mechanisms, laws and bodies for vulnerable sections

Reservation is India's principal instrument of affirmative action, permitted by the Constitution as a facet of substantive equality, and continuously contested over its beneficiaries, its ceiling and whether caste or economic deprivation should define entitlement.

What it is

Reservation is the practice of setting aside a fixed proportion of a scarce public good — seats in educational institutions, posts in public employment, and constituencies in elected bodies — for members of specified groups. It is the dominant, though not the only, form of affirmative action in India; other forms include fee concessions, scholarships, hostels, relaxed qualifying standards, age relaxations and preferential procurement.

The constitutional anchor lies in the equality provisions. Article 14 guarantees equality before the law, Article 15 prohibits discrimination on grounds including caste, and Article 16 guarantees equality of opportunity in public employment. Reservation is authorised by enabling clauses attached to these: Article 15(4) permits special provision for socially and educationally backward classes and for Scheduled Castes and Scheduled Tribes; Article 15(5), inserted in 2005, extends this to admissions in private as well as state educational institutions, excluding minority institutions; Article 16(4) permits reservation of appointments for backward classes inadequately represented in the services; Articles 16(4A) and 16(4B), added in the late 1990s, permit reservation in promotion for Scheduled Castes and Tribes and the carrying forward of unfilled reserved vacancies; and Articles 15(6) and 16(6), inserted by the 103rd Amendment, permit up to ten per cent for economically weaker sections outside the existing reserved categories. Article 46 directs the state to promote the interests of Scheduled Castes, Scheduled Tribes and other weaker sections; Article 335 requires their claims to services to be taken into consideration consistently with the maintenance of efficiency of administration. Political reservation rests on separate provisions: reserved seats in the Lok Sabha and state assemblies, and reserved seats and chairpersonships in panchayats and municipalities, including for women.

Two doctrinal points define the instrument. First, the courts have moved from treating reservation as an exception to equality towards treating it as a facet of substantive equality — a way of achieving Article 14's promise rather than a derogation from it. Second, the enabling clauses confer power on the state, not a right on the individual: no citizen can compel a government to provide reservation, though a government that does so must do it lawfully. Who counts is settled differently for each group. Scheduled Castes and Scheduled Tribes are notified in Presidential lists, alterable only by Parliament. Backward classes are identified administratively, subject to judicial tests of social and educational backwardness, with a National Commission for Backward Classes given constitutional status in 2018. Economically weaker sections are identified by income and asset criteria alone.

Why it is contested

The first conflict is between equality of opportunity understood as procedural neutrality and equality understood as a distribution of positions that reflects society. Opponents of reservation argue that a public post should go to the best-performing applicant and that departing from this both wrongs the individual passed over and degrades administrative capacity, invoking the efficiency proviso in Article 335. Defenders answer that examination performance is itself the product of accumulated household advantage — schooling, language, coaching, nutrition, freedom from stigma — so that treating it as raw individual desert launders inherited privilege as merit, and that a bureaucracy drawn from a narrow social base is less legitimate and often less effective in dealing with the populations it governs. The Supreme Court has increasingly accepted a substantive account of merit, but the argument is live in every recruitment cycle.

The second conflict is about the axis of disadvantage. Caste-based reservation rests on the claim that the relevant injury is social — exclusion, stigma, denial of ritual and physical access — for which economic status is a poor proxy, since a poor Brahmin and a poor Dalit face different worlds. The economic-criterion position holds that deprivation today is largely material, that a caste-based scheme insults the poor of unlisted castes, and that public policy should target need directly. The introduction of an economically weaker sections quota accepted the second logic for the first time in the Constitution's text, and did so above the existing quotas, which is why it was challenged as breaching both the ceiling and the principle that reservation compensates for social exclusion.

The third conflict is between groups and within them. Because the goods being distributed are finite and, in the case of formal public employment, shrinking as a share of total employment, every extension of a quota is a redistribution among claimants rather than a costless addition. Dominant agrarian castes in several states — Marathas, Jats, Patidars, Kapus — have pressed for backward-class status on the strength of agrarian distress and thinning public jobs, which the older Mandal-derived constituencies read as an attempt to recapture what was conceded to them. Simultaneously, less advanced sub-groups within the Scheduled Castes and within the backward classes argue that a small number of relatively better-placed communities absorb most of the benefit, and demand internal sub-quotas. The creamy-layer device — excluding the advanced among the backward — pursues the same targeting objective within a household rather than a community.

Underlying all of this is a data vacuum and a federal quarrel. There has been no comprehensive enumeration of castes other than the Scheduled Castes and Tribes since the 1931 census, so the size of the backward classes, and the distribution of benefits among them, is contested arithmetic. States argue that they know their own social structures and should set their own quantum; the courts have insisted on a national ceiling and on evidence-based identification, and a constitutional amendment was needed to restore state power to identify backward classes after a judgment read it away. Finally, the instrument acts on caste relations while depending on caste categories, which is why the topic connects directly to caste and communal relations: reservation both compensates for caste hierarchy and gives every community a stake in maintaining a caste identity.

The competing positions

Social-justice parties and Ambedkarite movements, historically the Dravidian parties in Tamil Nadu and the Mandal-derived parties in the Hindi belt, hold that reservation is not welfare but power-sharing: representation in the institutions of the state is a condition of democratic legitimacy, and the benchmark should be proportionality to population. They therefore want the fifty per cent ceiling abandoned as an artefact of judicial invention rather than constitutional text, a caste census to establish shares, extension to the private sector and the higher judiciary, and vigorous filling of backlog vacancies.

A constitutionalist position, articulated in much of the Supreme Court's own reasoning and by liberal commentators, accepts reservation but insists on limits: that the general category must retain a majority of posts if equality of opportunity is to mean anything, that identification must rest on evidence rather than agitation, and that internal targeting through creamy-layer exclusion and sub-classification is required if the benefit is to reach those for whom it was designed.

The economic-criterion position, which underlies the economically weaker sections quota and has support across the political spectrum, holds that poverty is the operative disadvantage in a market economy and that a scheme blind to the poverty of unlisted castes is neither fair nor politically sustainable.

Proponents of sub-classification — Madigas in the Telugu states, Arundhatiyars in Tamil Nadu, the most-backward strata among the backward classes — argue that graded inequality exists inside the reserved categories and that a single undifferentiated quota entrenches it. Their opponents, including sections of the Dalit movement, argue that the Presidential list is constitutionally homogeneous, that stigma attaches to all Scheduled Castes alike, and that internal division invites political manipulation.

A final position, associated with development economists and some Dalit intellectuals, is not against reservation but against relying on it: since formal public employment and elite higher education absorb a small fraction of entrants to the labour force, quotas cannot substitute for school quality, land and credit access, urban tenancy rights and anti-discrimination enforcement in private markets.

How it developed

The instrument predates the Constitution. Princely Kolhapur and Mysore introduced preferences for non-Brahmin communities in the early twentieth century, Madras issued communal orders allocating posts by community, the Poona Pact substituted reserved seats for separate electorates for the Depressed Classes, and the Government of India Act of 1935 gave the Scheduled Castes a statutory schedule.

After 1950, an early judgment striking down community-wise allocation of college seats prompted the First Amendment, which inserted the enabling clause for backward classes. A first Backward Classes Commission in the 1950s produced a report that was not acted upon. In the 1960s the Supreme Court articulated the fifty per cent ceiling and the rule that caste alone cannot define backwardness. The Mandal Commission reported at the start of the 1980s; its recommendation of twenty-seven per cent for Other Backward Classes in central services was implemented in 1990 and upheld by a nine-judge bench in 1992, which simultaneously affirmed the ceiling as a general rule, struck down a separate economic quota, mandated creamy-layer exclusion for backward classes, and barred reservation in promotions. Parliament reversed the last holding by amendment, and added consequential seniority, relaxed promotion standards and carry-forward of backlog vacancies; Tamil Nadu's higher quantum was placed in the Ninth Schedule.

The 2005 amendment extended reservation to admissions in private unaided institutions and produced twenty-seven per cent backward-class reservation in central educational institutions. The Court in the mid-2000s upheld the promotion amendments subject to conditions of quantifiable data, inadequate representation and administrative efficiency, later relaxing the requirement to prove backwardness for Scheduled Castes and Tribes while extending creamy-layer exclusion to their promotions. A 2005 judgment barred sub-classification within the Scheduled Castes; a seven-judge bench overruled it in 2024 and permitted states to sub-classify on the basis of demonstrated relative backwardness, with several judges observing that the creamy-layer principle should also apply to Scheduled Castes and Tribes. In between, the economically weaker sections amendment of 2019 was upheld by a narrow majority in 2022; Maharashtra's Maratha quota was struck down in 2021 with the Court declining to revisit the ceiling; a commission on sub-categorisation of backward classes submitted its report; Bihar conducted a caste survey and raised its quota above sixty per cent, which was set aside by its High Court; a women's reservation amendment for legislatures was enacted in 2023 with implementation deferred to a future census and delimitation; and the government announced that caste would be enumerated in the next census.

Where it stands

The fifty per cent ceiling survives as doctrine but is porous in practice: the economically weaker sections quota sits above it, taking the notional total in central recruitment close to sixty per cent, and several states operate higher aggregates through Ninth Schedule protection or pending litigation. Whether the ceiling can be revisited by a larger bench, and on what evidence a state may claim the extraordinary circumstances that justify exceeding it, is unresolved.

Sub-classification is now permissible, and states are beginning to legislate internal sub-quotas within the Scheduled Castes; the quality of the data on which they do so will determine whether these survive challenge. The application of the creamy layer to Scheduled Castes and Tribes in direct recruitment remains open, judicially suggested and politically resisted. The report on sub-categorising the backward classes has not been acted upon publicly. Caste enumeration in the census promises to settle the arithmetic but will also generate fresh claims to proportionality, and the interaction of enumeration, delimitation and the women's quota makes the next few years unusually consequential. Reservation in private employment has no national statute; state attempts to reserve private jobs for local candidates have run into constitutional objections. Beyond quotas, anti-discrimination enforcement in private housing, credit and hiring remains thin.

Taking a view

A good answer draws the distinctions the litigation turns on. Reservation is enabling, not mandatory. Vertical reservation for caste categories works differently from horizontal reservation for women or persons with disabilities, which cuts across all categories. Scheduled Castes and Tribes are treated differently from backward classes: their backwardness is presumed rather than proved, and the creamy-layer question is settled for one and open for the other. Adequate representation is not the same as proportional representation. And economic criteria and social backwardness are distinct grounds with distinct justifications, not two versions of the same idea.

The usual trap is to write the topic as a morality contest between merit and quotas, which produces assertion rather than analysis. The second trap is a chronology of amendments and judgments with no trade-off in it. Better to organise around the real tension: the objective is to dismantle a hierarchy without hardening the categories that constitute it, using a shrinking pool of formal-sector positions.

A defensible view can be firm without being partisan. One can hold that the case for reservation as compensation for durable, ascriptive exclusion is settled, while arguing that its legitimacy now depends on targeting — better data, internal sub-classification on evidence, honest creamy-layer thresholds — and on being frank that quotas in public employment cannot reach most of the disadvantaged, so they must be accompanied by school quality, land and credit access and enforceable anti-discrimination law in private markets. Whichever side is taken on the ceiling or on economic criteria, say which goal is being subordinated to which, and why.