma Rao, D. M.

1963

“Disparities of Representation Among Direct Recruits to the 1.A.S.” Indian Journal of Public Administration 9:88-94.

Rao, M.S. A. 1967 “Education, Social Stratification and Mobility.” In Papers in the

1968

Sociology of Education in India, edited by M. S. Gore, 1.P. Desai, and Suma Chitnis. New Delhi: National Council of Educational Research and Training.

“Political Elite and Caste Association: A Report ofa Caste Conference.” Economic and Political Weekly 3:779-82. Rao, V. K.R. V. 1972 Keynote Address on “Social Change.” Bangalore: Polygraph Press.

15pp. Ratnam, K. J. 1965

Communalism and the Political Process in Malaya.

University of Malaya Press.

Kuala Lumpur:

Real Indian National Congress Party

Bibliography

597

1945 Destroy This Brahmin Raj. Bezwada: Sath Brahmana Ashram. 72 pp. Republican Party of India (B. D. Khobrogade, General Secretary) The Charter of Demands: Submitted to Shri Lal Bahadur Shastri, the 1964 Honorable Prime Minister of India. New Delhi. 32 pp.

Retzlaff, Ralph H.

1963

‘The Problem of Communal Minorities in the Drafting of the Indian Constitution.” In Constitutionism in Asia, edited by R. N. Spann. Bombay: Asia Publishing House.

Revankar, Ratna G.

1971

The Indian Constitution: A Case Study of Backward Classes. Rutherford, N. J.: Faireleigh Dickinson University Press. Risley, Sir Herbert 1908 Peoples of India. London: W. Thacker Co. (2d. ed. 1915).

Robbin, Jeanette.

1964

Dr. Ambedkar and His Movement. Hyderabad: Dr. Ambedkar Publication Society.

Roy, Sarat Chandra 1928 Oraon Religion and Customs. Ranchi: “Man in India” Office. Rudolph, Lloyd I., and Susanne Hoeber Rudolph

1967

The Modemity of Tradition: Political Development in India. Chicago:

Universiy of Chicago Press.

Rudolph, Susanne Hoeber, and Lloyd I. Rudolph, eds. 1972

Education and Politics in India: Studies in Organization, Society and Politics. Delhi: Oxford University Press. Saberwal, Satish

1972

“The Reserved Constituency: Candidates and Consequences.” Economic and Political Weekly 7:71—-80 Sachchidanada, assisted by A. K. Lal and S. N. Tiwary. 1977 The Harijan Elite: A Study of Their Status, Networks, Mobility and Role in Social Transformation. Faridabad: Thomson Press. Saksena, R. N. 1965 “Factor in Social Tension.” Seminar 70 (June):22-25. Sanjana, J. E.

1946

Caste and Outcaste. Bombay: Thacker & Co.

1946

The Fight Against Untouchability: A Brief Review of the Phenomenal Progress Made During the Last Three Decades in the Eradication of an Ancient Evil. New Delhi: Hindustan Times.

Santhanam, K.

Santhanam, K. 1964 Transition in India and Other Essays. Bombay: Asia Publishing House.

Sapru, Tej Bahadur 1945 Constitutional Proposals of the Sapru Committee. Moradabad: Sapru Committee. Saraswathi, S.

1974

,

= Minorities in Madras State: Group Interests in Modem Politics. Delhi: Impex India.

598

Bibliography

Sastri, D. Raghavachandrayya Sath. 1946 The Brahmin Non-Brakmin Hindu Problem. Bezwada: Sath Brahmana Ashram. 30 pp.

Sathi, Chedi Lal n.d.

1967

Nehru Sarkar Dvara Pichara Vargo ke Adhikaro ki Hatya? [Crushing Down of Rights of Backward Classes in Nehru Government?], Lucknow. [c. 1960] Educational Planning: Its Legal and Constitutional Implications in India. Bombay: N. M. Tripathi.

Sharma, R. C, and C. L. Sapra

1969

Wastage

and

Stagnation

in

Primary

and

Middle

Schools

in

India. New Delhi: National Council of Educational Research and

Training.

Sharma, R. S.

1958

Sudras in Ancient India: A Survey of the Position of the Lower Orders down

to circa A.D. 500. Delhi: Motilal Banarsidass.

Sheridan, L. A. 1961 The Federation of Malaya Constitution: Text, Annotations and Commentary. Singapore: University of Malaya Law Review. 1962 “Equal Opportunity of Public Employment (Non-discrimination in Public Service in India).” International and Comparative Law

Quarterly \1:782-805.

Shiva Rao, B. (Chairman of Project Committee), and others 1968 § The Framing of India’s Constitution: A Study. New Institute of Public Administration.

Delhi: Indian

Shrivastava, B. N.

1968

Exploitation in Tribal Areas. New Delhi: Bharatiya Adimjati Sevak Sangh.

Shukla, V.N.

1967

“Equal Protection Clause, Selective Techniques and Educational

Planning.” In Educational Planning: Its Legal and Constituional Impli-

cations in India, edited by G. S. Sharma. Bombay: N. M. Tripathi.

Silverberg, James, ed.

1967

Social Mobility in the Caste System in India, an Interdisciplinary Symposium, The Hague: Mouton Publishers. Supplement III to Comparative Studies in Society and History.

Sinder, Leon

1963

Caste Instability in Moghul India.” Seoul, Korea: Culture Research Center, Chung-ang University.

Singer, Milton, and Bernard S. Cohn, eds.

1968

International

Structure and Change in Indian Society. Chicago: Publishing Co. Singh, Harjinder 1967 “Social Grading of Castes and Occupations in an Village.” Indian Journal of Social Work 27:38 1-96.

Singh, Mohinder

1947

Aldine Indian

The Depressed Classes: Their Economic and Social Condition. Bombay:

Hind Kitabs.

Bibliography

599

Singh, Parmanand

1976

“*Equal Opportunity’ and ‘Compensatory Discrimination’: Con-

stitutional Policy and Judicial Control,” Journal of the Indian Law

Institute 18:300-19

Singh, Saint Nihal

1913

“India’s ‘Untouchables’ ”. Contemporary Review 103:376-85.

Saxena, Ishwar Chandra 1963 “The Doctrine of Precedent in India: A Study of Some of Its

Aspects.” Jaipur Law Journal 3: 188-214.

Sayaji Rao, III, Maharaja of Baroda c.1912 “The Depressed Classes.” In The Depressed Classes: An Enquiry into

Their Condition and Suggestions for Their Uplift. Madras: G. A. Nate-

san & Co. [Reprinted from the Indian Review of Dec. 1909].

1928

Speeches and Addresses. London: Macmillan Co.

1969

“Scheduled Caste Welfare: Public Priorities in the States.” Economic and Political Weekly 4:397-401 Comparative Ethnic Relations. New York: Random House.

Schermerhorn, R. A.

1970

Schmitthener, Samuel

1968-69 “A

Sketch

of the

Development

of the

India.” Law & Society Review 3:337-82.

Legal

Profession

in

Schuchter, Arnold

1970

Reparations: The Black Manifesto and Its Challenge to White America.

1978

A Historical Atlas of South Asia. Chicago:

Philadelphia: Lippincott. Schwartzberg, Joseph E., ed. Press.

Scott, Roland W.

1953

Social Ethics in Modem Calcutta: n.p.

Hinduism:

University of Chicago

The Religious Quest of India.

Seervai, H. M. 1968 Constitutional Law of India. Bombay: N. M. Tripathi. Seminar 1965 “Secularism.” No. 67. New Delhi. Setalvad, M. C.

1969

“‘President’s Page.” Indian Advocate 9 (9):1.

1977

Who Goes to College? Scheduled Caste/Tribe Post-Matric Scholars in Gujarat, Akmedbad: Rachana Prakashan.

Shah, Vimal P., and Tara Patel

Shapiro, Martin

1975

“Courts,”

In

A

Handbook

1968

“Secular State and Civil Service.” In Secularism in India, edited by

N. Polsby and F. Greenstein. Publishing Co., Inc. Sharma, B. A. V.

Political

Science,

Massachusetts:

edited

by

Addison-Wesley .

V. K. Sinha. Bombay: Lalvani.

Sharma, G. S.

1965

of

“Concept of Leadership Implicit in the Directive Principles of State

600

Bibliography Policy in the Indian Constitution.” Journal of the Indian Law Institute

7:174-88.

Sharma, G. S., ed.

1966

Secularism: Its Implications for Law and Life in India. Bombay: N. M.

1980

“Some Reflections on the Indian Experience with the Policy

Tripathi.

Reservation”. Unpublished ms.

of

Sinha, D.N. 1967 “Educational Planning: Discrimination and the Equal Protection of the Laws.” In Educational Planning: Its Legal and Constitutional Implications in India, edited by G. S. Sharma. Bombay: N. M.

Tripathi.

Sinha, V. K., ed.

1968

Secularism in India. Bombay: Lalvani.

1963

When Caste Barriers Fall: A Study of Social and Economic Change in an Indian Village. New York: Humanities Press.

Sivertsen, Dagfinn

Smith, Donald E. 1963

India as a Secular State, Princeton: Princeton University Press.

1965

South Asian Politics and Religion. Princeton: Princeton University

Smith, Donald E., ed

Press.

Smith, William Roy

1938

Nationalism and Reform in India. New Haven: Yale University Press.

1978

Israel: Pluralism and Conflict. Berkeley:

Smooha, Sammy

Press.

Srinivas, M. N.

1956

‘A

1962

Caste in Modern India and Other Essays. London: Asia Publishing House. Social Change in Modem India. Berkeley: University of California Press.

1966

Note

on Sanskritization

Quarterly 15:8 1-96.

and

University of California

Westernization.”

Far Eastern

Srinivas, M. N., Y. B. Damle, S. Shahani, and André Bétéille

1959

“Caste: A Trend Report and Bibliography.” Current Soctology/La Sociologie Contemporaine 7:135-83.

Srivastava, L., and others

1971

Subbiah, A. 1963

Educational and Economic Condition and Employment Position of Eighteen

Tribes. New Delhi: National Council of Educational Research and

Training.

‘The Caste System: Vested Interests of the High Castes.” Quest no. 37,pp. 85-7. Subramaniam. V. 1971 Social Background of India’s Administrators. New Delhi: Publications Division, Ministry of Information and Broadcasting.

Bibliography Suffian, Tan Sri Mohamed 1972 An Introduction to the Constitution of Malaysia.

Dichetak di-Jabatan Chetak Kerajaan.

Sunshine, Russell B., and Berney, Arthur L.

1970

“Basic Legal Education in India: An

Kuala

601

Lumpur:

Empirical Study of the

Student Perspective at Three Law Colleges.” Journal of the Indian Law Institute 12:39-118. Taub, Richard P. 1969 Bureaucrats under Stress: Administrators and Administration in an Indian State . Berkeley: University of California Press. Thakkar, A. V.

1949

‘The Problem of Aborigines in India.” (The R. R. Kale Memorial Lecture, Gokhale Institute of Politics and Economics, 1941). In Thakkar Bapa: Eightieth Birthday Commemoration Volume, edited by T. N. Jagadish and Shyamlal. Madras, n.p.

Tinker, Hugh 1962 India and Pakistan: A Political Analysis. New York: Praeger.

Titus, Murray T.

1959 Islam in India and Pakistan, Calcutta: YMCA Publishing House. Tope,T. K. 1967

“Protective Discrimination and Educational Planning.” In Educational Planning: Its Legal and Constitutional Implications in India, edited

by G. S Sharma. Bombay: N. M. Tripathi.

Tripathi, P. K.

1966

1972

“Secularism: Constitutional Provision and Judicial Review.” Journal of the Indian Law Institute 8: |-29.

Some

Insights

Bombay.

into

Fundamental

Rights.

Bombay:

University

of

Vagiswari, Alladi

1972

Income Earning Trends and Social Status of the Harijan Community in Tamil Nadu. Madras: Sangam Publishers.

Varadarajalu Naidu, T., comp.

1948

The National Dharma: Life, Speeches and Writings of Dr. P. Varadarajulu Naidu. Salem: The Tamil Nadu Co, Varadarajulu Naidu, T. 1931

The Justice Movement, Madras: Justice Printing Works.

Venkatarangiengar, S. K. 1961 “Article 16 of the Constitution and the Application of the Equality Clause to Civil Servants.” Indian Advocate | (3):29-34. Venkatasubramanian, C. S.

1965

“Separate Concurring Judgments in the Supreme Court,” A./.R. 1965 Journal 3.

Verba, Sidney, Bashiruddin Ahmed, and Anil Bhatt

1971

Caste, Race and Politics: A Comparative Study of India and the United States, Beverly Hills: Sage Publications.

Verma, Brind Bihari 1956 Agriculture and Land Ownership System Among the Primitive People of

Assam. Delhi: Bharatiya Adimjati Sevak Sangh.

602

Bibliography

Vivekananda, Swami 1955 Caste, Culture and Socialism. Calcutta: Advaita Ashrama. Von Furer-Haimendorf, Christoph

1967.

“The Position of the Tribal Populations in Modern India.” In India and Ceylon: Unity and Diversity, edited by Philip Mason. London: Oxford University Press.

Von Mehren, Arthur T.

1963a

1936b

“The Judicial Process with particular reference to the United States and to India.” Journal of the Indian Law Institute 5:27 1-80 “Some Observations on the Role of Law in Indian Society and on Indian Legal Education.” Jaipur Law Journal 3:13-18.

Wadhwa, Kamlesh Kumar 1975

Minority Safeguards in India (Constitutional Provisions and Their Imp-

lementation). Delhi: Thomson Press. Wagley, Charles, and Marvin Harris

1958

§ Minorities in the New World: Columbia University Press.

Six

Case

Walsh, Sir Cecil 1930 Crime in India, with an Introduction on Peculiarities. London. Ernest Benn, Ltd. Wasby, Steven

1970

The Impact

Homewood,

of the United States Supreme IIl.: Dorsey Press.

Studies.

Forensic

Court:

New

York:

Difficulties

and

Some Perspectives.

Weiner, Myron, and John Osgood Field

1975

“How Tribal Constituencies in India Vote.” In Electoral Politics in

the Indian States: Three Disadvantaged Sectors, edited by Jagdish Bhagwati et al. Vol. IT. Studies in Electoral Politics in the Indian States. Delhi: Manohar Book Service.

Weitbrecht-Stanton, H. U.

1920.

“The Untouchables Review 16:17 1-84. Whitcombe, Elizabeth

of India and Their Enumeration.” Asiatic

1972

Agrarian Conditions in Northern India. Vol. 1, The United Provinces Under

British Rule, 1860-1900. Berkeley: University of California Press.

Wilson, John

1877

Indian Caste. 2 vols. Edinburgh: William Blackwood 1976, New Delhi: Deep Publications. Wiser, William Henricks 1958 The Hindu Jajmani System. A Socio-economic System Interrelating Members of a Hindu Village Community in Services. Lucknow: Lucknow Publishing House. Wiser, William H., and Charlotte Viall Wiser

1963

Behind Mud Walls. 1930-1960. Berkeley: University of California

Press.

World Bank 1975 Health: Sector Policy Paper. Washington D.C: World Bank.

Bibliography 603 Yadava, Jai Narayan Singh 1955 The Problems of India’s Backward Classes. Delhi: Pahari Dhiraj, Ahiram. Zelliott, Eleanor

1966

1969

“Buddhism and Politics in Maharashtra.” In South Asian Politics and Religion, edited by Donald E. Smith. Princeton University Press.

“Dr. Ambedkar and the Mahar Movement.” Ph. D* dissertation,

University of Pennsylvania. Zinkin, Taya 1962 Caste Today. Issued under the auspices of the Institute of Race Relations. London: Oxford University Press. ZoBell, Karl M.

1959

= “Division of Opinion in the Supreme Court: A History of Judicial Disintegration.” Cornell Law Quarterly 44: 186-214.

GoveRNMENT DocumENTS Pre-Independence

Bombay (Presidency). Depressed Classes and Committee [O.H.B. Starte, Chairman] 1930 Report, March 1930. Bombay: Printed

Aboriginal at

Central Press. Great Britain. India Office. Indian Constitutional Reforms 1919

Government

Reports of the Franchise Committee and the Committee on Division of Functions [Lord Southborough, Chairman]. Calcutta: Superintendent of

Government Printing. Great Britain. Indian Franchise Chairman] 1932

the

Tribes

Committee

[Marquess

. of Lothian,

Report. Vol. 1 Calcutta: Government of India, Central Publication Branch.

Great Britain. Indian Round Table Conference [Second Session, 7 Sep. 1931-1 Dec. 1931] 1932 Sub-committees’ Reports; and Prime Minister’s Statement. Calcutta: Government of India, Central Publications Branch.

Great Britain. Indian Statutory Commission.

1930a

Memoranda

Submitted by the Government

of India and the India

Office.Vol. 5. Calcutta: Government of India, Central Publications Branch. Great Britain, Indian Statutory Commission 1930b Report, Vol. I (Survey). Calcutta: Government of India, Central

Publications Branch.

Great Britain. Indian Statutory Commission. Growth of Education [P. Hartog, Chairman]

1929

Auxilliary

Committee

on

Interim Report of the Indian Statutory Commission. London: His Majesty’s Stationery Office.

604 Bibliography India. Indian Central Committee [Sankaran Nair, Chairman] 1929

Report.

Branch.

Calcutta:

Government

of

India,

Central

Publication

India. Reforms Enquiry Committee [A.P. Muddiman, Chairman] 1925 Report of the Reforms Enquiry Committee, 1924 . . .and Connected Papers. London: His Majesty’s Stationery Office.

Post—Independence: Central Government (India, Dominion) Constituent Assembly Debates: Official Report. v. 1-12; Dec. 9, 1946—Jan. 24, 1950. Delhi: Manager

of Publications.

Post-Independence: Central Government (India, Republic) Administrative Reforms Commission

1969

Report of the Study Team on Recruitment, Selection, U.P.S.C./State

P.S.Cs. and Training. Delhi: Manager of Publications. All-India Agricultural Labour Enquiry

1954

1955

= Agricultural Labour:

How

They

Work and Live; Essential Stati-

shes. Balakrishna Ramamurti for the All-India Agricultural Labour Enquiry. Forward by V. V. Giri. Delhi: Manager of

Publications. Report on Intensive Survey of Agricultural Labour, Employment, Underemp-

loyment, Wages and Levels of Living. Delhi: Man: Vol.

1: All India; Vol. 2: North India; Vol. 3:

of Publications.

it India; Vol. 4:

South India; Vol. 5: West India; Vol. 6: Central India; Vol. 7:

North West India.

Backward Classes Commission. (K. Kalelkar, Chairman)

1955

Report. 3 Vols. Delhi: Manager of Publications.

Backward Classes Commission. (B.P. Mandal, Chairman) 1981 Report: 7 Vols. in 2. New Delhi: Controller of Publications Central Social! Welfare Board

1966-68, 1972-76 Reports. Mimeo. New Delhi. Commission on the Hill Areas of Assam 1966

Report, 1965-66. [H. V. Pataskar, Chairman]. New Delhi: Govern-

ment of India. Commissioner for Scheduled Castes and Scheduled Tribes 1951-56, 1957-58 to 1978-79 Reports. New Delhi: Manager of Publications. 1968 Handbook on Scheduled Castes and Scheduled Tribes. Delhi: Manager of Publications.

Department of Community Development

1966 _ Panchayati Raj at a Glance as on 31 March 1966. New Delhi: Ministry of Food, Agriculture, Community Development and Cooperation.

Department of Personnel 1972 Brochure on Reservation for Scheduled Castes and Scheduled Tribes in Services. (3d ed.)

Delhi: Manager of Publications.

Department of Social Security 1965 Report of the Advisory Committee on the Revision of the Lists of Scheduled Castes and Scheduled Tribes [Lokur Committee]. Delhi.

Bibliography 605, Department of Social Welfare 1965-66 Reports. New Delhi: Manager, Government of India Press. to

1975-76 1966 Report of the Committee on Customary Rights to Scavenging [N. R. 1968 1969

Malkani, Chairman]. Faridabad: Government of India Press.

A Statistical Handbook of Tribal Welfare and Development. New Delhi: Manager, Government of India Press.

Report of the Committeeon Untouchability, Economic and Educational Development of the Scheduled Castes [Elayaperumal Committee]. Delhi. Directorate of Backward Classes Welfare 1973-74 The Newsletter. New Delhi: Ministry of Home Affairs. Election Commission 1955 Report on the First General Elections in India, 1951-52. 2Vols. Delhi: Manager of Publications. 1959 Report on the Second General Elections in India, 1957. 2 Vols. Delhi: Manager of Publications. 1966 Report on the Third General Elections in India, 1962-66. 2 Vols. Delhi: Manager of Publications. 1968 Report on the Fourth General Elections in India, 1967. 2 Vols. Delhi: Manager of Publications. 1973 Report on the Fifth General Elections to the House of the People in India, 1976a

1971. Vol. 2 (Statistical.) New Delhi: Manager of Publications.

Delimitation of Parliamentary and Assembly Constituencies Order, 1976. New Delhi: Government Press. 1976b Report on the General Elections to the Legislative Assemblies in India, 1970-72. Vol. 2-A (Statistical) except U.P., Nagaland, and Pondicherry. Delhi: Government Press. Expert Committee on Legal Aid [V. K. Krishna Iyer, Chairman] 1974 — Processual Justice to the People. Report of the Expert Committee on Legal Aid. May, 1973. New Delhi: Department of Legal Affairs, Ministry of Law, Justice and Company Affairs. Law Commission

1958

Fourteenth Report (Reform of Judicial Administration), Vol. I. Delhi:

Manager of Publications. Ministry of Community Development and Cooperation, Department of Cooperation 1964 Conference of Registrars of Cooperative Societies, and Directors— in-Charge of Backward Classes and the State Ministers-in-Charge of Welfare of Backward Classes and State Minister-in-Charge of Cooperation; May, 1964. Proceedings and Agenda Notes. Simla: Manager, Government of India Press.

Ministry of Education 1962 Report of the Committee on Emotional Integration [Sampurnand, Chairman)}Delhi. Ministry of Education. Education Commission 1966 Report of the Education Commission 1964-66. Delhi: Manager of Publications.

606 Bibliography 1970

Report of the Education Commission, 1964-66, Vol. 2 (Supplement). Delhi: Ministry of Education and Youth Services. Ministry of Education and Scientific Research 1957 Progress of Scholarships for Scheduled Castes, Tribes and Other Backward Classes. 4th ed. Delhi: Manager, Government of India Press. Ministry of Education and Social Welfare 1955 Education in India, 1951-52. Delhi: Manager of Publications.

1968

1973

Education in India, 1961-62. Delhi: Manager of Publications.

Report, 1972-73. New Delhi: Manager of Publications.

Ministry of Finance 1959 Report of the Commission of Enquiry on Emoluments and Conditions of

Service of Central Government Employees, 1957-1959. Delhi. Ministry of Home Affairs. 1956 § Memorandum on the Report of the Backward Classes Commission. Delhi: 1962a

1962b

Government of India Press.

Conference of State Ministers in Charge of Welfare of Backward Classes. Agenda and Notes; 26-27 July 1962. Mimeographed. 71 pp.

Report of the Special Working Group on Cooperation for Backward Classes

[M.P. Bhargava, Chairman]. 2 vols. New Delhi: Manager, Government of India Press. Ministry of Information and Broadcasting 1952 Harijans Today. Delhi: Director, Publications Division. 1953 India: A Reference Annual. New Delhi. 1963 Welfare of the Backward Classes. Delhi: Director, Publications

Division.

Ministry of Law, Justice and Company Affairs 1977 Judges of the Supreme Court and the High Courts, as on 1-4-77, Delhi: Ministry of Law, Justice and Company Affairs.

1978

Parliament

Report on National Juridicare: Equal Justice—Social Justice [P.N.

Bhagwati, Chairman]. New Delhi: Ministry of Law, Justice and

Company Affairs.

1950-52 Parliamentary Debates. Official report.

Parliament.

House of the People (Fourth Lok Sabha). Committee on the

Welfare of Scheduled Castes and Scheduled Tribes 1969-70 Reports 1-19. New Delhi: Lok Sabha Secretariat.

Parliament. House of the People (Fifth Lok Sabha). Committee on the Welfare of Scheduled Castes and Scheduled Tribes. 1971-76 Reports 1-14, 16-38, 40-51. New Delhi: Lok Sabha Secretariat.

Parliament. Lok Sabha [House of the People]

1952Debates. Delhi: Lok Sabha Secretariat. Parliament. Rajya Sabha

1952— Debates. Delhi: Rajya Sabha Secretariat. Planning Commission 1951

1953

The First Five Year Plan: A Draft Outline. New Delhi: Manager of

Publications.

First Five Year Plan: Questions and Answers. Delhi: Publications Division, Ministrv of Information and Broadcasting.

Bibliography 1960 1961 1965 1970 1976

607

Social Welfare in India, Forward by Jawaharlal Nehru. Revised and abridged ed. Delhi: Publications Division, Ministry of Information and Broadcasting.

Third Five Year Plan. Delhi: Manager of Publications. Report of the Seminar on Employment of Scheduled Castes and Scheduled Tribes. Delhi: Manager of Publications. Fourth Five Year Plan, 1969-74. Delhi: Government of India. Fifth Five Year Plan, 1974-79. Delhi: Controller of Publications.

Planning Commission, Programme Evaluation Organization 1956 Study of Village Artisans. Delhi: Government Press. Scavenging Conditions Enquiry Committee [N. R. Malkani, Chairman] 1960 Report. Delhi: Manager of Publications.

Scheduled Areas and Scheduled Tribes Commission [U. N. Dhebar, Chairman] 1961 Report, 1960-61. 2 vols. Delhi: Manager of Publications, Civil Lines.

Study Group on the Welfare of the Weaker Community b. P. Narayan, Chairman] 1961,

Sections of the Village

1963 Report, 2 vols. Delhi: Ministry of Community Panchayati Raj and Cooperation.

Development,

Study Team on Social Welfare and Welfare of Backward Classes [Renuka

Ray, Chairman]

1959 Report. 1 vol. New Delhi: Committee on Plan Projects. Study Team on Tribal Development Programmes 1970 Report. New Delhi: Committee on Plan Projects, Planning Commission. University Education Commission

1950-51 Report. Dec. 1948-Aug. 1949. 2 vols. Delhi: Ministry of Education.

Post—Independence: States Andhra

Pradesh.

Backward

Classes

Commission

Chairman] 1970 Report. Hyderabad. Andhra Pradesh. Department of Social Welfare 1975

Bihar 1968

Performance Budget, Central Press.

1975-76.

[Manohar

Hyderabad:

Pershad,

Government

A Report on the Situtation in the Tribal Areas of Bihar. Patna: Printed by the Superintendent, Secretariat Press.

Bihar. Welfare Department 1966a

Annual Administration Report. 1964-65. Patna: Superintendent, Sec-

1966b

The Bihar Welfare Manual:

retariat Press.

:

Compendium of Important Circulars and

Letters Relating to the Programme for the Welfare of Backward Classes. Patna: Superintendent, Secretariat Press. Bombay. Backward Classes Department 1957, 1958 Administration Report on the Working of the Backward Classes Department 1954—55, 1955-56. Poona: Printed at the Yervada Prison Press.

608 Bibliography Bombay. Director of Backward Class Welfare 1955

Privileges Provided by Bombay State Government for Backward Classes. Poona: Yervada Prison Press. Bombay. Ex-Criminal Tribes Rehabilitation Committee [K. B. Antrolikar, Chairman] 1951 Report, 1950. Bombay: Government Central Press.

Gujarat. Directorate of Social Welfare 1963-64 Annual Administration Report on the Working of the Social Welfare Department. Ahmedabad: Stationery.

Government

Printing,

Publications

and

Gujarat. Education Department

1959-60 Administration Report. Anmedabad: Government Central Press. Gujarat. Socially and Educationally Backward Classes Commission [A. R.. Bakshi, Chairman]

1976

Haryana.

Report of the Socially and Educationally Backward Commission. 2 vols. Ahmedabad: The Commission.

Classes

Legislative Assembly. Committee on the Welfare of Scheduled

Castes and Scheduled Tribes 1975

Jammu

Report. Ist. January. Chandigarh: Vidhan Sabha Secretariat.

and

Kashmir.

Backward

Classes

Committee

[J.

N.

Wazir,

Chairman] 1969 Report. Srinagar: Printed at the Government Press. Karnataka. Karnataka Backward Classes Commission [L. G. Havanur, Chairman]

1975

Kerala

Report, Karnataka Backward Classes Commission, Constituted under the Commission of Inquiry Act, 1952 (Central Act 60 of 1952). 4 vols. in 5 parts. Bangalore: Government of Karnataka.

1960

Kerala Services Manual. Vol. 1. Trivandrum: Superintendent of Government Printing. Kerala. Administrative Reforms Committee [E. M. S. Nambudiripad, Chairman] 1958 Report. Vol. 1, Part 1, 1958. Trivandrum: Printed at the Government

Press.

Kerala. Backward Classes Reservation Commission [Nettur P. Damodaran, Chairman] 1971 Report, 1970. 2 vols. Trivandrum: Printed at the Government Press.

Kerala. Commission for Reservation of Seats in Educational Institutions [G. Kumara Pillai, Chairman} 1966 Report. Trivandrum: Printed at the Government Press. Kerala. Harijan Welfare Department.

Administration Report. Annual. 1956-57 to 1963-64; 1967-68 to

1967

1971-72. Ernakulam: Printed at the Government Press. First Supplement to the Hand Book of the Department of Harijan Welfare:

Rules and Standing Orders., as on Shoranur: Government Press.

0th September 1965.

Bibliography 609 Kerala. Legislative Assembly. Committee on the Welfare of Scheduled Castes and Scheduled Tribes 1974a

1974b

Report on the Alenation of Tribal Lands, Ist, 1973-74. Trivandrum: Secretariat of the Kerala Legislature.

Second Report on the Representation of Scheduled Castes and Scheduled Tribes in the Services of Various Departments under the Government of Kerala. Trivandrum: Secretariat of the Kerala Legislature. Kerala. Revenue Department 1964 A Handbook of the Department of Harijan Welfare (as on 31 August 1962). Trivandrum: The Government Press. Madras n.d.

Harijan Welfare in Madras State. Madras: Director of Information and Publicity, Government of Madras.

Madras. Harijan Welfare Department

Administration Report, Annual. 1955-56; 1958-59 to 1960-61; 197172. Madras: Government Press.

Maharashtra n.d. Schemes for the Welfare of Backward Classes and Provisions Made ThereSore in the Budget Estimates for 1973-74. Bombay: Government Central Press. Maharashtra. Backward Class Hostels Committee 1962 Report. Poona: Printed at the Yervada Prison Press. Maharashtra. Backward Classes Welfare Wing

1960-61 Annual Administration Report on the Welfare of Backward Classes. to Bombay: Director. Government Prinitng. Publications and 1966-67 Stationary. Maharashtra. Committee on Reservation for Backward Classes in the Service [B. D. Deshmukh, Chairman] 1964 Report. Bombay: Printed at the Government Central Press. Maharashtra. Legislature. Committee on the Welfare of Scheduled Castes and Scheduled Tribes 1973.

Fifth:

Report:

Education.

Bombay:

Maharashtra

Legislative

Secretariat. Maharashtra. Public Service Commission 1963 Report, 196-62. Bombay: Director, Government Printing, Publications and Stationery. Maharashtra. Social Welfare Department 1960-61 Annual Administration Report of Scheduled Areas in Maharashtra State. to

Bombay:

Director,

Government

Printing,

Publications

and

1971-72 Stationery. Mysore 1965 Administration Report on the Welfare of Backward Classes for the Year 1963-64. Bangalore: Cyclostyled. 84 pp.

Mysore. Backward Classes Committee [R. Nagan Gowda, Chairman]

1960 1961

Interim Report. Bangalore: The Government Press. Final Report. Bangalore: The Government Press.

610 Bibliography Mysore. Bureau of Economics and Statistics 1972

— Bidar District: A Survey of a Backward Area. Bangalore: n.p.

Mysore. Commissioner for Depressed Classes 1957 Summary Record and Main Recommendations of the Mysore State Depressed Class Seminar Held at Kengeri, February 11th to 14th, 1956 Mysore: Deputy Director, Govenment Branch Press. Mysore. Education Department 1969

Handbook of Scholarships and Educational Concessions for the Students of

1962

Scheduled Castes and Scheduled Tribes Appointments Committee Report.

Mysore State (as on 2nd January 1969). 2 vols. Bangalore: The Government Press. Mysore. General Administrative Department

Coorg: Manager, Government Branch Press. Mysore. Legislature. Committee on the Welfare of Scheduled Castes and Scheduled Tribes 1973 First Report: Education of Scheduled Castes and Scheduled Tribes. Banga-

lore: Mysore Legislature Secretariat. Mysore. Planning, Housing and Social Welfare Department 1962 The Welfare of Scheduled Castes in Mysore State. Bangalore: The Government Press. Mysore. Social Welfare Department 1961

Detailed

Administration

Report,

1958-59.

Bangalore:

1970

Welfare of the Weaker Sections and the Handicapped, 1969-70. (Report of

Government Press. 1968-69 Annual Administration Report (draft). Mimeographed.

The

the Social Welfare Department for 1969-70.) Bangalore: The Government Press. Orissa. Tribal and Rural Welfare Department

Administration Report. Annual. 1965-66; 1968-69; 1970-71; 1971-

1958 1972

72. Cuttack: Orissa Government Press. Welfare of Backward Classes in Orissa. Bhubaneswar: Orissa Govern-

ment Press. Report on the Annual Administration of Scheduled Areas in Orissa for the Year

1970-71. Cuttack: Orissa Government Press. Punjab. Welfare Department 1966 Report of the Evaluation Committee on Welfare Regarding the Welfare of Scheduled Castes, Backward Classes and Denotified Tribes in Punjab State Sor the Period Commencing from 15 August 1947 (December 1965—August 1966). Chandigarh: Controller of Printing and Stationery, Punjab. Tamil Nadu. Backward Classes Commission [A. .N. Sattanathan, Chairman]

1971

Report. 3 vols. Madras: Director of Stationery and Printing.

Tamil Nadu. Directorate of Backward Classes

1971-72 Performance Budget. Madras: Director of Stationery and Printing. Tamil Nadu. Harijan Welfare Department 1971-72 Performance Budget. Madras: Government Press. to 1973-74

Bibliography

611

Travancore—Cochin. Committee Enquiring into the Condition of Scheduled Hill Tribes in Travancore—Cochin [W. S. S. Mackay, Chairman.]

1956

Preliminary Report of the Committee Enquiring into the Condition of Scheduled Hill Tribes in Travancore-Cochin, 1955-56. Munnar: Imperial Press.

West Bengal. Bureau of Applied Economics and Statistics.

1971

Backwardness of the Districts of West Bengal—A Comparative Study.

Calcutta: Reliance Printing Works. West Bengal. Development Department n.d. India’s Villages. A collection of articles originally published in the Economic Weekly of Bombay. Calcutta: n.p. West Bengal. Legislature. Legislative Assembly. Committee on the Welfare of Scheduled Castes and Scheduled Tribes. 1976 Third Report, 1974-76. Calcutta: West Bengal Legislative Assembly Secretariat. West Bengal. Tribal Welfare Department. Progress, Statistics and Co-

ordination Cell

1961-62, Administration Report. Alipore: Government Press

1962-63, 1965-66

West Bengal. Tribal Welfare Department. Cultural Research Institute 1962-63, Bulletin. Annual. Calcutta

1963-64

1962

Impact of Industrialization on the Life of the Tribals of West Bengal.

1958

Report on Social Welfare and Welfare of Backward Classes in West Bengal.

Calcutta: Government of West Bengal. West Bengal. Tribal Welfare Department Alipore: Government Press.

Index

Backward Class: definition of, 3, 154-59, 185-86; education, 62, 64, 162, 177;

Abraham, Amrita, 40 Adam, William, 19

Adi Dravida, 315, 327-32 Adler, Chaim, 553

Adoption, 337-40, 341 CB. 4 Aggarwal, Partap C., 66, 67, 68, 78, 108, 109, 110, 356, 357

Ahmed, S. A., 429 Aiyappan, A., 16

Alberuni, 13 Alexandrowicz, C. H., 2, 366, 494, 498

All-India Backward Classes Federation, Ambedkar,

Bhimrao

Ramji,

18, 29-40

passim, 45, 127, 129, 307, 319, 326; in Constitutent Assembly, 39, 222, 368,

399-400; in First Amendment debate,

160, 166, 167

Anandan, P. I., 20

Anant, Santokh Singh, 75, 76, 77, 78

Anantanarayanan, M., 315

Andhra Pradesh Backward Classes Commission (1970), 101, 227, 236, 237, 244, 248, 253, 254, 260, 466 Angell, James W., 554 Anglo-Indians, 42, 294

Ant-Brahminism, 209, 447 Anti-disabilities legislation, 28, 36. see also

Untouchability (Offences) Act, 1955 Arya Samaj, 23, 29, 292-93, 303

Ashraf, Mohammed 78, 108, 356, 357

Siddiq, 66, 67, 68,

Austin, Granville, 44, 85, 363, 561-62

Ayyangar, M. A., 166 Ayyar, S. V., 79 Azad, Maulana, 162

Assembly, 159-61; Centre's policy regarding, 167-79; varied state policies

regarding,

161-63,

179-87; judicial

control on permissible “‘classes” 188-

221 passim; judicial determination of backwardness, 222-81 passim, 341-48; compartments within and layers of,

Ahmad, Nazirrudin, 146 Ahmed, Karuna, 560

162, 174, 178, 179

estimated population, 43, 158, 161, 168-69, 195; discussion in Constituent

463-72; most backward classes, 468—

70; litigation concerning, 501-13 Backward Classes Commission (1953—

55), 165, 166, 167-74, 176, 179, 180, 217, 241, 245, 260, 262

Backward

80), 187

Classes Commission

(1978-

Baines, Athelstan, 123 Bajpai, G. S., 123 Bakhale, R. R., 125, 130 Balaji, M. R., 428 Balakrishnan, B.G., 75

Banaji, D.A., 19 Banerjee, A. C., 159 Banias, 223

Barnabas, A. P., 73, 84, 175

Baroda, Maharaja of [Gaikwad]

Rao III), 25, 30 Barrier, N. G., 25 Basi, P. Reddy, 446 Bastedo, T. G., 514, 515 Baxi, Upendra, 416, 520

Becker, Theodore L., 522 Bedars, 288

Beg, M. H., 220, 385, 389, 424, 481 Behara, Baidhar, 69 Beidelman, Thomas O., 9 Berney, Arthur L., 514

(Sayaji

614

Index

Berreman, Gerald D., 12

204;

Béteille, André, 329 Bhandare, R. D., 267, 321

Cohn, Bernard S., 20

Compensatory

. Bhargava, Thakur Das, 166

Bhovis, 285 Bittker, Boris, 554 Black, Donald, 498 Blaug, Mark, 85

Discrimination

Polides:

and other requirements, 87, 372, 389, 420-21; priority of commitment to SC

and ST

Bonayji, N. B., 85

Bose, Nirmal Kumar, 168, 169 Brahmins, 10, 222, 223, 235 Brown, F. H., 123

Buddhists, 16, 39-40, 144, 307-10, 311, 319-23, 337 7; system described,

7 -13; legal position under British rule, 19-21; reform of caste as public issue,

23—25; as unit and as measure of backwardnessness, 139, 189-90, 229-30,

535; lost on conversion, 315-22, 332;

judicial views of, 318, 331-32; excom333~35;

diluted, 391, 393, 435-37, 537;

lessons from India’s experience, 563-67

Bopegamage, Albert, 235

munication,

elastic

terminology, 2-3; expenditures under five year plans, 56-57; waiver of age

Bobbitt, Phillip, 437

Caste: terminology,

and

Cohen, Stephen P., 349

Bhangis, 35

Calabresi, Guido, 437

historic, modernist

views, 188-89, 204-08, 535

constitutional

position of, 350; role in Indian life, 559-61 —telation to backws-dness: correla-

tional and causal, 209-13; absence of data, 213-15 Census: definition of Untouchables, 127— 29; eschews caste data, 164; use of old data, 259-60; collection of caste data, 260-6 1; impact on identities, 354

Chamars, 14, 137, 286, 288, 289 Chandrachud, Y.V., 445, 481 Channiah, T., 160 Chauhan, D.N.S., 51

Chaurasia; S. D. Singh, 158, 161, 168, 172, 234 Chavan, Y. B., 321

Chayes, Abram, 395

Chidambaram Pillai, P., 21

Chintamani, C. Y., 125, 130 Chitnis, Suma, 77, 79

5. L., 51 Christians, 16, 223, 291, 293, 294, 312-13, 330, 335-36

“Classes”: communities as classes, 188—

—— administration: expenditures,

65;

shortfalls

benefits

in

separate

rather than additional, 65-67; absence

of goals, 67-68; low prestige and priority, 68-69; of economic tests of backwardness, 342-45, 347-48; imy

tation of judicialrules, 351, 359, 52-33 perceptions of: by beneficiaries,

66, 77-79; by intellectuals, 73-75, 77;

by public, 75-77, 560-61 ——measurement of

2 Se

55, 72-75, 79-83; government ment, 93-117, 213-15, 356; problems of, 417—18; limitation of courts in effectuating,

541-46;

effects

marized, 547-52: ——Justifications: fairness, 552-59; nondiscrimination

sum-

114-15, theme,

552-53; general welfare theme, 55354; reparations theme, 554~555, 556-59

——maldistribution: problem,

“‘lion’s

share”

137, 270, 276; “creaming”,

468-69 Commission for the Scheduled Castes and Scheduled Tribes, 71

Commissioner for Scheduled Castes and

Tribes, 70-71, 161, 541

Communal Award, 31-33 Community, Indian usage of term, 16-17

Constituent

Assembly,

363, 368, 371, 399-400

37,

39,

159-61,

Constitution of India: basic format, 37— 39; Directive Principles, 41, 365, 366,

397, 443, 538; Article 338, 70; debate over Article 15(4), 165-67; forbidden

classifications, 215—16; burden of proof as to backwardness, 218-20; tests for

allowable classifications, 218, 314; jud-

icial requirement of “reasonableness” in compensatory discrimination provi-

EPR Re ie RE

OS

Index sions,

220-21;

Article

12, 313,

129;

antidiscrimination provisions, 364; Article 335, 368, 398; its relation to Article 16(4), 369; Article 320(4), 369; Ar-

ticle 46, 397; Forty-Second

ment, 478, 520, 530, 538

Conversion:

among

Scheduled

Amend-

Tribes,

138, 316-17; to Christianity, 312-13,

324-25; loss of caste theory, 315, 327-

31, 332; reconversion. 327-28; to ob-

tain benefits, 337 Coupland, R., 375

scholarships,

62,

162,

formal

vs. substantive,

“horizontal

378,

and_

vertical”

equality,

384-89;

views, 379-80; classification to promote

substantive

equal opportunity, varied meanings,

404-06; accommodation of substantive

Deb, Dasaratha, 151 Denotified Tribes, 43 Depressed Classes, see Untouchability, M., 9, 19, 300, 301,

Desai, Morarji, 341 Deshmukh, Punjabrao, 45, 162, 165, 166, 369 De Vos, George, 562 Dharmalingam, A. M., 209, 213, 447 Dhavan, Rajeev, 479, 481, 485, 489, 490, 508, 539, 540

Dhavan, S. S., 479, 485 Dhebar Commission, 66

Dhobis, 137-42 India

education of SC and ST, 58-62; backward classes, 62, 64, 162, 177; post-

391-93;

De, Arunangshu, 172

Directive

Education: increase and deficiencies in

Equality:

Dave, P. K., 100 Davis, Kingsley, 259

305, 306, 325 Desai, K. K., 445

Edelman, Murray, 528

Ekbote, Gopal Rao, 446 Elayaperumal Committee, 53, 60, 70, 102 Eldersveld, Samuel J., 84, 560 Emergency Rule (1975-77), 139, 483, 520

Das, S. R., 429 Das, Seth Govind, 166 Daschowdhury, B. K., 317 Datar, B. N., 320, 325

Derrett, J. Duncan

* Dutta, Ratna, 51

177,267,

Dalton, Dennis, 18, 37 Damaska, Mirjan, 359

Scheduled Castes

Dumont, Louis, 8

Dusadhs, 141 Dushkin, Lelah, 18, 27, 32, 33, 43, 46, 53, 54, 63, 64, 66, 85, 128, 129, 130, 137, 149, 209, 222, 322, 325, 456, 527, 550

matriculation

Dahl, Robert, 85 Dalal, M. N., 375

615

Principles, see Constitution of

Dohars, 286 Dolbeare, Kenneth M., 529, 542 Double-Member Constituencies,

Legislative Bodies: Reserved seats

Dravida Munnetra Kazagham, 236 Driver, Edwin, 259

D’Souza, Jerome, 165

D'Souza, Victor S., 61

Dubey, H. P., 478, 479

see

equality and formal equality, 561-62

Ezhuvas, 16, 23, 195, 235-36

Fazl Ali, S. Murtaza, 382, 388, 419, 420, 424, 439 Feeley, Malcolm, 522

Fisher, Fred, 123

Fisher, Marguerite J., 70

Fox, Richard D., 11, 295, 349 Franchise (Southborough) Committee,

123, 125

Friedman, Lawrence M.. 359, 483, 533

Gadbois, George H., 479, 480, 481, 482,

484, 489, 490, Gaikwad, B. K., Gajendragadkar, 486, 488, 526 Galanter, Marc, 215, 305, 318, Gandhi, Indira, Gandhi, M. K.,

Gauba, K. L., 51

500, 503, 504 320 P. B., 232, 302, 381, 410,

18, 19, 20, 21, 52, 54, 103, 358, 359, 364, 395, 75, 477, 480, 481, 509 18, 29-40 passim, 375

Ghurye, G. S., 8, 13, 18, 21, 22, 23, 28, 111, 147, 153, 213 Gledhill, Alan, 496, 498, 499 Gopalan Nambiar, V. P., 218, 242, 248 Gough, E. Kathleen, 329

616

Index

Government employment, attraction of, 84-86 Government employment, reservation in: origins, 86; age, fee and other concessions, 87, 372, 389; measurement of effects, 93-117; pre-examination coaching centers, 99; promotions, 99— 101, 370-71, 446; public sector undertakings, 102-03 Government of India Act 1919, 27 Govinda Menon, P., 255

Grossman, Joel, 489 * Grossman, Kurt R., 554 Grover, A. N., 481

Gujarat Backward Classes Commission (1976), 258 Gupaa, A. C., 388

Gupta, M. G., 480, 482 Gurdon, P. R. T., 294

Gupta,

Kumar, 378

Gurung, Ari Bahadur, 159 Hajarnavis, R. M., 529 Halbas, 143

Hammond, Phillip, 529 Handler, Joel F., 533

Dis-

ram

of Fundamental

pluralities

dependent

“seats, 138

Rights,

on

34;

reserved

Indian society: compartmental and hierarchic organization, 7-17; empiri-

cal and formal approaches to group membership, 348-51; “realist” and “relativist” views of I. S., 352-58

Interviews, see Selection Committees

Irschick, Eugene F., 18, 27, 154, 343, 447 Isaacs, Harold R., 108, 110, 113, 319,

Jackson, A. M.R., 19

Hanumanthaiya, K., 186, 317, 341 Hardgrave, Robert L., 16, 23, 171, 262

Harijan Sevak Sangh, 34 Harijans, see Untouchability, Scheduled Castes

Harrison, Selig S., 175

Hart, Henry C., 47, 52 Hartog Committee (1928), 123, 156

Havanur Commission, 228, 239, 240, 245, 256-58, 260, 454 Havanur, L. G., 190, 192, 194, 206, 210,

211, 222, 226, 231, 232, 239, 256, 262, 268, 513 Hayden, Robert, 428 Hazari, 324

Hindu Mahasabha, 35

see Compensatory

crimination Policies: Administration Inden, Ronald, 8, 10 Indian Central (Nair) Committee (1929), 124, 156 Indian Franchise (Lothian) Committee, 124-25, 127, 129 Indian National Congress: excludes social reform questions, 24; adopts first antidisabilities resolution, 26; 1931 Prog-

324 Israel, preferences for oriental Jews, 553

Hah, Chong-Do, 562

Hegde, K. S. 199, 206, Heimsath, Charles H., Hidayatullah, M., 366, Hindu: population, 33; i

Income Ceilings, 62, 243, 263-66, 273

Income Tests,

380, 481, 513 18, 23 384, 486 definition of, 305-

Holderness, T. W., 123 Horowitz, Donald L., 395 Hutton, J. H. , 8, 9, 12, 28, 126, 127-28, 129, 130, 134, 143, 300

Jaganmohan Reddy P., 485 Jagannadham, V., 84

Jajmani System, 9 Jammu

and Kashmir Backward Classes

Committee (1969), 227

Janata Party, policy on OBC’s, 186-87 Jangan, R. T., 51 Jatavs, 137, 286

Jennings, Ivor, 38

Jha, A.N., 178 Joint Parliamentary Committee on the Welfare of Scheduled Castes and Scheduled Tribes, 71 Jones, Kenneth W., 25 Judiciary: Supreme Court, 477-78; high

courts,

477,

478-80;

appointment,

479-80; turnover, 481, 488; formalistic style, 483-85, 539; citation of authorities, 485-86, 493; dissenting and

concurring

hierarchy,

opinions,

490-92,

495-97,

489-90;

531-32;

writ jurisdiction, 489-99, 500, 517; tendencies to incoherence and diffuseness,

539-40;

pensatory 541-46

effectuation

of

com-

discrimination _ policies,

Index

Kalelkar, K. K., Chairman of B.C. C., 172, 260

540-41,

Legislative

Kayasthas, 223

Keer, Dhananjay, 18, 23. 30, 36, 319, 324, 326 Kelkar, G. S., 98, 113 Kennedy, Duncan, 483

Backward

Classes

Reservation

Commission 1971, 227, 237, 244, 245, 256, 273-74, 467, 468

Ketkar, Sridhar V., 25, 123 Khandekar, H. J., 160

Khanna, H. R., 423, 435, 481

Khare, R.S., 514 Khasis, 294 Khatiks, 141 Khobrogade, B. D., 321 Khosla, G. D., 479

Kirpal, Viney, 98, 113, 422 Chhatrapati), 30, 154

of

(Shahu

Konda Kapus, 140 Koppell, Oliver, 520

Krishna, Gopal, 47, 48, 51, 54 Krishna lyer, V. R., 147, 249, 302, 386, 387-88,

393, 419, 420, 424, 436-37,

468, 486, 520, 540

Kshatriyas, 10, 23, 295-302 Kuchelar, G., 340 Kumar, Dharma, 19 Kumara

Pillai

Commission,

236, 255, 263 Kunzru, Hirday Nath, 159

Bodies,

electorates,

375; extension thorization,

reserved _ seats: 25,

31-32,

45,

of constitutional au-

46.

375-76; — desig-

nation of con. tituencies, 46-; doublemember constituencies, 46, 456, 457—

58;

local

bodies,

occupants,

2;

50;

profile

performance

of of

occupants, 3 scope, 37576 Legislative Bodies, SC’s & ST’s in non-

Lingayats, 16, 225, 226, 231, 232 Litigation: to enforce or challenge caste prerogatives, 20; regarding reservations in government service, 104-05; claims for inclusion among beneficiaries of compensatory discrimination, 279-81; about group membership,

282-83;

capability

of litigants,

358-59; possible litigation for affirmative pursuit of substantive equality,

394-95;

government

success

in

Supreme Court, 482, 487; about com-

pensatory

discrimination,

486-87,

500-07, 509-13; interest group participation, 507-09; tactical prob-

lems in reservations cases, 519; remedies, 522-24; compliance and resistance to, 524-27; relation to government policy, 528-29, 530-31; relation

Krishnamachari, T. T., 160 Krishnan, G. Y., 285

Kumar, Santosh, 136

separate

516-17,

reserved seats, 49, 50-51

Kidder, Robert L., 500, 514, 515 Kochanek, Stanley A., 53 Kolhapur, Maharajah

Lecky, Robert S., 554 Legal Scholarship, 485, 514-15,

Katzenstein, Mary F., 427, 436

Kerala

Layard, Richard, 85 Legal Aid, 520-21

Kamble, N. M., 321, 323 Kapur, J. L., 303 Karve, Irawati, 12, 322, 349

617

197,

227,

to public opinion, 529; absence of litigation to implement compensatory discrimination policy, 396-98, 537-39

Lockard, Duane, 543

Kuppuswamy, B., 462

Lohia, Ram Manohar, 234 Lokur, B.N., 136

Lajpat Rai, Lala, 292

~ Luthera, Ved Prakash, 305

Lapp, Christopher C., 562 Latchanna, G., 72, 212

Lokur Committee, 53, 136-38, 152, 286

Lynch, Owen M., 23, 108, 111

Latin Catholics, 195

McCormack, William, 19 Latthe, A. B., 154 McDonald, Ramsay, 3! Lawyers: training, 514-15; orientation to MacDougall, John Arthur, 563 courts, 515-16; emphasis on litigation,

Madhavan, K. K., 323

516, 517-18, 538

Madras, Henry

515-16; low degree of specialization,

Madigas, 335-36

123

618

Index

Mahanubhava Panth, 290-92

Narayan, Jayaprakash, 176

Mahars, 14, 21, 29, 137, 290-92

Narayana, G., 51, 53, 71 Narula, R. S., 272 Natarajan, D., 58, 59

Majithia, Sunder Singh, 127

Majumdar, N. D., 152, 162, 168

Malas, 137 Malik, Suneila, 548, 549

Natarajan, K., 25 Natarajan, S., 18, 23, 25, 27 Nayakas, 287, 288

Mallahs, 219

Mandal, B. P., 187

Nehru, Jawaharlal, 159, 161, 162, 166, 167, 169, 173, 241, 379, 482

Mandelbaum, David G., 8, 9, 259, 349 Mariappa, T., 172

Marriage, as means of acquiring group membership, 138, 294-95, 340, 341 Marriot,

349

McKim,

8,

10,

11, 235,

302,

Mathew, K. K., 382, 385, 387, 423, 424,

486 Maurya, B. P., 138 Mayakrishnan, V., 195

Merillat, H.C. L., 494 Miller, Robert J., 291, 319 Misra, Lokanath, 323

Pant, G. B., 325

Mohapatra, M. K.. 51, 52, 53, 75 Moka Doras, 295-302 Montagu, Edwin, 154

Paraiyans, 14, 312, 329 Parsis, 223 Parvathamma, C., 51 Patankar, Bharat, 14, 29, 31, 32 Patars, 289

Mookerjee, S. P., 166 Moonje, B. S., 33 Moothan, Orby, 498

Patel, Tara, 61, 548 Patel, Vallabhbhai J., 170 Patil, M. S., 225, 232 Phadnis, Urmila, 321 Phule, Jyotirao, 23, 29

Morkhandikar, R. S., 40 Morrison, Charles, 500, 514

Mudholkar, J. R., 194, 402 Mukerji, P.,156 Mundas, 289, 294

Munshi, K. M., 160

Murphy, Arthur, 520

Murphy, Walter, 523

Muslims, 16, 25, 26, 51, 195, 198, 201, 202, 222, 231, 247, 306

Committee,

Nair, Kusum, 321

Nalavade, R. M., 159 Namasudras, 137

Naoroji, Dadabhai, 24

see

Compensatory Discrimination Policies: Administration O'Malley, L. S.S.,9, 12

Palekar, J. J., 445 Palen, Frank, 395

Mochis, 286, 288, 289

Gowda

Occupational Test of Backwardness,

Pal, G.S., 127

Misra, Sheokaran Nath, 158

232, 234, 262 Nagappa, Shri, 212

Noronha, Dionysius, 329

Other Backward Classes, see Backward Classes

Minto-Morley Reforms (1909), 25-26

Nagan

Nehru, S. S., 157 Non-Brahmin Movements, 27, 36, 156

Omvedt, Gail, 14, 18, 23, 29, 31, 32, 156 Oraon, Kartik, 317 Oraons, 293

Mayhew, Leon H., 357, 543 Mehta, S. C., 73, 175

Nadars, 16, 23, 171, 262, 304

165,

224,

230,

Pickett, J. Waskon, 324 Ponnarangam, V. A., 209 Poona Pact, 32-33 Prakash, Dharam, 159 Precedent, 490-95

Protection of Civil Rights Act, Untouchability (Offences) Act

se

Pyarelal, 18, 32

Race, 217, 364 Radhakrishnan, N., 190, 194, 195, 206, 485

Rai, Kalp Nath, 323 Rajagopala Ayyangar, N., 371

Index Rajah, M. C., 33, 85, 157 Rajbanshis, 137 Rajshekar Shetty, V. T., 74

619

Saxena, I.C., 491

Scheduled Areas, 147-51 passim Scheduled Castes: legislative

reserva-

tions, 44-55 passim; population, 42, 133; education, 55-64 passim; government employment, 84-105 passim; de-

Ram, Chandrika, 160 Ram, Jagiivan, 86

Rama Rap, T. S., 235, 447 Ramaswami, V., 429

signation of S. C., 122-31; inclusion of

Ramaswamy, G., 510 Ramaswamy, Uma, 27, 36 Ranade, M. G., 24 Rao, M.S. A., 179 Ratnan, K. J., 563 Ray, A. N., 145, 382, 384, 388, 419, 481,

Sikhs, 144, 305, 323; determination of membership in, 283-93, 305-41. See also Untouchability — S.C., Lists: 134-35; 1936 Order, 190; 1951 Order, 132; 1976 amendments, 135, 139; area limitations, 139-42; re-

Reddi, K. Brahmananda, 136, 323 Reforms Enquiry (Muddiman) mittee (1925), 27, 123, 156

Scheduled Tribes: legislative reservations, 44-55 passim; education, 55-64

486

Com-

Religion: Religious test for Scheduled Castes, 143-44; religious groups as

Backward Classes, 201-02. See also Conversion, Buddhists, Christians, Muslims, Sikhs

Reservations: layers of preference, 273— 74, 467-68; carry-forward of reservations from year to year, 403, 407-09, 410-11; for groups other than Back-

ward Classes, 425-31; above merit” vs. minimums,

ments,

439,

455-63;

“over and guaranteed compart-

463-67, 536; “carry-over” of

unfilled places to other groups, 474— 76. See also Government employment, Legislative Bodies, Education

Retzlaff, Ralph H., 363 Risley, Herbert, 122

Round Table Conference, 30-31, 33

Rowe, Peter, 514 Roy, Sarat Chandra, 293 Rudolph, Lloyd I., 2, 18, 19, 74, 321, 349 Rudolph, Susanne Hoeber, 2, 18, 19, 74, 321, 349 Saberwal, Satish, 52 Sachchidananda, 77

Sahas, 286-87

Sahib, Mhd. Ismail, 160 Saksena, R. N., 73 Santhanan, K., 75

Sapra, C. L., 60

Saraswathi, S., 36, 154, 159 Saraswati, Dayananda, 23

Sathi, Chedi Lal, 179, 206

ligious test, 143-44; judicial interpretation, 144-45, 283-90

passim; government employment, 84105 passim;

population,

148-50;

de-

148;

1950

termination of membership in, 28390, 293-305, 316-17

——S.

T.

Lists:

1936

list,

Order, 149; 1976 amendments, 139 Schermerhorn, R. A., 69 Schmitthener, Samuel, 514, 515

Schuchter, Arnold, 554 Schwartzberg, Joseph E., 11 Secularism, 305, 559, 561 Seervai, H. M., 479, 494

Selection Committees, 446-55 Separate electorates, see Legislative Bodies: reserved seats Setalvad, M. C., 497

Shah,J. C., 198, 199, 370 Shah, K. T., 146, 164, 166, 368 Shah, P. G., 172 Shah, Vimal P., 61, 548 Shankaraiya, Shri, 165 Shapiro, Martin, 359 Sharma, B. A. V., 73 Sharma, G. S., 305 Sharma, R. C., 60 Sharma, R. S., 13, 14 Sharp, Henry, 123 Shelat,J. M., 481

Shiva Rao, B., 322

Sikhs, 16, 35, 144, 305, 312, 323

Silverberg, James, 12, 234 Simon Commission

157

Singer, Milton, 349 Singh, Anup, 172

(1930), 30, 124, 125,

620

Index

Singh, Charan, 187

Singh, Deva Charan,

161, 162

Singh, Harjinder, 235 Singh, Hukum, 160, 305

Singh, K., 126 Singh, M., 130 Singh, Parmanand,-2 Singh, Saint Nihal, 26, 123 Singh, V. P., 178 Sinha, V. K., 305 Smith, Donald E., 305, 326 Smooha, Sammy, 553 Somnath Iyer, A. R., 372 Sonars, 292-93 Srinivas, M. N., 12, 16, 17, 18, 175, 295,

349

Srinivasan, R., 34, 85 Starte Committee (1930), 157

Subba Rao, K., 196, 198, 205-06, 247, 377, 408, 481, 488 Subbiah, A., 175 Sudras, 10, 13, 35, 145, 158, 234

Suffian, Tan Sri Mohamed, 563 Sunris, 286-87 Sunshine, Russell, 514

Suryabashis, 141

Tainbe, S. B., 125, 130 Tambe, Y. S., 445 Tamil Nadu Backward Classes Commission (1971), 234, 236-37, 460 Taub, Richard P., 77

Terminology,

Regarding Compensatory

Discrimination ward Classes,

Policies, 2-3; Back3, 154-59, 185-86;

caste, 7; untouchability, 13, 14, 25, 34;

community, 16-17 Thakkar, A. V., 161 Thomas, John M., 395

Tinker, Hugh, 26, 36

Titus, Murray T., 324 Tripathi, P. K., 180, 193, 204, 481 Tussman, Joseph, 486 Untouchability: general account, 13-16; terminology, 13, 14, 25, 34; emergence

as political issue, 25; evangelical and .Secular approaches to abolition, 28;

dispute about population of, 122-25,

129, 130, 132; dispute about meaning of, 125-130; among Muslims and

Christians, 143; definition of, 131-32, 134, 145-47

Untouchability 103, 44, 54 Untouchables,

(Offences)

see

Untouchability

Act,

Scheduled

1955,

Castes:

Vaishyas, 10

Varna, 10-11, 12, 23, 29, 35, 37, 145, 211, 234-35, 300-02, 324, 349-50

Veeraswami, K. N., 315

Venkata Ranga Iyengar, S. K., 428

Venkataraman, S., 166, 167 Venkataraman, S. R., 158 Venkatasubramanian, C. S., 493 Venkataswamy, Shri, 179 Voddars, 285 Vokkaligas, 225, 226, 231, 232 Von Furer-Haimendorf, Christoph, 151 Von Mehren, Arthur T., 484 Wagatsuma, Hiroshi, 562

Walsh, Cecil, 499

Wanchoo, K. N., 371, 410 Wasby, Steven, 522 Weitbrecht-Stanton, H. U., 25, 123 Wells, Richard, 489 Wiser, W. H., 9 Women, deemed backward class, reservations for, 217, 426 Wood, Glynn, 455 Woodhall, Maureen, 85 Wright, H. Elliott, 554

168;

Yadav, O. K., 210 Yadav, Shyam Lal, 323 Zelliot, Eleanor, 18, 21, 22, 25, 26, 27, 28,

29, 30, 31, 32, 33, 36, 39, 40, 108, 291,

319, 320, 326.

ZoBell, Karl M., 489

Index of Cases

Abdul Aziz v. State of Maharashtra, 445 Abdul Kadirv. Dharma, 301, 318

Abdul Latif v. State of Bihar, 414-17 Abhoy Pada Saha v. Sudhir Kumar Mondal, 286, 487 Abodha Kumar v. State of Orissa, 216, 445 Abraham v. Abraham, 291

A. R. V. Achar 0. State of Madras, 201, 222, 376 Agurchand v. Deochand, 492

All-India Station Masters v. General Manager, 370 Anandji Haridas & Co. v. Engineering Mazdoor Sangh, 494

Anjali v. State of West Bengal, 426 Annam Adinarayana v. State of Andhra Pradesh, 502 K. Appa Rao v. Director of Posts and Telegraphs, 140

C. M. Arumugam v. S. Rajgopal (11), 316, 329, 330, 334-35, 338-39, 501 Ashok v. Dean, 337, 511 Atmaram v. King-Emperor, 145

Baij Nath Prasad v. State of Bhopal, 493 Balaji v. State of Mysore, 178, 180, 191-97,

201-02, 204, 207, 220, 222, 226-29, 232-33, 235-36, 238, 241-42, 246-48, 250-51, 254-55, 258, 262, 267, 275-77, 314, 371, 377, 381, 389, 397, 400, 402-04, 410-16, 420, 32, 438-39, 441, 466-68, 470, 472, 475-76, 491, 495, 507, 510-11, 20, 524, 526, 528-32, 543

273, 393, 431474, 518—

Balaram v. State of Andhra Pradesh, 459 Balchand v. Laxminarain Mateh, 286 Basavalingappa v. Munichinnappa, 136, 285 Basunta Kumar v. Chief Electrical Engineer, 491 Baula v. Chief Executive Officer, 491

Bhagwan Bakhsh Singh v. Drigbijai Singh, 306 Bhagwan Koer v. Bose, 306

Bhaiya Lal v. Harikishan Singh, 140, 285, 288 Bhaiya Ram Munda v. Anirudh Patar, 283, 289 Bhau Ram v. Baij Nath, 215 Bhavan v. Commercial Tax Officer, 491, 495

Bhopalsingh v. State, 215, 314, 376

Bidi Supply Co. v. Union of India, 217, 314 Biman Chandra v. Governor, 313

Bisanchand Lalchand v. Union of India, 491 Bombay Education Society v. State of Bombay, 42 Brahmachari Research Institute v. Their Workmen, 497

Budhan Chaudhry v. State of Bihar, 217, 314 Cannabasaviah v. State of Mysore, 445

Chait Ram v. Sikander, 197, 200, 219, 255,

373

Chamaraja v. State of Mysore, 456, 473

Champakam Dorairajan v. State of Madras [see State of Madras v. Champakam Dorairajan} Chanchala v, State of Mysore, 384, 393, 402, 426-27, 433, 435, 524

Chandhlal v. Babulal, 492

Chandra Sekhera v, State of Mysore, 372

Chandrasekhara Mudaliar v. Kulandaivelu Mudaliar, 306, 311 Chatturbhuj Vithaldas Jasani v. Moreshwar Parashram, 290, 292-93, 295-300, 304, 315-17, 327, 333, 348, 484 Chauhan v. State of Gujarat, 408, 420, 445

Chhotey Lal v. State of Uttar Pradesh, 184, 432, 467 Chinnaswammy v. Anthonyswamy, 306 Chitra Ghosh v. Union of India, 428, 429, 434

622

Index of Cases

Chitralekha v. State of Mysore, \92, 194-98,

204, 485, Chunku Daryao

207, 233, 247, 264, 314, 402, 448, 488, 491, 513, 530 Manjhi v. Bhabami Mayhan, 296, 306 v. State of Uttar Pradesh, 479

Dasa Rayudu v. Andhra Pradesh Public Service

Commission, 196, 242, 397, 446, 511 Dass Bank v. Kali Kumari, 491 Dattatraya v. State of Bombay, 217, 369, 376, 416 Daya Krishnan v. Assessing Officer, 492 Daya Ram v. State of Haryana, 384, 403, 427, 451 Devadasan v. Union of India, 93, 217, 267,

371, 372-73, 377, 379, 390, 398, 403-— 05, 408, 410-21, 431, 442, 445, 476, 488, 505, 537

Devarajiah v. Padmanna, 134, 146, 309

Guntur Medical College v. Mohan Rao, 335, 341, 503, 511, 519, 523

Gurinder Pal Singh v. State of Punjab,

183,

265, 277, 432

Gurmukh Singh v. Union of India, 310, 31213, 319

Gurunath v. Kamalabai, 491

Gurusami Nadar v. Irulappa Konar, 319, 328 Hadibandhu v. Banamali, 146, 333

Hariharan Pillai v. State of Kerala, 196, 202,

218, 242, 248, 255-56, 261, 269, 378,

402, 416, 459, 464, 476, 523

Harpartap Singh v. Union of India, 244, 440, 456 Horo v. Jahan Ara, 294—95, 335, 340, 34849 Hridaya Narain Singh v. Mhd. Sharif, 197,

219, 241

Devchand Totaram v. Ghaneshyam, 324 Dhanki Mahajan v. Rana Chandubha, 492 Didar Singh v. Sohan Singh, 289

1. - T. Officer Shillong v. Rymbai, 279

458 Dilip Kumar v. Government of Uttar Pradesh, 228, 243, 258, 277, 421, 459, 492, 523 Dippala Suri Dora v. V. V. Giri [See V. V. Giri v. Dippala Suri Dora) Director General of Posts and Telegraphs v. Natarajan. 409, 505 Dulichand v. Union of India, 376

Indian Hume Pipe Co., Lid. v. Their Workmen,

Digambar Rao Bindu v. Dev Rao Kamble, 406,

Durgaprasada Rao v. Sundarsanaswami, 319,

328

Dwarka Prasad v. State of Uttar Pradesh, 496

Ellappa v. Ellappa, 329

Emperor v. Dholraram, 491

Gadipalli Parayya v. Boyina Rajayya, 294,

298 Ganpat v. Presiding Officer, 308, 325 Garg v. State of Punjab, 207

General Manager v. Rangachari, 100, 121, 217, 369-372, 409-412, 421, 487-88, 507, 531 V. V. Girt v. Dippala Suri Dora, 47, 295, 298-99, 302, 348, 406, 456, 458, 488, 506, 527

Gokul Prasad v. Sohani, 217

Golak Nath v. State of Punjab, 367, 384 Gopal v. Hanmant, 300 Gudigar v. State of Mysore, 452

Gullapalli Nageswara Rao v. Principal, 426

Gupta v. Union of India, 421-22

Income Tax Commissioner v. Vazir Sultan, 493 Income Tax Officer v. Nadar, 491

Inder Singh v. Sadhan Singh, 301, 318 497

Jacob Mathew v. State of Kerala [See State of Kerala v. Jacob Mathew) Jagan Nath v. State of Jammu and Kashmir [See State of J & K v. Jagar Nath] Jagdish Rai v. State of Haryana, 392-93, 435 Jagwant Kaur v. State of Bombay, 367, 442 Jai Kaur v, Sher Singh, 492 Jaisri v. Rajdewan, 492 Jaiwant Rao v, State of Rajasthan, 494 Jalali v. Principal, 525 Janardhan Surraraya v. State of Mysore, 475 Janardhana Rao v. Deputy Transport Commissioner, 494

Jankilal v. Jabarsingh, 304

Janki Prasad Parimoo v. State of Jammu and Kashmir,

92,

183,

191, 200-01,

207,

217, 228, 238-39, 243, 248-49, 269, 274, 277, 377, 422-23, 438, 452, 453, 526.

Januma Rai v. Chandradip Rai, 492

Jayantilal v. Rana, 492

“Jayasree v. State of Kerala, 92, 183, 204, 238, 243, 249, 264, 280 Jayaswal v. Principal, 275, 342

Jiwan Khan v. Habib, 306 Joseph Thomas v. State of Kerala, 216 Joshi v. Madhya Bharat, 216 Jyoti Bhushan v. Bodh Ram Muritram, 141

‘Index of Cases Kayjari Saha v. State of West Bengal, 283, 287,

623

Kalinath v. Nagendra Nath, 495

Mulai v. Lal Dan Bahadur Singh, 302 Municipal Committee v. Hazra Singh, 491 Muralidhar v. State of Andhra Pradesh, 216

Kartik Oraon v. David Munzni, 293, 295, 298, 317, 348 Karwadi v. Shambharkar, 307

300, 319, 349 Nain Sukh Das v. State of U.P., 45, 215, 305, 375-76

348, 499

Karkare v. Shevde, 313

Kathi Raning Rawat v. State of Saurashtra, 218

Kelkar v. Chief Controller of Imports and

Exports, 408

In re Kerela Education Bill, 1957, 42, 370,

478, 538

Kesava Iyengar v. State of Mysore, 201, 245,

249, 369, 399-400, 438

Keshavananda v. State of Kerala, 481,538

Kishan Chand v. Ram Babu, 491 Kishori v. Board of Revenue, 216

427, 445

Chancellor,

384,

Laila Chacko v. State, 195, 263 . Lalita Shuri Tikku v. State of Jammu and Kashmir, 252, 426, 444-45, 525

Lalwani v. Collector, 141 Laxman

Siddappa

Naik

v.

Kattimam

Chandappa_Jampanna, 287, 329, 487

Mahadeolal v. Administrator-General, 491

Maharajah of Kolhapur v. Sundaram Iyer, 296

Mahendra Nath Pathak v. State of Assam, 217, 282, 291, 364, 373, 446 Makhan Lal Waza v. State of Jammu and Kashmir, 423, 453, 525

Maktul v. Manbhari, 492 Mangilal v. Mittilal-Radheylal Rastogi, 445

Meera Bai v. Director of Medical Education, .

Michaelv. Venkataswaran, 306, 309, 312-13, 318

Khan

Public

Pardhasaradhi Rao v. Srinavasa, 495 Pars Ram v. Shiv Chand, 288

S.A. Partha v. State of Mysore, 191, 218, 220,

222, 250, 279-80, 406, 456-57, 465,

471, 474-75 Pasha v. State of Mysore, 345

Periakaruppan v. State of Tamil Nadu, 199, 259, 380, 426, 438, 451, 460, 484, 523 Periakaruppan Chettiar v. Stateof Tamil Nadu, 219, 452, 454

Phulchand v. Hukumchand, 493 Plessy v. Ferguson, 494

Pradip Tandon v. State of Uttar Pradesh, 180, 200, 204, 278, 374, 421, 435, 492

Pramanick v. Union of India, 446 Pritam Kaur v. State of Pepsu, 218

144, 307, 310-11,

Puppala Sudarsan v. State of Andhra Pradesh, 456

Raghuramulu v. State of Andhra Pradesh, 406, 455, 458, 464-65, 511 Raj Narain v. Indira Nehru Gandhi, 477

S. Rajagopal v. C. M. Arumugam [1}, 315—

16, 318, 327-28, 331

Michael Pillai v. Barthe, 318 Mohd. Abdul Mabood Maharashtra, 445

467,475 Panduram v. Biswambar, 333 Pandurangarau v. Andhra Pradesh

315, 320, 327, 487, 513

Mastanaiah v. Delimitation Commissioner, 48

Mira Devi v. Aman Kumari, 306

Om Prakash v. State of Punjab, 366, 494 S. G. Pandit v. State of Maharashtra, 439,

Punjabrao v. Meshram,

Manickam v. Poongavanammal, 145 Marthamma v. Munuswami, 328

315, 522-23

524 Narasimha Reddy v. Bhupathi, 327, 500 Narayana Swami v. State of Mysore, 431 Nataraja v. Selection Committee, 306, 339, 343

Parameswaran Moothathu v. Vasudeva Kurup, 146

Kolandei v. Gnanavarum, 326 v. Pro-Vice

145,

Nanda Kishore Sharma v. State of Bihar, 196,

Service Commission, 218

Konkani v. Shankar Rao ,285 Kulkarni v. State of Mysore, 372 Kushma Joshi

Muthusami Mudaliar v. Masilamani,

v. State

of

Mohd. Hanif Qureshi v. Stateof Bihar, 218, 366 Moosa v: State of Kerala, 373, 442 Moti Das v. S. P. Sani, 215, 218, 310

S. Rajagopal v. C. M. Armiugam [II] [See C. M. Arumugam v. S. Rajgopal] Rajasthan State Electricity Board v. Mohan

Lal, 370 C. A. Rajendran v. Union of India, 377, 379, 396-98, 408, 508, 511, 537

624

Index of Cases

P. Rajendran v. State of Madras,

197-98,

216, 233, 252, 262, 267, 426, 477, 488,

519, 523 Rayt v. Deputy Tahsildar, 493 Raju v. Chief Electoral Officer, 373, 384 Rayu v. State of Gujarat, 420 Ram Kishore v. Union of India, 491 Ram Narain v. Director of Consolidation, 491, 494 Rama Suddarayalu v. Rengammal 492 Ramakrishna Singh v. State of Mysore, 175, 180-81,

190, 201, 222, 241, 246, 249,

314, 400, 438, 464, 472, 475

Ramalingam v. Boddu Abraham, 484 Ramaswami v. Chandra Kotayya, 491

Ramayya v. Josephine Elizabeth, 328

Ramchandra Mahton v. State of Bihar, 416 Ramchandra Vishnu v. State of Madhya Pradesh, 426-27

Ramesh Chander Garg v. State of Punjab, 265, 431-32, 459, 474, 476, 519 Rangachari v. General Manager [See General Manager v. Rangachari} Ranjit Kumar v. State of West Bengal, 493 Ratansi D. Morarji v. Admr. General of Madras, 300. 306

Rattan Singh v. Devinder Singh, 307, 310 Reddi v. Savitramma, 494

Registrar v. Shashi Pal Singh, 428

Resham Bibi v. Khuda Bakhsh, 306 Rex v, Ram Dayal, 491

Rhagava Dass v. Sarju Bayamma, 292 Rura Ram v. Gurbachna, 207

Sagar v. State of Andhra Pradesh [See State of Andhra Pradesh v. Sagar)

Sahdeonarain v. Kusumkumari, 296

Saifuddin Saheb v. State of Bombay, 146, 334 Sangannagonda v. Kallangonda, 301 Sanghar Umar v. State, 215, 314

Sankalinga Nadan Dorai, 145

v.

Raja

Rajeshwari

Sankaram Namboodri v. Madhavan, 303

Sarda v. Mizo District Council, 487

Sardool Singh v. Principal, Medical College, 277, 384, 427, 451, 499

Sastri Yagnapurushadasi v. Muldas Bhundardas Vaishya, 146, 309-10, 327 G. Satyanandhan and others v. State of Andhra Pradesh, 446 Shamdasani v. Central Bank, 500 Shameem v. Medical College, 263, 275, 344

Shanshikalabai v. Election Officer, 457

Shantha Kumar v. State of Mysore, 339, 347,

484 Sharda v. Visveswariah Engineering College, 444

Sheokaransingh v. Daulatram, 215

Sherbert v. Verner, 325 Shetty v. State of Mysore, 200, 268, 342, 34547

Shimoga Durga Bovi v. Returning Officer, 406 Shri Prithvi Cotton Mills v. Broach Municipality, 492 Shyam Behari Tewari v. Union of India, 369 Shyamsundar v. Shankar 295, 298, 316, 348

Deo

Udgir,

292,

Sidhrajbhai Sabbaj v. State of Gujarat, 370

Sri Venkataramana Devaru v. State of Mysore,

146, 334

Sridhara v. Director of Technical Education,

269, 345, 346 Srinivasa Atyar v. Saraswati Ammal, 215

State of Andhra Pradesh v. Balaram, 184, 200,

227, 238, 248, 254, 377, 426, 438, 460, 466, 487, 523 State of Andhra Pradesh v. Sagar, 196, 198—

201, 211, 216, 218-20, 233, 248, 250,

252-53, 377, 389, 418, 426-28, 431, 485 State of Bihar v. Abdul Majid, 491 State of Bombay v. Narasu Appa, 215, 301 State of Bombay v: Yagna Sastri Purushadasyi [See Sastri Yagnapurushadasyi v. Muldas Bhundardas Vaishya) State of Jammu and Kashmir v. Jagar Nath, 201, 222, 245, 305, 494, 524, 526 State of Kerala v. Jacob Mathew, 180, 195, 202, 216, 235-36, 254, 258, 426-27, 440, 460, 464, 476, 507

State of Kerala v. Krishna Kumari, 264-65 State of Kerala v. N. M. Thomas, 145, 147, 220, 244, 249, 302, 361, 379, 382-84, 389-95, 418-20, 423-25, 435-37, 439, 468, 484-85, 487, 494, 508, 541

State of Madhya Pradesh v. Puranchand, 309 State of Madras v. Champakam Dorairajan,

164-65, 217, 314, 365-67, 369, 406,

State of Madras v. V. G. Row, 221

State of Mysore v. Narasinga Rao, 217 State of Mysore v. Bidap, 494

State of Punjab v. Hira Lal, 92, 220, 425, 440,

487 State of Rajasthan v. Pratap Singh, 215, 305

Index of Cases State of Uttar Pradesh v. Deoman, 494

State of Uttar Pradesh v. Pradip Tandon, 201,

625,

B.N. Tewariv. Union of India, 403, 476, 523 In re Thomas, 312, 319, 538

State of Uttar Pradesh v. Ram Chandra, 491 Subhash Chandra v. State of Uttar Pradesh,

Thomas v. State of Kerala (See State of Kerala v. N. M. Thomas] Trilochan Choudhury v. Dayanidhi Patra, 491

Subhash Mohan Jalali v. Principal, Medical College, 101, 450

Kashmir, 181, 198, 202, 211, 220, 222, 242, 244, 245, 252, 261, 379, 399, 423,

216-18, 228, 239, 278, 421, 435, 487 277-78, 492

Subhashini v. State of Mysore, 426-28, 431—

32, 450 Subrao v. Radha, 300-301 Sudama Prasad v. Divisional Superintendent, 445, 505 Sudha v. Selection Committee, 345

Sudharkar v. State of Orissa, 425 Sudhir Kumar Ghose v. Ank Kameya, 121 Sukhdev v. Government of Andhra Pradesh,

196, 218, 248, 250, 372, 421, 426-27, 474, 494 Sukhnandan Thakur v. State of Bihar, 429,

430, 435

Sukhvinder Kaur v. State of Himachal Pradesh, 183, 428, 523 Sundarambal v. Suppiah, 325 Sunder Devi v. Jheboo Lal, 300

Superintendent v. Corporation of Calcutta, 491 Superintendent v. Ram Ajodhva, 491

Surendrakumar v. State of Rajasthan, 427, 523

P. Susila v. State of Madras, 282, 304, 444,

sll Swain v. Secretary, 219

Triloki Nath Tiku v. State of Jammu and 444, 453, 506, 524-26

Umesh Chandra Sinha v. V. N. Singh, 197, 427 University of Madrasv. Shantha Bat, 369, 426 Upoma Kachain v. Bholaram, 300 Urmila Ginda v. Union of India, 340-41, 484 Vinod Sagar Sood v. State of Madhya Pradesh,

217 Varadiah v. Parthasarathy, 333 Venkataramana v. State of Madras, 165, 167, 190, 201, 217, 220, 245, 367, 369, 399

400, 406, 474, 523

Vinayak Shamrao v. Moreshwar Ganesh, 491 Vinod Sagar Sood v. State of Madhya Pradesh, 427 Vishnu v. State of Madhya Pradesh, 217 D. G. Vishwanath v. Government of Mysore, 192, 207, 233, 258, 271, 276, 346, 380,

402, 411, 416, 440, 448-49, 451, 454, 502, 532

Wilson Readev. C. S. Wasudeo v. Union of Yusuf Abdul Aziz v. Yusuf Beg v. Maliq,

Booth, 294-95, 298, 348 India, 141, 143 State of Bombay, 217 318