Between 1928 and 1947, competing groups drafted blueprints for India’s future constitution. These documents reveal a divergent views about what a constitution could or should do about caste. For Dalit leaders, especially Ambedkar, the constitutional question was inseparable from the caste question—any constitution that failed to dismantle the social and material structures sustaining untouchability would be a betrayal. For Congress leaders and liberal reformers, caste was one form of discrimination among others, addressable through declarations of equality and gradual social reform. Examining documents from this period shows how constitutional drafting confronted the problem of caste, and why the Constitution that emerged in 1950 contained both victories and compromises.
Ambedkar and Dalit organizations produced documents across nearly two decades that shared a distinctive approach. From the beginning, they rejected euphemism. The 1930 Scheme of Political Safeguards opened: “The Depressed Classes cannot consent to subject themselves to majority rule in their present state of hereditary bondsmen.” Where Congress documents spoke carefully of “disabilities” and “backwardness,” Ambedkar named bondage, deliberately echoing America’s 1866 Civil Rights Protection Act.
By 1947, when States and Minorities was submitted to the Constituent Assembly, “hereditary bondsmen” had softened to “social status.” But the enumeration remained specific: inns, educational institutions, roads, wells, public conveyances, theatres. Ambedkar listed them because generalized equality provisions would never address the thousand small exclusions through which caste operated.
Every Dalit document insisted that ending untouchability required criminalization. The 1930 Scheme proposed five years imprisonment for denying Dalits access to public facilities. The 1947 document established jurisdiction in special tribunals rather than ordinary courts, recognizing that caste Hindu judges could not be trusted to prosecute caste discrimination.
The most radical criminalization targeted social and economic boycott. States and Minorities defined boycott with exhaustive precision—refusing to rent land, conduct business, maintain social relations, or employ labor. This was far more threatening to the caste order than banning temple entry denial. Boycott was the enforcement mechanism. When a Dalit challenged an upper-caste landlord, the entire dominant caste would refuse to employ them, sell to them, or rent them land. Criminalizing this struck at how caste hierarchy actually worked.
Ambedkar’s documents show a growing conviction that legal equality without economic transformation was worthless. The 1930 Scheme focused on access and representation. The 1944 Political Demands introduced “separate settlements” that would physically remove Dalits from Hindu villages. But the 1947 States and Minorities made the most extraordinary leap, declaring that key industries and agriculture “shall be State Industry.”
This was central to caste abolition. Ambedkar understood that untouchability was economic as much as ritual. Dalits remained bondsmen because they lacked land. The village economy—Dalits landless, dependent on upper-caste landlords for work—made untouchability enforceable. Laws against discrimination meant nothing if a Dalit needed employment from the landlord who denied them well access. This is why States and Minorities proposed constitutionally mandating collective farms.
All Dalit documents insisted on separate electorates. The 1932 Poona Pact had been a defeat because it replaced separate electorates with reserved seats in joint electorates. Separate electorates gave Dalits the power to elect militants who would fight for their interests. Joint electorates meant upper-caste voters determined which Dalits got elected—favoring those willing to compromise.
Congress and liberal documents took a strikingly different approach. The 1928 Nehru Report, the 1931 Karachi Resolution, and the 1945 Sapru Committee Report all treated caste as one form of prejudice among others. The Nehru Report’s language was typical: “no person shall by reason of religion, caste or creed be prejudiced in public employment.” Caste appeared in lists, alongside religion, as prohibited grounds for discrimination.
For Dalit documents, caste was not comparable to religious discrimination. It was simultaneously economic, social, ritual, and occupational, enforced through mechanisms like boycott that had no parallel elsewhere. Treating it as one item in a list missed its specificity. Congress documents could mention “public wells” in the same breath as roads and schools, making untouchability seem like an access problem—just let them use wells and the problem is solved. This erased the economic relations that made denial possible.
None of the Congress documents proposed criminalization. The Nehru Report and Karachi Resolution declared equality principles but established no enforcement mechanisms. The Sapru Report proposed a “Minorities Commission” to “watch” and “report” but with no prosecution powers. This reflected a theory that education and development would erode caste prejudice. Gandhi epitomized this in opposing criminalization of untouchability, preferring appeals to conscience and moral suasion.
Congress documents also treated property and caste as separate issues. The Karachi Resolution addressed economic matters—land rent reduction, progressive taxation, living wages—but when it discussed caste, it spoke of non-discrimination. When it discussed land, it spoke of rent reduction. The two never connected. That landlessness made untouchability enforceable, that upper-caste monopoly over land was the material basis of caste hierarchy—this never fully registered in Congress thinking.
What is striking is that no document—not even Ambedkar’s most radical—declared “caste is hereby abolished.” This reflected political realism about what constitutions could accomplish. Consider land reform, far less radical than caste abolition. The Constitution included Directive Principles calling for land redistribution. Yet even this proved impossible. Zamindari Abolition Acts eliminated some intermediaries but left distribution untouched. Land ceiling laws were gutted through loopholes. By the 1970s, meaningful redistribution was dead, killed by landowning castes who controlled state legislatures.
If land reform could be so thoroughly sabotaged, what hope for abolishing caste? Caste structured marriage, kinship, occupation, residence, ritual life. Simply declaring it abolished would have been unenforceable. More fundamentally: who would enforce such abolition? Police, judiciary, bureaucracy were staffed overwhelmingly by upper castes. The entire apparatus of local governance was controlled by dominant castes who benefited from the system.
This is why Ambedkar’s documents never called for abolishing caste as such. They targeted caste’s enforcement mechanisms: the economic dependence created by landlessness, the boycotts that punished challengers, the denial of public resources. Ambedkar understood you could not abolish caste by declaration. You could only dismantle the material and social conditions that made caste hierarchy enforceable. His proposal for state ownership of agriculture was an attempt to destroy the economic basis of caste power. If Dalits were not dependent on upper-caste landlords for work, the everyday mechanisms of enforcement would collapse.
Ambedkar’s documents had shown what anti-caste transformation required: not just legal equality but economic redistribution, not just non-discrimination but criminalization of boycott, not just representation but structural change in who controlled land and capital. Congress documents showed liberal constitutionalism’s limits: acknowledging oppression while refusing to encode strong enforcement. The Constitution’s caste provisions are stronger because Ambedkar fought for them. Article 17 abolished untouchability and made its practice an offense punishable by law. Yet they are weaker than he demanded—a compromise between those who understood caste as economic exploitation requiring structural dismantling and those who saw it as social prejudice amenable to gradual reform that could be left to future legislatures rather than constitutionally mandated