New Constitutional Rule: States May Sub‑Classify Scheduled Castes for Reservations, Subject to Rational Criteria and Data – EV Chinnaiah Overruled

Introduction

On 1 August 2024, a seven‑Judge Constitution Bench of the Supreme Court of India (Dr. D.Y. Chandrachud, CJI; B.R. Gavai, Vikram Nath, Pankaj Mithal, Satish Chandra Sharma, and Bela M. Trivedi, JJ.) delivered a landmark judgment in State of Punjab v. Davinder Singh, 2024 INSC 562. The Court resolved a question that had vexed Indian equality jurisprudence for two decades: whether sub‑classification among Scheduled Castes (SCs) for the purpose of reservations is constitutionally permissible. The Court (by majority) held that it is, thereby overruling E.V. Chinnaiah v. State of A.P. (2005) 1 SCC 394.

The litigation arose from challenges to three measures: (i) Section 4(5) of the Punjab Scheduled Castes and Backward Classes (Reservation in Services) Act, 2006, which gave first preference within the SC quota to Balmikis and Mazhabi Sikhs; (ii) a 1994 Haryana notification splitting SCs into Block ‘A’ and Block ‘B’ with a 50:50 split and mutual interchangeability; and (iii) the Tamil Nadu Arunthathiyars (Special Reservation) Act, 2009, reserving a sub‑quota for Arunthathiyars from within the SC list for education and State services. These measures had been invalidated by reference to Chinnaiah’s holding that once the President notifies the SC list under Article 341, that list forms an indivisible, homogeneous class immune from State‑level sub‑division for reservation.

In 2020, a five‑Judge Bench in State of Punjab v. Davinder Singh (2020) 8 SCC 1 doubted Chinnaiah and referred the issue to a larger Bench. The present seven‑Judge Bench answered that reference, laying down a detailed framework for when and how States may sub‑classify SCs for affirmative action under Articles 15(4)/15(5) and 16(4), while preserving the integrity of the Presidential List under Article 341.

Summary of the Judgment

• Sub‑classification within SCs is constitutionally permissible. Article 14 permits reasonable classification and, by parity, reasonable sub‑classification of a class that is not homogeneous for a given legislative purpose. The equality code (Articles 14–16) embodies substantive, not merely formal, equality. The Court overruled E.V. Chinnaiah and held that States may sub‑classify SCs for reservations to secure substantive equality and equal opportunity.
• Article 341 does not create an immutable, internally homogeneous class. The Presidential List identifies which castes/tribes are “Scheduled Castes” in relation to each State/UT but does not say that all such groups are identically situated inter se for every constitutional purpose. Article 341(1) creates constitutional identity; Article 341(2) vests inclusion/exclusion power exclusively in Parliament. Sub‑classification for reservations does not add to or subtract from the List and therefore does not offend Article 341(2), provided the policy does not exclude other SC groups from the benefit entirely.
• Historical and empirical heterogeneity within SCs is acknowledged. The Court canvassed colonial and post‑colonial materials (e.g., Hutton’s 1931 tests, the 1936 and 1950 Orders, sociological studies) to establish that degrees and modalities of untouchability and social exclusion vary widely across SC groups, and even across regions. This heterogeneity justifies rational sub‑classification to reach the “most deprived among the deprived”.
• Source of State power. The State’s power to design reservations and allied measures for SCs derives from Articles 15(4), 15(5), and 16(4). These provisions permit “any special provision” for advancement (Art. 15) and reservation for “any backward class” inadequately represented (Art. 16), which includes SCs and STs. Legislative competence flows from Articles 245–246 read with the Seventh Schedule (Entry 41 List II for services; Entry 25 List III for education) insofar as the State is not altering the Presidential List.
• Criteria & yardsticks. A constitutionally valid sub‑classification must:
  • Be based on a rational principle (intelligible differentia) linked to the statutory/constitutional objective (nexus).
  • Track inter‑se social backwardness among SC groups and/or inadequacy of effective representation in State services/education (with “effective” understood across posts/grades, not mere headcount).
  • Be supported by quantifiable, demonstrable data showing that the benefitted subgroup is more disadvantaged and inadequately represented.
• Design models & limits. The Court sketched permissible design options:
  • Preference model: Give first preference to a subgroup over a defined percentage of the SC quota, with unfilled seats reverting to other SC groups. A preference over all SC‑reserved seats risks exclusion of others and is unconstitutional.
  • Exclusive sub‑quota model: Earmark a percentage within the SC quota exclusively for a subgroup, carry forward unfilled seats under Article 16(4B), provided the design does not bar other SC groups from accessing the balance of the SC quota.
  Any model that effectively excludes one or more SC groups from accessing the SC quota violates Article 341(2)/Article 14.
• Judicial review. Courts will review whether: (a) the chosen rational principle actually distinguishes the targeted subgroup’s disadvantage; and (b) effective inadequacy of representation exists (backed by quantifiable data). The State’s opinion is reviewable for manifest arbitrariness and evidentiary support; courts will not substitute their policy choices.
• Efficiency reimagined; Article 335 not a brake. Efficiency in Article 335 is not an external limit on Articles 15/16, but a facet of substantive equality. “Efficiency” cannot be reduced to exam scores or neutral criteria that a priori exclude historically disadvantaged groups. The proviso to Article 335 (relaxation of qualifying marks/standards in promotion) was reaffirmed.
• Creamy layer for SC/ST (majority view). Separate concurring opinions (Gavai, Nath, and Satish Chandra Sharma, JJ.) affirm that the creamy layer principle applies to SCs and STs for reservation benefits, though criteria for exclusion may differ from OBCs given the nature of caste‑based stigma. The Chief Justice’s lead opinion does not resile from Jarnail Singh and a majority endorses creamy layer application.
• Outcome and next steps. EV Chinnaiah is overruled. The registry is to place the matters before appropriate Benches to test the specific State measures against the framework (rational basis, data, effective representation, non‑exclusion).

Precedents Cited and Their Influence

• M.R. Balaji v. State of Mysore (AIR 1963 SC 649): Recognised caste as a relevant (but not sole) indicator of social backwardness; set the 50% ceiling. Often read as formalist; later re‑oriented by NM Thomas.
• State of Kerala v. N.M. Thomas (1976) 2 SCC 310: A turning point. Article 16(4) held as an emphatic facet—not an exception—of 16(1). Introduced substantive equality, upheld classification within an integrated class, and questioned narrow “efficiency”. Krishna Iyer, Mathew, Fazl Ali, and Ray, JJ. laid the jurisprudential foundation used here.
• Indra Sawhney v. Union of India 1992 Supp (3) SCC 217: Upheld OBC reservations; recognised heterogeneity within backward classes; endorsed sub‑classification among OBCs; clarified that SCs/STs fall within “backward class” under Article 16(4) even though 15(4) mentions them distinctly.
• General Manager, S. Railway v. Rangachari (1962) 2 SCR 586; CA Rajendran (AIR 1968 SC 507); Devadasan (1964) 4 SCR 680; RK Sabharwal (1995) 2 SCC 745; M. Nagaraj (2006) 8 SCC 212; Jarnail Singh (2018) 10 SCC 396; BK Pavitra (II) (2019) 16 SCC 129; Neil Aurelio Nunes (2022) 4 SCC 1: Together, these cases reflect the evolving balance of representation, efficiency, promotions, consequential seniority, creamy layer, and the 50% ceiling. The present judgment harmonises these strands, clarifying that efficiency is inclusive and that sub‑classification furthers equal opportunity.
• EV Chinnaiah v. State of A.P. (2005) 1 SCC 394: Held that SCs in the Presidential List form a single, indivisible class; sub‑classification impermissible. Overruled for (i) conflating Article 341 identity with homogeneity for reservation design, (ii) reading Article 341(2) as barring any intra‑SC prioritisation even if no inclusion/exclusion occurs.

Legal Reasoning: How the Court Reached Its Conclusions

1) Equality Code = Substantive Equality, Not Formal Sameness

Articles 14, 15 and 16 are a single “equality code”. Reservation is an instrument to realise equal opportunity (Article 16(1)). Formal equality—treating unlike persons alike—entrenches historical inequality. Sub‑classification is a facet of reasonable classification under Article 14: if a class is not homogeneous for a given purpose, the State may carve sub‑classes on a rational basis tied to the purpose. This principle has long been recognised (e.g., Triloki Nath Khosa).

2) Article 335: Efficiency Is Inclusive

The Court rejects the binary of merit vs reservation. Efficiency of administration must be defined in a manner consistent with Article 16’s promise of factual equal opportunity. Minimum qualifications/standards can be relaxed; higher marks are not a true proxy for administrative efficiency, given structural inequities.

3) Article 341: Identification ≠ Immutable Homogeneity

Article 341(1) assigns constitutional identity to notified groups as “Scheduled Castes” for a given State/UT. Article 341(2) reserves to Parliament the power to include/exclude. The majority carefully parses the “deeming” language: even if a “deeming fiction” is assumed, it is limited to identity for constitutional purposes, not to a presumption that all notified SC groups are similarly situated inter se for every policy. Sub‑classification that does not change the List (no inclusion/exclusion) does not violate Article 341(2).

4) Demonstrated Heterogeneity Within SCs

The Court details how “untouchability” and social exclusion have manifested differently across geographies and occupations (e.g., scavengers vs leather workers vs weavers), and how intra‑Dalit hierarchies often replicate discriminatory patterns. This historically grounded heterogeneity justifies tailoring within the SC quota to ensure benefits reach the “most downtrodden”.

5) State Power and Competence

States may design reservations and allied measures within their legislative and executive domains (Entry 41 List II: services under the State; Entry 25 List III: education), so long as they do not alter the Presidential List. Articles 15(4)/15(5) (“any special provision”) and 16(4) (“any backward class”) are the constitutional fonts for sub‑classification policies within the SC quota for advancement/equal opportunity.

6) Criteria and Yardsticks for Sub‑Classification

• Backwardness standard: For Article 15(4): social and educational backwardness (which are causally linked). For Article 16(4): social backwardness plus inadequacy of representation—measured by effective representation across grades/posts, not mere headcounts.
• Rational principle and nexus: The principle used to distinguish SC groups (e.g., traditional occupation stigmas, persistently low representation in higher posts, compounded stigma) must be rational and tied to the goal of equal opportunity and advancement.
• Data: States must collect quantifiable, demonstrable data showing inter‑se backwardness/inadequacy. Pure assertions or political expediency will fail judicial scrutiny.

7) Design Models and Constitutional Limits

• Preference over a sub‑percentage within the SC quota is valid; preference over all SC‑reserved seats is not.
• Exclusive sub‑quotas are permissible for a portion of the SC quota, coupled with carry‑forward under Article 16(4B), provided other SC groups have real access to the remaining SC quota.
• No micro‑classification: Avoid atomising into countless sub‑groups without meaningful differences; the basis must be substantial (cf. Nakara).

8) Judicial Review

Courts will examine (a) the rationality of the differentiating principle; (b) the nexus to the objective; and (c) the data establishing inter‑se backwardness and effective inadequacy of representation. The State’s “opinion” is not immune; it must be informed and supported by evidence. However, courts will not micromanage policy choices that clear these hurdles.

Positions Within the Bench

• Lead Majority (Chandrachud, CJI): Sub‑classification permissible; Article 341 identification does not bar rational sub‑grouping for reservation; Article 335 read as supporting inclusive efficiency; detailed framework on criteria, limits, and review.
• Concurring (Gavai, J.): Fully concurs on sub‑classification. Additionally, expressly affirms the applicability of creamy layer to SC/ST reservation, with the caveat that exclusion criteria for SC/ST may differ from OBCs due to the unique nature of caste stigma. Adds design guardrails and analogies (e.g., “general compartment” metaphor).
• Concurring (Vikram Nath, J.): Agrees with the Chief Justice and Gavai, J., including on creamy layer for SC/ST.
• Concurring (Pankaj Mithal, J.): Joins the majority; additionally urges a broader policy re‑look at reservations, including: (a) limiting reservation benefits for SC/ST to one generation within a family that has substantially progressed; (b) periodic reviews; and (c) moving towards a casteless society.
• Concurring (Satish Chandra Sharma, J.): Agrees with the majority and with the applicability of creamy layer to SC/ST.
• Dissent (Bela M. Trivedi, J.): Would uphold EV Chinnaiah and deny State power to sub‑classify SCs. Emphasizes Article 341’s text and the need for stare decisis; considers sub‑classification an impermissible “tinkering” with the Presidential List; cautions that affirmative action cannot be a vehicle to bypass express constitutional restraints.

Impact and Implications

• Policy space opened: States can now design intra‑SC sub‑quotas or preferences to channel benefits to the most deprived SC groups (e.g., scavenging communities, Arunthathiyars, Madigas), provided they base measures on sound data and avoid exclusionary designs.
• Data systems will be decisive: Expect a wave of State commissions/surveys to assess inter‑se backwardness and representation patterns across grades/services (and in education). Poorly supported designs risk being struck down.
• Reconfiguration of existing schemes: Sub‑quotas/preferences within SC reservations can be recalibrated; Tamil Nadu’s Arunthathiyar sub‑quota, Punjab’s Balmiki/Mazhabi Sikhs preference, and similar initiatives in other States can be reconsidered against the new framework.
• Creamy layer for SC/ST: With a majority affirming applicability, States may move to craft SC/ST‑specific creamy layer criteria (distinct from OBC thresholds) to prevent the better‑off from cornering benefits meant for the most deprived.
• Integrity of the List preserved: Parliament alone may include/exclude under Article 341(2). The judgment strictly prohibits any State policy whose effect is to “write out” some SC groups from the benefit.
• Efficiency doctrine modernised: Recruitment, promotion, and evaluation systems will need to reflect inclusive metrics, rejecting proxies that perpetuate a priori exclusion.

Complex Concepts Simplified

• Sub‑classification vs micro‑classification: Sub‑classification is dividing a class into meaningful subgroups where heterogeneity exists for the policy purpose (e.g., the “most deprived” among SCs). Micro‑classification is creating numerous tiny categories without substantive differences—this is impermissible.
• Effective representation: Not a crude headcount. The question is whether a group is represented across posts and grades in a way that gives it a real voice and presence, not confined to the lowest rungs.
• Article 341’s “deeming”: The President’s notification “deems” listed communities to be SCs for constitutional purposes. It fixes identity (who is an SC) but does not bar rational allocation within the SC quota so long as no inclusion/exclusion occurs.
• Preference vs exclusive sub‑quota: A preference gives first claim for a part of the SC quota but allows others to fill leftovers; an exclusive sub‑quota ring‑fences a portion (with carry‑forward), leaving the balance open to all SC groups. Both are permissible if they do not lead to exclusion of other SC groups.
• Creamy layer for SC/ST: The principle excludes the socially advanced within a beneficiary class to ensure benefits reach those who still bear stigma/deprivation. For SC/ST, criteria may look different than OBCs (e.g., high‑status public positions attained, demonstrable social mobility), given the unique dimension of untouchability.

Conclusion: Key Takeaways

• Sub‑classification among SCs for reservations is constitutionally valid, provided it rests on a rational principle tied to inter‑se backwardness and effective inadequacy of representation, and is backed by quantifiable data.
• Article 341’s Presidential List fixes identity; it does not bar States from designing intra‑SC allocation so long as no inclusion/exclusion occurs and other SC groups are not shut out.
• Design matters: avoid “preference over all SC seats” and exclusionary models; permissible models include (a) preference over a sub‑percentage with fallback, and (b) exclusive sub‑quotas within limits, with carry‑forward under Article 16(4B).
• Courts will scrutinize both the principle and the data. The State’s opinion is reviewable for arbitrariness, but courts will not supplant policy if the constitutional standard is met.
• “Efficiency” in Article 335 is not a veto; it is to be read as enhancing inclusion and representativeness.
• A majority affirms that the creamy layer principle applies to SC/ST reservations, with SC/ST‑appropriate criteria, to prevent capture by the relatively advanced.
• EV Chinnaiah is overruled; the equality code is re‑aligned to substantive equality, ensuring that affirmative action actually reaches the most deprived among the Scheduled Castes.