Affirmative Action in India: Reservation, Judgments and Debates

Official Syllabus: Equality — Social, political and economic; relationship between equality and freedom; Affirmative action.

Affirmative action is the point at which the theory of equality becomes policy, and therefore the point at which it becomes contested. Every argument about formal versus substantive equality, about opportunity versus outcome, and about merit versus circumstance, arrives here and has to produce an answer that can be written into a statute.

• The policy occupies the ground between equality of opportunity and equality of outcome. Its object is a society that will not be equal but will be more equal and less unequal, while every individual retains the opportunity to progress.
• It is the most institutionally developed anywhere in the world in India, where it is constitutionally mandated rather than merely permitted, and where it reaches legislatures, public employment and education simultaneously.
• The debate about it is genuinely unresolved. It draws as much strong criticism as strong support, and both sides argue from the value of equality rather than against it.

Introduction

PSIR Optional Notes {PDF}

Complete Notes Package (Paper I & II) | 1500+ Pages | Instant Download

Discounted Price | Buy Now @ ₹1000 →

A. What Affirmative Action Means

• Affirmative action denotes positive measures taken to improve the position of groups that have suffered systematic disadvantage, going beyond a prohibition on discrimination to a requirement of preference.
• Several terms are used interchangeably, and their connotations differ.

• The choice of label is itself an argument. “Reverse discrimination” presumes the measure is a wrong requiring justification; “protective discrimination” presumes it is a duty already justified.
• Its distinguishing feature is that it operates on groups rather than individuals. Benefits are allocated by membership of a category, which is what makes it powerful as a remedy and vulnerable as a principle.

B. Where It Sits in the Theory of Equality

• Formal equality prohibits discrimination and stops there. Affirmative action rests on the proposition that treating unequals equally perpetuates inequality, so identical treatment must sometimes be set aside.
• In Rawlsian terms it is fair equality of opportunity supplemented by the difference principle: positions must be genuinely open, and where they are not, arrangements must work to the benefit of the least advantaged.
• In Aristotelian terms it is the second half of his formula finally taken seriously — it is unjust to treat unequals equally, and proportionate treatment therefore requires differentiation.
• In Young’s terms it is differentiated citizenship: the recognition that a universal standard applied to a differentiated population is not neutral.

The Case For Affirmative Action

A. Compensatory Justice

• The most widely stated justification is that the policy compensates a group for past injustices and past suffering. Present disadvantage is the accumulated residue of historical wrong, and the wrong requires redress.
• Its strength is moral clarity: where a group was excluded by law for centuries, the burden of proof falls on those who would leave the consequences in place.
• Its weakness is the transmission problem. The individuals compensated are not those who suffered, and those bearing the cost are not those who inflicted the wrong.

B. Distributive Justice

• A second and stronger justification is forward-looking rather than backward-looking: present inequality of starting points makes formal equality of opportunity a fiction, so the remedy corrects a current unfairness rather than a historical one.
• This avoids the transmission problem entirely. The claim is not that a person’s ancestors were wronged but that the person is disadvantaged now, by a structure still operating.
• It also explains why the policy should be assessed by whether disadvantage is actually reducing, rather than by whether the historical debt has been discharged.

C. Diversity and Representation

• Institutions function better when they reflect the society they serve. A judiciary, a bureaucracy or a university drawn from a narrow social base makes worse decisions because it lacks relevant knowledge and lacks legitimacy with those it governs.
• Representation also has a democratic dimension. A state whose institutions visibly exclude a fifth of its population is not fully a democracy for that fifth, whatever its franchise rules.
• The role-model argument adds that visible success alters expectations within the disadvantaged group and reduces the stereotype within the advantaged one.

D. The Critique of Merit

• The case against merit as a sole criterion is that merit is substantially produced by opportunity. Measured attainment records the schooling, nutrition, language exposure and family capital a person happened to receive.
• Rawls’s formulation carries the argument: people no more deserve their native abilities, including their propensity for hard work, than they deserve the advantages of family and social background.
• The practical version, developed in India, is that assessment should be against a disadvantage index — measuring achievement relative to circumstances rather than absolutely, so that a given score from an under-resourced school is read as the higher accomplishment it represents.
• Affirmative action on this reading does not compromise merit; it corrects the measurement of it.

The Case Against Affirmative Action

The objections deserve to be stated at full strength, since the honest position is that they are not fully answerable.

A. The Individual-Justice Objection

• The central moral difficulty is that an individual is made to pay for wrongs they did not commit. A candidate from an advantaged group is disadvantaged not for anything they did, but because they are seen as belonging to a category associated with past injustice.
• This offends the principle that persons are to be treated as individuals rather than as instances of a group — the same principle on which the case against the original discrimination rested.
• The objection is strongest where the burdened individual is themselves poor, and weakest where they are the direct beneficiary of the structure being corrected.

B. The Stigma Objection

• Preferential treatment can undermine respect for talented individuals from the disadvantaged group, since their achievements are attributed to the preference rather than to their capacity.
• It can also unleash resentment among those who believe they were passed over, which corrodes precisely the fraternity the policy was meant to build.
• The internal version is the more serious: beneficiaries may internalise the doubt, and the policy meant to affirm equal worth can end by qualifying it.

C. The Perpetuation Objection

• The most consequential practical criticism is that the policy tends to become a self-perpetuating mechanism. Material equality may be achieved to some degree while caste or ethnic consciousness grows stronger, because the category has been given institutional value.
• A remedy intended to dissolve a social division can end up entrenching and politicising it, since constituencies form around the benefit and no group willingly exits the list.
• The corollary is that affirmative action must be conceived as a temporary measure to create a level playing field, and that its costs and moral difficulties must be weighed before it is adopted rather than after.

D. The Efficiency and Elite-Capture Objections

• The efficiency objection holds that selecting other than by measured attainment reduces institutional performance. The reply is that measured attainment is a poor predictor once circumstances are controlled for, but the objection retains force in highly specialised fields.
• Elite capture is the empirically best-supported criticism: benefits concentrate among the already advantaged within the beneficiary group, so the policy redistributes within the disadvantaged category rather than across the social divide.
• This objection is not fatal but corrective — it is an argument for creamy-layer exclusion and sub-classification, not for abolition, and Indian law has moved in exactly that direction.

Affirmative Action in the Indian Constitution

India’s scheme is distinctive in three ways: it is written into the Constitution rather than adopted by policy, it is mandated rather than merely permitted, and it operates across legislatures, employment and education at once.

A. The Constitutional Provisions

• The structural point is that Articles 15(4) and 16(4) are not exceptions to the equality clause. The Court has held they are facets of the equality guaranteed by Articles 14, 15(1) and 16(1) — the means by which formal equality is made substantive, not a departure from it.
• Article 46, a Directive Principle, supplies the purpose: the state shall promote with special care the educational and economic interests of the weaker sections, particularly SCs and STs.

B. Ambedkar and the Origins

• The scheme reflects Ambedkar’s conception of justice more than any liberal theory. His premise was that Indian society was organised as graded inequality, in which equal treatment would confirm the existing order rather than dislodge it.
• He regarded political safeguards as necessary but insufficient: representation without social transformation would leave the structure intact, which is why Annihilation of Caste rather than reservation is the centre of his programme.
• His Constituent Assembly warning frames the whole policy — India was entering a life of political equality alongside social and economic inequality, and unless that contradiction was removed, those who suffered from it would blow up the structure of political democracy.
• The original design was explicitly temporary for political reservation — ten years for reserved legislative seats, extended repeatedly since — while reservation in employment and education carried no sunset clause.

The Backward Classes Commissions

Identifying who qualifies has been as contested as deciding what they receive, and the machinery for doing it has a distinct history.

A. The Kalelkar Commission

• The first Backward Classes Commission was appointed in 1953 under Kaka Kalelkar, pursuant to Article 340.
• It identified over 2,000 backward castes and recommended caste-based criteria, reservation in services and large-scale educational provision.
• Its report was effectively shelved. The Chairman himself appended a note expressing misgivings about caste as the basis of classification, and the government declined to act on it — a rare instance of a commission undermined by its own head.

B. The Mandal Commission

• The second Backward Classes Commission was appointed in 1979 under B.P. Mandal and reported in 1980.
• It used eleven indicators of backwardness grouped as social, educational and economic, with social indicators weighted most heavily — an attempt to avoid resting the classification on caste alone while still using it.
• It identified roughly 3,743 castes as socially and educationally backward and estimated the OBC population at about 52% of the total.
• It recommended 27% reservation in central government services and public undertakings — a figure set not by the population estimate but by the 50% ceiling, since SC and ST reservation already accounted for 22.5%.
• The recommendation lay unimplemented for a decade until announced in 1990, provoking widespread agitation and the litigation that became Indra Sawhney.
• Its lasting significance is that it converted a category into a constituency. OBC identity as a political force in northern India dates substantially from this period.

C. The Rohini Commission

• A commission under Justice G. Rohini was constituted in 2017 to examine the sub-categorisation of Other Backward Classes, on the concern that benefits were concentrated among a few dominant OBC communities.
• It submitted its report to the President in 2023, and its contents have not been made public.
• The question it addresses is the same one Davinder Singh later answered for the Scheduled Castes — whether a reserved category may be graded internally so that the most disadvantaged within it actually reach the benefit.

D. The Caste Census

• Reservation policy has operated for decades without reliable data on the size of its principal beneficiary group. Caste enumeration was discontinued after 1931 for all groups other than SCs and STs, so the Mandal Commission’s 52% estimate rested on extrapolation from ninety-year-old figures.
• Bihar conducted a caste-based survey in 2023, the first substantial state-level enumeration in independent India, finding backward classes to constitute close to two-thirds of the state’s population and prompting an expansion of state reservation.
• The Union Government announced on 30 April 2025 that caste enumeration would be included in the forthcoming census — a decision with consequences that run in both directions.
  • It supplies the empirical basis that judgments from Balaji to Davinder Singh have repeatedly demanded and that states have rarely been able to produce.
  • It also makes the 50% ceiling harder to sustain politically if enumerated populations substantially exceed their quotas, which is the argument its critics anticipate.

The Judicial Evolution

Indian reservation law has been built almost entirely by the Supreme Court, and the sequence is the substance of the topic.

A. The Early Period

• Champakam Dorairajan (1951) struck down communal reservation in education, holding that the Directive Principles could not override Fundamental Rights. Parliament responded with the First Amendment, inserting Article 15(4) — the first of many legislative reversals of judicial limits.
• Balaji (1963) introduced the 50% ceiling on reservation and held that caste could not be the sole criterion of backwardness. It also gestured at the idea later formalised as the creamy layer.
• Devadasan (1964) struck down the carry-forward rule where it breached the ceiling, later addressed by Article 16(4B).

B. Indra Sawhney: The Governing Framework

The Mandal litigation produced the judgment that still governs, and its holdings should be known individually.

• The Court upheld 27% reservation for Other Backward Classes in central government employment, sustaining the Mandal Commission’s recommendation.
• It struck down the additional 10% reservation for economically backward sections among forward castes, holding that economic criteria alone could not define a backward class.
• It confirmed the 50% ceiling as a general rule, breachable only in extraordinary circumstances and for good reason.
• It required the exclusion of the creamy layer from OBC benefits — those among the backward classes who have advanced sufficiently that they no longer need the protection.
• It held that reservation under Article 16(4) applies to appointments, not promotions — a holding Parliament reversed by inserting Article 16(4A).
• It clarified that caste can be the starting point for identifying a backward class, provided backwardness itself is established, since caste in India is a social class.

C. After Mandal

• Nagaraj (2006) upheld Article 16(4A) but conditioned promotion reservation on the state demonstrating backwardness, inadequacy of representation and administrative efficiency, with quantifiable data.
• Jarnail Singh (2018) removed the requirement to prove the backwardness of SCs and STs — since their backwardness is presumed by their listing — while extending the creamy-layer principle to them for the purposes of promotion.
• Janhit Abhiyan (2022) upheld the 103rd Amendment and its 10% reservation for economically weaker sections, by 3:2.
  • The majority held three things. Economic criteria alone are a permissible basis for reservation.
    • The EWS quota breaching the 50% ceiling does not violate the basic structure, since the ceiling was laid down for Articles 15(4) and 16(4) and EWS operates under separate clauses.
    • Excluding SCs, STs and OBCs from EWS is not impermissible, since those groups are separately provided for.
  • The dissent held that the exclusion of the poor among SCs, STs and OBCs was discriminatory and constitutionally impermissible, and that permitting the breach of the ceiling would open the door to its erosion.
  • The judgment is a genuine inflection point: it shifts the constitutional basis of affirmative action from social exclusion toward economic disadvantage, which is a change in what the policy is understood to remedy.

D. Davinder Singh: Sub-Classification

The most significant recent development, and the one most reference material predates.

• In State of Punjab v. Davinder Singh, decided on 1 August 2024, a seven-judge Constitution Bench held by 6:1 that sub-classification within the Scheduled Castes is constitutionally permissible.
• States may apportion the reserved quota among sub-groups to give greater benefit to the most disadvantaged among them, provided the sub-classification rests on empirical data demonstrating relative backwardness and not on political preference.
• The judgment overruled E.V. Chinnaiah (2004), which had held the Scheduled Castes to be a homogeneous class incapable of subdivision because the Presidential list under Article 341 created a single group.
• Several judges in the majority also endorsed applying the creamy-layer principle to Scheduled Castes and Scheduled Tribes, so that benefits reach those who most need them.
• Justice Bela Trivedi dissented, holding that the Constitution intended the Scheduled Castes to remain a single homogeneous class, and that state sub-classification amounts to tampering with the Presidential list, which only Parliament may alter.
• Its significance is conceptual as much as legal. It converts the beneficiary category from a bloc into a gradient, and accepts that disadvantage varies within a disadvantaged group — which is the elite-capture objection absorbed into constitutional doctrine rather than resisted.

The Creamy Layer

• The creamy layer is the socially and economically advanced section within a backward class, excluded from the benefit so that it reaches those who actually need it.
• Its rationale is precise: without exclusion, the benefit is captured by those who have already escaped the disadvantage, and the policy redistributes within the group rather than across the social divide.
• The doctrine has expanded steadily — introduced for OBCs in Indra Sawhney, extended to SC and ST promotions in Jarnail Singh, and endorsed more broadly for SCs and STs in Davinder Singh.
• The objection to extending it to SCs and STs is serious and constitutional rather than merely political. SC and ST reservation remedies social exclusion and untouchability, which income does not cure — a wealthy Dalit may still face exclusion in marriage, housing and social interaction.
  • On that reasoning, applying an income filter to a social remedy mistakes what the remedy is for.
• The counter-argument is the empirical one: without exclusion, benefits concentrate in a narrow stratum across generations, and the most disadvantaged within the category never reach the point of competing for them.
• This is the sharpest live disagreement in Indian equality law, and it turns on whether the disadvantage being remedied is fundamentally social or material.

Reservation Beyond Caste

A. Women’s Reservation

• Local government came first and remains the largest such intervention anywhere. The 73rd and 74th Amendments reserve not less than one-third of seats and chairperson posts for women in panchayats and municipalities, since raised beyond a third by several states.
  • The result is over a million women in elected office, and a documented shift in the composition of local public spending.
  • The counter-phenomenon is the sarpanch pati — the elected woman’s husband exercising the office in practice — which demonstrates that formal access is not substantive power and that patriarchy adapts to reform rather than yielding to it.
• Legislative reservation took far longer. The Constitution (106th Amendment) Act, 2023 — the Nari Shakti Vandan Adhiniyam — reserves one-third of seats in the Lok Sabha and state assemblies for women.
  • Its commencement was made conditional on a census and a subsequent delimitation, which deferred the benefit indefinitely.
  • An attempt to break the deadlock failed. The Constitution (131st Amendment) Bill, which would have enabled delimitation on 2011 census data so the quota could operate sooner, secured 298 votes on 17 April 2026 — 54 short of the two-thirds required.
  • The reason for its defeat is analytically important: opposition came not from resistance to women’s representation but from southern states’ objection to seat redistribution, and from a demand that reservation be delinked from delimitation and applied to the existing 543 seats.
  • The episode shows a gender entitlement held hostage to a federal dispute it has nothing to do with — a reminder that affirmative action is implemented through a political system with its own cleavages.

B. Persons with Disabilities and Other Categories

• The Rights of Persons with Disabilities Act, 2016 raised reservation in government employment to 4% and in higher education to 5%, and expanded the recognised categories from seven to twenty-one.
• Reservation for transgender persons has been directed in principle following the recognition of the third gender, though implementation remains uneven.
• Several states operate reservation for domicile, women in state services and specified minorities, and the constitutional limits of these remain contested.

C. The Private Sector Question

• Constitutional reservation binds the state, so the shrinking share of public employment steadily reduces the reach of the policy — which is the structural argument for extending it.
• Article 15(5) already extends reservation to private unaided educational institutions, so the principle that private bodies may be bound is established.
• Arguments for extension point to the fact that liberalisation moved employment to a sector the remedy does not reach; arguments against point to efficiency, to the difficulty of enforcement, and to the availability of alternatives such as supplier diversity and skilling.
• No general private-sector reservation exists at the national level, and the debate remains unresolved.

Assessment

A. What It Has Achieved

• Political representation is the clearest success. Reserved constituencies have produced continuous SC and ST presence in every legislature since 1952, and local reservation has restructured village-level politics.
• Access to public employment and higher education has created a substantial first-generation professional class among groups previously excluded from both by law and custom.
• The dignity effect is real though unquantifiable. The presence of members of formerly excluded groups in positions of visible authority alters what both groups regard as normal.

B. What It Has Not Achieved

• Economic inequality has widened over the same period. The top 1% now holds roughly 22.6% of national income and 40.1% of wealth, the highest recorded concentration in Indian history — which indicates that a policy operating on public employment and education cannot by itself address a distribution determined mostly elsewhere.
• Benefits remain concentrated within sub-groups and generations, which is the problem sub-classification and the creamy layer are attempts to address.
• Social exclusion persists independently of economic advance. Caste endogamy remains near-universal and discrimination in housing and social interaction is widely documented, which supports the argument that a purely economic criterion misidentifies the harm.
• The category has hardened. Group identity has been institutionally reinforced, competitive claims to inclusion have multiplied, and no group has ever been removed from a list — which is the perpetuation objection borne out.

C. The Design Questions That Remain

• Should the criterion be social or economic? Janhit Abhiyan moved toward the economic; the dissent and the SC/ST creamy-layer debate push back. The answer depends on what the policy is remedying.
• Should benefits be graded within groups? Davinder Singh answers yes, and the implementation question — what empirical basis is adequate — is now the live one.
• Should there be an exit? The policy is justified as temporary and has expanded for seventy-five years. No mechanism exists for determining when a group no longer needs it, and none is politically available.
• Should the reach extend to the private sector? As public employment shrinks, the remedy reaches a smaller share of the labour market each year.
• What replaces or supplements it? Improving school quality, nutrition and early-childhood provision addresses disadvantage at the point where it is created rather than at the point of selection — and is the least contested and least implemented of the alternatives.

Affirmative Action in Comparative Perspective

India is not the only country to have tried this, and the comparison sharpens the Indian argument by showing which features are essential and which are local.

A. The United States

• American affirmative action arose from executive action and judicial interpretation rather than constitutional text, which made it structurally weaker than India’s from the outset.
• Bakke (1978) struck down fixed racial quotas while permitting race to be considered as one factor among many — the origin of the “holistic review” model.
• Grutter (2003) upheld race-conscious admissions on a diversity rationale rather than a compensatory one, and expressed the expectation that such preferences would no longer be necessary in twenty-five years.
• Students for Fair Admissions (2023) ended it. The Supreme Court held that race-conscious admissions at Harvard and the University of North Carolina violated the Equal Protection Clause, effectively terminating the practice in university admissions.
• The contrast with India is instructive on two counts.
  • The American programme rested on a diversity justification, which is vulnerable because diversity can be pursued by other means; the Indian programme rests on historical exclusion and constitutional mandate, which is harder to argue away.
  • American affirmative action was never constitutionally entrenched, so a change in judicial composition could end it. India’s is written into the text, and ending it would require amendment.

B. Other Jurisdictions

• Malaysia’s bumiputera policy, in operation since 1971, is the closest analogue in scale and duration. It reserves extensive advantages for the ethnic Malay majority, and is the standard cautionary example of a programme that benefits a numerical majority and has proved politically impossible to withdraw.
• South Africa’s Employment Equity and Black Economic Empowerment framework addresses a majority excluded by law until 1994, and faces the familiar criticism of elite capture — benefits concentrating in a politically connected stratum.
• Brazil introduced racial quotas in federal universities in 2012, with the distinctive feature of combining race with public-school attendance and income, which is one of the few designs that attempts to target social and economic disadvantage simultaneously.
• The comparative lesson is consistent across cases. Programmes justified by historical exclusion are more durable than those justified by present diversity; programmes without internal gradation concentrate benefits at the top of the beneficiary group; and no country has successfully designed an exit.

Conclusion

• Affirmative action is where the argument between formal and substantive equality stops being theoretical. Anyone who accepts that treating unequals equally perpetuates inequality is committed to something in this family; the disagreement is about form, extent and duration.
• The policy is best defended on distributive rather than compensatory grounds. The forward-looking argument — that present starting points are unequal, so formally equal opportunity is a fiction — avoids the transmission problem that the historical-debt argument cannot escape.
• The strongest objections are not to the principle but to the design. Elite capture, category hardening and the absence of an exit are real, evidenced, and remediable — and Indian constitutional law has been slowly remedying them through the creamy layer and now sub-classification.
• India’s experience is the most instructive available, precisely because the policy has run longest and reached furthest here. It shows that affirmative action can transform representation and dignity while leaving the underlying distribution of wealth substantially untouched — which is an argument for supplementing it, not for abandoning it.
• The deepest unresolved question is Ambedkar’s rather than Rawls’s.
  • If the disadvantage being remedied is social exclusion, then economic criteria misidentify it, and economic advance will not end it.
  • If it is material deprivation, then caste is a proxy that will eventually outlive its usefulness.
  • Indian law has not chosen between these — and until it does, the policy will keep expanding along both axes at once.

Affirmative action is a bridge, not a destination — and the argument about it is really an argument about how far there still is to go.