India borrowed due process from America and pretended it didn't

There is a story Indian constitutional lawyers tell about Article 21. It begins with a narrow text — ‘procedure established by law’ — and ends, sometime around 1978, with a Court that liberated the provision from its formalistic past and made it speak the language of substantive justice. The story is true in its broad outlines. It is also a polite fiction in one important respect. Indian constitutional law did not evolve into substantive due process. It borrowed the doctrine, by name and by content, from the United States, after the Constituent Assembly had specifically refused to write it into the text. The borrowing was unacknowledged in 1978 and it remains unacknowledged in 2026. That refusal to call the move what it was has cost us more than honesty. It has corrupted the way Indian scholars do comparative constitutional law, and it has left the doctrine that emerged — our own version of due process — standing on a foundation that no one wants to inspect.

The settled view, stated in its own words

The standard account runs roughly as follows. The framers of India's Constitution were nervous about the American due process clause because of the Lochner era and so chose the narrower Japanese formulation, ‘procedure established by law’. The early Court took the framers at their word in A.K. Gopalan v State of Madras (1950) and held that Article 21 only required a validly enacted procedural law. By the late 1970s, after the Emergency and the moral collapse of the Court in ADM Jabalpur, that formalism had become untenable. Maneka Gandhi (1978) corrected it by reading Article 21, in combination with Articles 14 and 19, to require that the procedure be ‘fair, just and reasonable.’ The doctrine continued to grow — Sunil Batra, Bachan Singh, Mithu, eventually Selvi v State of Karnataka in 2010 — and what Indian constitutional law produced was not a copy of American due process but a richer indigenous version of the same insight.

Almost every word of this is defensible. The Lochner fear was real; the formalism of Gopalan was costly; ADM Jabalpur was a disaster; Maneka was the right decision and it was correctly decided. The standard story does not need defending against the charge that its outcome was wrong. It needs defending against the charge that it has misdescribed how the outcome was reached, and that the misdescription matters.

Why the settled view appeals. Begin with what the settled view gets right, because there is a great deal. The Court in 1978 was not engaged in some unprincipled act of free invention. There was a textual hook — the word ‘procedure’ — and a structural hook — the relationship between Articles 14, 19 and 21 that the Gopalan majority had artificially severed. The Maneka bench did the work of stitching the silos back together, and on that work everything decent in modern Indian fundamental-rights jurisprudence has been built. To strike at the doctrine is to strike at the doctrinal infrastructure of Indian liberty as we now know it, and any honest critic of the borrowing has to hold those two facts together.

There is a second strength to the settled view, more important than the first. Common-law constitutional development does not march in step with constitutional text. American due process itself is the strongest example. Nothing in the Fourteenth Amendment, read straight, contains the rule of Roe v Wade or the libertarian streak of Lochner or the equality reasoning of Lawrence v Texas. The American doctrine is a judicial creation built on a textual scaffold that says the State shall not deprive a person of life, liberty or property without due process of law — ten words doing the work of a library. The British constitutional order produces enormous doctrinal innovation without any textual scaffold at all. To demand of Indian constitutional law that every doctrinal shift trace itself to a textual amendment is to demand of it something that no common-law constitutional culture supplies. On this view the Indian Court did what every common-law constitutional court does — it took a sparse text and made working law out of it.

I find this defence the strongest, and the rest of this essay should be read against the weight of it. It is not a foolish defence. It is the reason the borrowing has gone uncalled-out for half a century. But it is, in the end, an alibi for a more specific failure, and the failure is not in the doctrine. The failure is in the description.

What the framers actually refused, and when

The historical record is unusually clean here. In notes of September 1946 and April 1947, B.N. Rau — Constitutional Adviser to the Constituent Assembly — wrote that the American due process clause had given American courts ‘far too much discretion over legislation’ and might empower Indian courts ‘to stand in the way of beneficient social legislation’.¹ The fear was named: Lochner, Adkins v Children's Hospital, Mugler v Kansas — the line of cases in which the American doctrine had been used to invalidate maximum-hour laws and minimum-wage statutes through the construction of a ‘liberty of contract’ that nothing in the constitutional text required.⁶ Rau's misgiving predates his trip abroad. It was not imported by Frankfurter. It was confirmed by him.

In October-December 1947, Rau met Frankfurter at Harvard. Frankfurter, sitting on the Supreme Court that had only recently buried the Lochner line, told Rau that the due process clause was ‘undemocratic’ because ‘it gives a few judges the power of vetoing legislation enacted by the representatives of the nation’, and that it imposed an ‘unfair burden’ on the judiciary.² Rau wrote home to Rajendra Prasad on 11 November 1947 reporting the conversation. The Drafting Committee accepted the deletion. The phrase ‘due process of law’ vanished from the draft and was replaced with ‘procedure established by law’, borrowed from Article 31 of the Japanese Constitution — a clause that the Japanese Supreme Court was at the time reading narrowly.

None of this was a quiet bureaucratic adjustment. When the deletion came up in the Constituent Assembly on 6 December 1948, members protested vigorously. Numerous amendments were moved to put ‘due process of law’ back into the text. Kazi Syed Karimuddin warned that the new phrase would not allow courts to look into ‘the injustice of a law or into a capricious provision in a law’: ‘As soon as the procedure is complied with there will be an end to everything and the judges will be only spectators.’³ Only Alladi Krishnaswami Ayyar spoke firmly for the deletion. Even Ambedkar was, on the record, ambivalent. On 13 December 1948 the amendments to restore ‘due process of law’ failed and the deletion stood.

This is not a story about a casual omission. It is a story about a constitutional choice, made deliberately, after consultation with a sitting US Supreme Court Justice who had seen what the doctrine could do and counselled against importing it. To say that Maneka Gandhi read ‘due process’ back into Article 21 is to say that the Court of 1978 reversed the Constituent Assembly's most consequential textual decision in the fundamental-rights chapter. It is a defensible reversal. But it is a reversal, and we should be willing to call it that.

Where the settled view actually fails: the silent re-import

The first Gopalan bench understood what had happened in the Assembly because the events were a year old. Sastri J wrote that to read ‘procedure established by law’ to mean procedural due process or natural justice would import ‘those “subtle and elusive criteria” implied in [the due process] phrase which it was the deliberate purpose of the framers of our Constitution to avoid’. Das J wrote: ‘it is common knowledge that our Constitution-makers deliberately declined to adopt the uncertain and shifting American doctrine of due process of law… To try to bring in the American doctrine, in spite of this fact, will be to stultify the intention of the Constitution as expressed in Article 21.’⁴ These are not the words of judges who have stumbled across an interpretive disagreement. They are the words of judges who know exactly what the Assembly chose, and why.

Twenty-eight years later in Maneka Gandhi, Bhagwati J asked whether Article 21 required ‘that there must be some semblance of procedure, howsoever arbitrary or fanciful, prescribed by law’, or whether ‘the procedure must satisfy certain requisites in the sense that it must be fair and reasonable’.⁵ The answer was the second. The reasoning ran through Article 14 and the proposition that arbitrariness offends equality, and through the integration of Articles 14, 19 and 21 that Gopalan's silos doctrine had blocked. Nowhere in the majority opinion is there an honest reckoning with the framers' deletion. The 1948 record is not engaged. The Frankfurter-Rau exchange is not addressed. The proposition that Sastri and Das JJ said could not be done in 1950 because the framers had refused it is now done, and the refusal is treated as if it had not happened.

The pattern repeats. Sunil Batra, decided eight months after Maneka, has Krishna Iyer J writing that ‘our Constitution has no “due process” clause, but, in this branch of law, after Cooper and Maneka Gandhi, the consequence is the same.’⁷ The candour is striking and isolated. Bachan Singh tested Section 302 IPC on the touchstone of Article 21 — substantive due process by any honest description — without saying so. Mithu introduced the ‘savage sentence’ doctrine, an explicit substantive limit on legislative power to define punishment, and described it as a reading of Article 21 alone. By Selvi v State of Karnataka in 2010 the Court was using the words ‘substantive due process’ as if the doctrine had always been part of Indian law, citing Articles 14, 19 and 21 ‘in combination’ as its source.⁸ In Suresh Kumar Koushal v Naz Foundation the Court conceded openly that the ‘requirement of substantive due process’ had been ‘read into the Indian Constitution through a combined reading of Articles 14, 21 and 19’ — a sentence that, taken at face value, admits the doctrine is a judicial composite rather than a textual entitlement.¹⁰ By Re Ramlila Maidan Incident (2012) a Division Bench was writing that ‘due process of law’ is ‘contained in Article 21 of the Constitution’, full stop, without footnote or qualification.⁹ The contradiction with the framers' choice is not addressed because, by 2012, the contradiction has been forgotten.

What we have here is not doctrinal evolution. Doctrinal evolution acknowledges what it is doing. The American Court that decided Brown v Board said ‘we cannot turn the clock back to 1868… We must consider public education in the light of its full development’. The Court was honest that it was overriding original intent, and gave reasons. The Indian Court that decided Maneka overrode the framers' textual choice without saying so, and the legal academy that wrote about Maneka for the next forty years did not press the point. The borrowing was hidden in plain sight — the words ‘due process’ stayed out of the constitutional text but the substance walked in through the back door, and the Court that opened the door pretended it had always been open. The honourable exception in the academic literature is Manoj Mate, who treated the post-Maneka arc explicitly as ‘borrowing’ rather than as evolution; nearly every other treatment slips back into the language of organic doctrinal growth.¹¹

What the silence has cost comparative reasoning

So far the argument is historical. The doctrinal arc, properly described, is one of borrowing without acknowledgement. That alone is not yet a serious indictment — unacknowledged borrowing happens in every constitutional culture. The serious indictment is what the unacknowledged borrowing has done to the Indian comparative-law conversation.

Sujit Choudhry has written sharply about how the Indian Constitution looks from outside the country: a document that comparative scholars admire from a respectful distance because they cannot trace the doctrines back to anything in the text they recognise. American constitutional scholars looking at our ‘substantive due process’ line do not know what to do with it. The phrase is American. The doctrine, in its Indian application, is not the American doctrine — it is a freer hand operating on a wider statutory field. But Indian textbooks describe the doctrine in language drawn straight from American constitutional theory, with American case-citations as wallpaper, while insisting that the Indian doctrine is independent. The two propositions cannot both be true. Even the standard monograph on the subject, Abhinav Chandrachud's Due Process of Law, lays out the historical gap between the framers' deletion and the Court's restoration without pressing the comparative-jurisprudence consequences hard enough to force a reckoning.¹²

Consider what an honest comparative description would look like. It would say: India refused American due process at the framing stage. The Court reversed that refusal, by reading rather than by amendment, in 1978. The reading was facilitated by the structural integration of Articles 14, 19 and 21 that Gopalan had wrongly severed; the integration was a domestic move and is to that extent a genuine Indian innovation. But the substantive content that flowed through the integration — the ‘fair, just and reasonable’ standard, the substantive scrutiny of penal statutes, the unenumerated rights doctrine — was American in pedigree, mediated through the Court's own reading of US case law, and it brought with it the specific doctrinal pathologies that the framers had explicitly tried to avoid.

That description costs nothing in pride and gains a great deal in clarity. It tells a comparative scholar exactly what to compare. It exposes which parts of Indian doctrine are answers to Indian problems and which are inheritances from American debates that the Indian Court accepted unargued. It allows us, finally, to ask the question the standard story makes it impossible to ask: which American debates have we inherited along with the doctrine, and have we noticed?

We have not noticed several. The American substantive-due-process tradition carries with it, by its own internal logic, a recurring tension between economic substantive due process (the Lochner line, formally dead) and personal-liberty substantive due process (the Roe-Lawrence-Obergefell line, formally alive). The Indian Court has helped itself to the second without ever facing the first. Lochner was the spectre that drove the framers' deletion. The deletion was meant to ensure that an Indian court could never strike down a maximum-hours law on the ground that it interfered with someone's liberty of contract. The Court that decided Maneka was right not to revive that fear: in 1978 the Indian Supreme Court was not going to invalidate land reforms or labour statutes under Article 21. But the doctrinal tools the Court took on were the same tools that, in different judicial hands and a different political moment, could do exactly that. There is nothing in the Indian doctrine — as written — that internally distinguishes ‘good’ substantive due process from ‘bad’. The distinction has been kept by judicial culture, not by doctrine. We have a Lochnerian instrument we have promised ourselves we will not use Lochnerianly. The promise is not enforceable.

The alternative: a doctrine that knows what it is

The point of this essay is not to call for the deletion of substantive due process from Indian constitutional law. It cannot be deleted; the cases are there, the doctrines are stitched into adjacent fields, the litigation patterns are settled. The point is to insist that the doctrine be described correctly, because correct description is a precondition for the principled limits the doctrine needs.

An honest description does three things. First, it concedes that the Indian doctrine of ‘fair, just and reasonable’ is, in its substantive operation, the American doctrine of substantive due process applied to a different statutory landscape. Second, it acknowledges that this importation reverses a deliberate framing-stage choice and that the reversal was effected by judicial reading rather than by constitutional amendment. Third, it accepts the duty that follows from the second: the Court that read in the doctrine owes the constitutional order an articulable, principled account of why the doctrine should reach personal-liberty questions but not, say, the economic-policy questions the framers were specifically trying to keep out of judicial reach.

The third duty is the one the Court has consistently dodged. Indian doctrine has no internal limit on what counts as a permissible substantive-due-process inquiry. Selvi uses substantive due process to scrutinise narco-analysis. The Naz Foundation litigation used it to scrutinise sodomy laws. Mithu used it to scrutinise mandatory death sentences. Saroj Rani used it to uphold the restitution-of-conjugal-rights provision against substantive challenge. There is no doctrinal rule that distinguishes the cases the Court takes from the cases the Court refuses. There is only judicial culture, and judicial culture is not law.

An Indian doctrine that knew what it was would either limit substantive review of legislation to enumerated rights and the structural prohibitions of the Constitution — a Hart-style narrow reading — or it would adopt and openly defend a tier-of-scrutiny framework like the American one, distinguishing strict scrutiny for fundamental personal liberties from rational-basis review for economic and social legislation. Either move would put the doctrine on a foundation it currently lacks. Either move would force an honest engagement with the framers' deletion. Either move would require the Court to say in public what it has, since 1978, said only in private through the disposition of cases.

The strongest defence, addressed

The objection runs as follows. Common-law constitutional doctrine never tracks textual change. To require an amendment for every doctrinal shift would freeze constitutional development. The Maneka Court was facing the consequences of Gopalan's silos formalism, in which Articles 14, 19 and 21 were forced to operate as airtight chambers and a law that violated none of them on a narrow reading violated none of them at all. There was no other interpretive route to a working fundamental-rights doctrine. The Court took the only available route. The fact that the route led to a destination the framers had wanted to avoid is unfortunate, but the alternative — to leave Indian constitutional law in the Gopalan straitjacket pending an amendment that was never going to come — was worse.

Every clause of this objection has weight. The integration of Articles 14, 19 and 21 was necessary; the Gopalan silos were untenable; an amendment was politically impossible. The Court did what it had to do. I accept all of this. The objection lands against an essay this is not.

This essay is not arguing that Maneka was wrongly decided. It is arguing that Maneka, having been decided, owed the constitutional order an honest account of what it had done. ‘We are reading back into Article 21 a doctrine the framers explicitly removed, because the framers' alternative — leaving the field to a strict procedural reading — has proved unworkable in the face of Emergency-era abuses; here is the account of why we are doing it and the principled limits we accept on what we have done’ would have been a defensible judgment. ‘We are reading Article 21 to require fair, just and reasonable procedure, full stop’, with no acknowledgement of the framers' refusal and no internal limits on how far the reading can travel, was the easier judgment. The easier judgment is what we got. Forty years on, we are paying for the missing paragraphs.

The objection's deepest version says: an honest account would have been politically impossible in 1978, because the Court was rebuilding its legitimacy after the Emergency and could not afford to advertise itself as the framers' opponent. There is something to this. But the price of the silence is not historical — it is current. Every Indian Supreme Court judgment that uses substantive due process today inherits the unargued foundation. The judges of 2026 have no political reason to maintain the silence the judges of 1978 had institutional reasons to maintain. They maintain it because the academy has not pressed the point.

What changes if the essay is right

If the essay is right, three things change in how Indian constitutional law is taught, written about and litigated.

The Constituent Assembly Debates of 6 and 13 December 1948 stop being a curiosity in the footnotes of casebooks and become a live source. The deletion of ‘due process of law’ — with the speeches against the deletion preserved in full — is the most consequential drafting decision in the fundamental-rights chapter, and it has been treated as if it were a footnote because the doctrine that grew up around Article 21 made it embarrassing to remember. An honest curriculum teaches the deletion alongside Maneka, not as a quaint backdrop to it.

Comparative scholarship gains a category it currently lacks: the unacknowledged transplant. Indian constitutional law is full of these — due process is the largest, but the equality jurisprudence borrows from American equal-protection law in similar ways, and the basic-structure doctrine borrows German basic-law thinking with no more candour. A scholarship that can describe these as transplants, rather than as evolutions of the indigenous text, is a scholarship that can engage with comparative interlocutors as equals rather than as defenders of a story whose structure those interlocutors do not share.

Litigation gains an honest doctrinal vocabulary. A petitioner challenging a penal statute under Article 21 today must use the language of ‘fair, just and reasonable’ and pretend that the words are doing analytic work. They are not doing analytic work. They are a placeholder for substantive due process, and the inquiry the petitioner is actually inviting the Court to undertake is the inquiry an American court would undertake under the Fifth or Fourteenth Amendment. Calling the inquiry by its proper name would force petitioners to articulate which strand of substantive due process they are invoking — personal liberty, equal treatment, anti-arbitrariness — and would force the Court to be principled about which strands it accepts and on what terms.

None of this requires a constitutional amendment. None of it requires the Court to abandon any case it has decided. It requires a paragraph, somewhere in the next major Article 21 judgment, that says: in 1948 the framers refused American due process; in 1978 the Court read its substance back into Article 21; the reasons for the reading are X, Y and Z; the principled limits we accept on what we have read in are A, B and C. That paragraph would do more for Indian constitutional law than another decade of substantive-due-process jurisprudence operating on an unargued foundation. The borrowing happened. The pretence that it did not has lasted long enough.

Get the week's biggest decisions, distilled.

A weekly digest for lawyers, students and the legally curious. Every Friday morning. Free.