Caste System in India: Constitutional Provisions & Modern Reality

The caste system is not a single system. It is two overlapping systems that students collapse into one, and that confusion costs marks every single year. Varna is a fourfold theoretical classification found in Vedic texts. Jati is the lived, ground-level reality of thousands of endogamous groups that actually govern marriage, occupation, and social status across India. If you write about "four castes" in your answer, you have already lost the examiner.

I spent years watching students at Rau's IAS mix these up in GS1 answers. The distinction is not academic trivia. UPSC Mains 2019 GS1 asked directly about the "social impact of caste" and the toppers who scored well were the ones who demonstrated they understood caste as a living, mutating institution, not a frozen textbook category.

Varna vs Jati: The Distinction That Matters

Varna refers to the fourfold classification in the Rig Veda’s Purusha Sukta: Brahmins (priests/teachers), Kshatriyas (warriors/rulers), Vaishyas (traders/agriculturists), and Shudras (service providers). This is a theoretical framework. It was never a neat description of how Indian society actually functioned.

Jati is the operational unit. India has over 3,000 jatis, each with its own hierarchy, customs, and rules of endogamy. A Brahmin in Tamil Nadu and a Brahmin in Kashmir belong to different jatis with different social standing, dietary habits, and ritual practices. Jati determines who you marry, who you eat with, and in many parts of India, what work you do.

Common student mistake: Writing "India has four castes" in an answer. This is factually wrong. India has four varnas and thousands of jatis. The Portuguese word "casta" (meaning breed or race) was applied indiscriminately to both systems, creating the English word "caste." Always specify whether you mean varna or jati.

The colonial period hardened jati boundaries. The 1881 Census under Risley attempted to rank all jatis into a linear hierarchy, turning fluid social relationships into rigid administrative categories. This is what scholars call the "colonial construction of caste", and it is a legitimate UPSC perspective. But do not overcorrect: caste hierarchy, untouchability, and occupational restrictions existed well before the British arrived. The truth is that colonialism did not invent caste but did freeze and bureaucratize it.

Constitutional Provisions: The Framework of Equality

The Constitution does not ignore caste. It acknowledges caste as a social reality and then systematically dismantles its discriminatory aspects while using caste categories for affirmative action. This dual approach is what makes India's constitutional treatment of caste unique.

Article 17 is the most revolutionary provision. No other constitution in the world specifically abolishes a social practice by name. The framers knew that legal equality (Article 14) alone would not end untouchability. They needed a specific prohibition backed by criminal penalties.

Mandal Commission & the Reservation Landscape

The Mandal Commission (Second Backward Classes Commission) was appointed in 1979 under B.P. Mandal’s chairmanship and submitted its report in 1980. Its core recommendation: 27% reservation for Other Backward Classes (OBCs) in central government services and educational institutions.

The report sat unimplemented for a decade. In August 1990, Prime Minister V.P. Singh announced its implementation. The political firestorm was immediate. Anti-reservation protests erupted across North India. Rajiv Goswami's self-immolation became the defining image of that moment.

The legal challenge came in Indra Sawhney v. Union of India (1992), the most important reservation judgment in Indian constitutional history. The nine-judge bench ruled:

• 27% OBC reservation is valid
• Total reservation cannot exceed 50% (the "50% ceiling")
• "Creamy layer" among OBCs must be excluded
• No reservation in promotions for OBCs (later modified for SCs/STs)
• Backward class identification should be based on caste plus other factors, not caste alone

Common student mistake: Stating that the 50% ceiling is a constitutional provision. It is not. It is a judicial guideline from Indra Sawhney. The Constitution does not prescribe any cap. Tamil Nadu exceeds 50% (69% reservation) under Ninth Schedule protection. The 103rd Amendment (2019) adding 10% EWS reservation also effectively breaches this ceiling, bringing total reservation to 59.5% at the central level.

The 103rd Amendment itself survived constitutional challenge in Janhit Abhiyan v. Union of India (2022), where a 3-2 majority upheld EWS reservation. The two dissenting judges argued it violated the basic structure by excluding SCs, STs, and OBCs from EWS benefits.

SC/ST Prevention of Atrocities Act 1989

The Scheduled Castes and Scheduled Tribes (Prevention of Atrocities) Act, 1989 was enacted because existing criminal law was inadequate to address caste-based violence. The Protection of Civil Rights Act 1955 dealt with untouchability but not with the full spectrum of atrocities.

Key features of the Act:

• Defines specific offences against SCs/STs (forcing to eat noxious substances, dumping waste, sexual exploitation, land dispossession)
• Exclusive Special Courts for speedy trials
• Non-bailable and non-compoundable offences
• Provision for preventive action by district administration
• Rehabilitation and compensation for victims

The Act was amended in 2015 to expand the list of offences (added garlanding with footwear, denying access to public places, social/economic boycott). The 2018 amendment came after the Supreme Court's controversial ruling in Subhash Kashinath Mahajan v. State of Maharashtra (2018), which had diluted the Act by requiring preliminary inquiry before arrest. The 2018 amendment restored the original stringent provisions and specified that no preliminary inquiry was needed for registering FIRs.

NCRB data consistently shows over 50,000 cases registered annually under this Act. Conviction rates remain low (around 30-35%), which is itself a point worth raising in your UPSC answers.

The Caste Census Debate

The Socio-Economic Caste Census (SECC) 2011 was the first attempt since 1931 to enumerate caste across India's entire population. The economic data was released in 2015. The caste data has never been officially released by the central government. This non-release is itself a politically charged issue.

Bihar's caste survey (2023) broke new ground. Conducted under the Nitish Kumar government, it found:

• OBCs constitute 63% of Bihar's population
• EBCs (Extremely Backward Classes) form 36%
• SCs at 19.65%, STs at 1.68%
• Upper castes at approximately 15.5%

This data reignited the national demand for a caste census. Proponents argue that without accurate caste data, welfare policies are shots in the dark. Opponents argue that enumeration entrenches caste identity and that the 50% reservation ceiling would face political pressure if OBC numbers are shown to be vastly larger than assumed.

My view: accurate data is always better than guesswork. Policy should be evidence-based. The reluctance to count caste while simultaneously administering caste-based policies is an incoherent position.

Modern Reality: Caste Beyond the Village

Caste has not disappeared with urbanization. It has adapted. Anyone who tells you "caste does not matter in cities" has not looked at the data or the matrimonial pages.

In elections: Caste arithmetic remains the foundation of seat allocation, candidate selection, and coalition building across every major party. The "KHAM" formula in Gujarat, the "MY" combination in UP, caste-based vote bank calculations are not aberrations but standard operating procedure.

In matrimony: Over 90% of Indian marriages remain endogamous (within caste). Matrimonial sites explicitly list caste. Even among urban, educated, English-speaking professionals, caste matching in marriage is the norm, not the exception.

In urban spaces: Research by Sukhadeo Thorat and others has documented caste discrimination in housing rental markets in Delhi, Mumbai, and other cities. Muslim and Dalit applicants face systematic exclusion. The Indian Institute of Dalit Studies has produced rigorous evidence on this.

Digital caste discrimination: Caste slurs on social media, targeted harassment of Dalit activists and writers online, and caste-based trolling are well-documented phenomena. The anonymity of digital spaces has not dissolved caste. It has given it new platforms.

Intersectionality: Caste, Gender, and Class

A Dalit woman experiences caste differently from a Dalit man. She faces triple marginalization: caste, gender, and often class. The concept of intersectionality, while originating in Western feminist scholarship (Kimberle Crenshaw, 1989), applies directly to the Indian caste context.

Dalit women face disproportionate rates of sexual violence, manual scavenging (which is overwhelmingly performed by Dalit women), and exclusion from both mainstream feminist movements and male-dominated Dalit movements. The Hathras case (2020) brought national attention to this intersection.

Dalit Christians and Dalit Muslims face a constitutional anomaly. The Constitution (Scheduled Castes) Order, 1950 originally restricted SC status to Hindus. It was amended in 1956 to include Sikhs and in 1990 to include Buddhists. But Dalit converts to Christianity and Islam remain excluded from SC benefits, despite continuing to face caste discrimination. The Ranganath Misra Commission (2007) recommended their inclusion, but no action has been taken.

For UPSC, the intersectionality angle is gold. It shows analytical depth. When discussing caste, always consider how it interacts with gender, class, religion, and geography.

PDF