The basic-structure doctrine has outgrown its job

The basic-structure doctrine is the constitutional achievement most often celebrated by the Indian legal profession, and it is celebrated for a reason: in 1973 it stopped a Parliament that had begun to amend the Constitution at will. I have no quarrel with what Kesavananda Bharati did in 1973. I have a quarrel with what the doctrine has been turned into since. Conceived as a check on Article 368, the doctrine now polices ordinary legislation, executive action, and the Court's own structural reforms. Each step of that expansion was justified in its own moment. The cumulative result is a doctrine that has no internal limits, no textual anchor, and no honest answer to the question its founders most wanted answered: who guards the guardian.

What Kesavananda actually decided, and why the question matters now

It is worth being precise about the holding, because almost every contemporary use of the doctrine drifts from it. The thirteen-judge bench in Kesavananda Bharati did one thing: it held that Article 368, the constitutional-amendment power, does not authorise Parliament to abrogate the basic structure of the Constitution.¹ The Court's target was specific. The Twenty-fourth Amendment had been enacted to neutralise Golak Nath; the Twenty-fifth, to dilute the right to property and to immunise certain laws from Part III review. Six of the seven judges in the majority found implied limitations in the word "amendment" itself. Khanna J., whose opinion broke the tie, located the limit in the meaning of the constituent power Parliament was exercising under Article 368: an amendment, he said, contemplates a Constitution that survives the change without loss of identity. To destroy the identity is to abrogate.

That is a holding about the amendment power. It is not, on its face, a holding about ordinary legislation, about executive action, about the appointment of judges, about the way the Centre talks to the States, about the autonomy of tribunals, about the boundary between political questions and justiciable ones. The judgment was authored by a Court that had been picking a fight with Parliament for nearly a decade; it was a doctrine narrowly tailored to that fight. The holding, narrowly read, says: Parliament may amend the Constitution; it may not replace it. Khanna J. was clear that the change to the Constitution must leave the document recognisably itself; he was equally clear that subject to that limit, the amending power is plenary, including over fundamental rights. The line drawn in 1973 was a hard one, and it was drawn in a particular place.

Madhav Khosla, in his Oxford Handbook chapter on amendment, makes the conceptual point: the basic-structure doctrine emerged from a worry about the location of sovereignty — who gets to act as the constituent power, and on what conditions.⁹ That is a question about the amendment power. It is not a question about every contested matter on which the Court might want a final word. The doctrine has answers for the first; it has invented answers for the second. Khosla notes that the doctrinal payoff of basic-structure review of statutes — over and above the rights and federalism analysis already available under Articles 14, 19, 21, and the Seventh Schedule — has never been clearly stated. The honest answer is that the payoff is rhetorical: a basic-structure label adds nothing legally that the underlying provisions do not already supply, but it adds a great deal to the political register of the Court's decision.

I am not writing this as an academic complaint. The doctrine governs the constitutional argument every senior counsel makes in every important matter. If a litigant cannot win on the text of a provision or on the established rights jurisprudence, basic-structure rhetoric becomes the next move. Tribunal autonomy, electoral freedom, the secularism of a particular government policy, the fiscal relationship between Centre and State on a given dispute, the design of a statute regulating online speech, the manner in which a Speaker rules under the Tenth Schedule, the architecture of judicial appointments through which the Court itself is constituted — all of these are, today, basic-structure arguments. Each of them was, in the Constitution as drafted, a question for some other branch, or for the rights chapter, or for both.

The cost of this expansion is not abstract. When everything contested is a basic-structure question, two things happen. The Court becomes the routine forum for political disagreement. And the political branches stop developing the institutional culture that makes constitutional self-restraint possible, because there is no reason to develop it; the Court will do the work. We have a constitutional system in which the political class has learned that constitutionality is what the Court says it is on a given Tuesday — and a Court that has learned, in turn, that there is no constitutional question it lacks the equipment to answer. That is a quiet but real shift in how Indian constitutionalism actually functions, and it has happened without a single deliberate decision to bring it about.

The expansion in three moves

The doctrine's outward migration came in three steps, none of which were planned as a programme. Each was a sensible response to the matter then in front of the Court. Together they remade the doctrine into something its founders did not build.

The first move was to amendments unconnected with the structure of the Constitution. Indira Nehru Gandhi v. Raj Narain (1975) struck down clause (4) of the newly inserted Article 329A, which had retroactively validated the Prime Minister's election and shielded it from challenge.² The amendment was self-serving in the most obvious sense, and the Court was right to refuse to bow to it. But the doctrinal move was significant in a way the political optics obscured. The Thirty-ninth Amendment was not an amendment to the structure of the Constitution; it was a partisan piece of legislation dressed up as a constitutional amendment. By striking it down on basic-structure grounds, the Court established that the doctrine could be used wherever Parliament tried to change the Constitution to win a particular fight. That was a useful tool to have in 1975. It also cracked open a category — "amendments that touch basic structure" — into which almost any contested amendment could thereafter be fitted.

The second move was to make judicial review of basic structure itself basic structure. Minerva Mills (1980) struck down clauses (4) and (5) of Article 368, inserted by the Forty-second Amendment to put amendments beyond the courts' reach.³ The Court's reasoning has the clean elegance of a snake eating its tail: the donee of a limited power cannot use that power to convert it into an unlimited one. As a piece of doctrinal construction this is unanswerable. As a structural move it locks the doctrine in place forever. The basic-structure doctrine is no longer merely a judicial reading of Article 368; it is a feature of the Constitution that Parliament itself cannot touch. From Minerva Mills onward, the doctrine is unfalsifiable in the strict sense — there is no constitutional move available to the political branches that can undo it.

The third move was the one we should worry about most, because it was barely argued at the time it happened. By the late 1990s the doctrine had begun to migrate from challenges to constitutional amendments into challenges to ordinary statutes. L. Chandra Kumar v. Union of India (1997) is the watershed.⁴ The Administrative Tribunals Act provisions excluding the writ jurisdiction of the High Courts were struck down because, in the Court's reading, judicial review under Articles 32 and 226 is basic structure, and Parliament cannot abridge it through any law. The conclusion may be right. The doctrinal route is the problem. The Court had moved from "Parliament cannot abrogate the basic structure through Article 368" to "Parliament cannot abrogate the basic structure through any law," with the latter now functioning as a freestanding ground of unconstitutionality. The shift was justified on the unobjectionable ground that any other rule would let Parliament do indirectly what it could not do directly. That is true. It is also a change in kind, not in degree. The doctrine that originally lived inside Article 368 had broken out of its containing provision.

By I.R. Coelho (2007), the migration had become formal: the Ninth Schedule, that catch-all device through which Parliament had immunised hundreds of statutes from Part III review, was now subject to basic-structure scrutiny for any law inserted after 24 April 1973.⁵ The Court was unsparing about why. The provision had been used so promiscuously that the Schedule had grown from 13 entries at the moment of its insertion in 1951 to 284 by the time Coelho was decided; the Court itself called the practice rampant. The reasoning was rigorous, and on its own terms hard to argue with: if Article 31B can shield any law Parliament chooses, the limit on Article 368 is functionally undone, since Parliament can get to any constitutional outcome it wants by adding a statute to the Ninth Schedule and then adding the Schedule entry through an amendment. The Court closed that route. The practical effect, however, was to make almost any statute now justiciable on the same expansive ground that originally policed only constitutional amendments. Kuldip Nayar (2006) had tried in passing to limit the doctrine to amendments;¹² Coelho and the Coelho-style reasoning that followed it cut the other way. The line did not hold.

None of these three moves was, in isolation, indefensible. Indira Nehru Gandhi dealt with a constitutional amendment so transparently self-serving that it would have offended any serious doctrine of constitutional self-government. Minerva Mills was a defensive move against a Forty-second Amendment that had in fact tried to abolish judicial review of constitutional amendments. L. Chandra Kumar was responding to a real risk that Parliament could use tribunalisation to launder away the writ jurisdiction of the High Courts. The trouble is not in the individual cases. The trouble is in the accumulated doctrinal posture they leave behind. After these three moves, basic-structure review is no longer a narrow tool aimed at a specific class of constitutional events; it is a general theory of constitutional supremacy in which the Court is the only institution whose readings are not subject to override.

The cost: judicial supremacy without the discipline of text

The standard defence of the expansion is that the Constitution is a transformative document and that a flexible reading is the price of protecting it. I find that defence unconvincing, for two reasons.

The first is that the doctrine has no textual foundation. Article 368 says nothing about a basic structure. The list of features that count as basic structure is a judicial inventory built case by case: democracy, federalism, secularism, separation of powers, judicial review, free elections, the rule of law, the balance between Parts III and IV, and now (after the NJAC matter) the primacy of the judiciary in its own appointments. Indira Sawhney added the rule of law in the Mandal context;⁷ Kihoto Hollohan added free and fair elections;⁶ I.R. Coelho added the Part III–IV balance. Each addition was sensible in isolation. Each addition was also a transfer of the constitutional last word from the political branches to the Court. The cumulative roster is now so capacious that any plausible constitutional question can be raised under it; that is not a hidden flaw in the doctrine, it is its design.

The second is that the doctrine has no internal limits. There is no test for what counts as basic structure other than what five judges, on a particular day, decide counts. Kesavananda itself acknowledged this: the basic features could not be exhaustively enumerated, only illustratively listed, and different judges in the majority offered different illustrations. That intellectual modesty was tolerable while the doctrine policed only Article 368, because the targets of basic-structure review were rare and visible. It is no longer tolerable. When the doctrine is used to strike down ordinary statutes, when it is used to evaluate the Court's own structural reforms, the absence of an internal limit becomes the defining feature, not a technical lacuna.

D.D. Basu's commentary on Article 368 captures the formal position the Court has settled into: Parliament holds constituent power, but cannot, in the exercise of a limited amending power, convert that limited power into an unlimited one.¹¹ That is true of Parliament. It is also true of the Court. A Court that holds the only key to what counts as basic structure, and that admits no test other than its own conscience, has done exactly what it forbade Parliament from doing — converted a limited reviewing power into an unlimited one.

The NJAC ruling and the doctrine turning inward

The clearest illustration of the doctrine's reach is the 2015 ruling on the National Judicial Appointments Commission. The Ninety-ninth Amendment to the Constitution and the NJAC Act had been passed with overwhelming parliamentary majorities, ratified by the required number of State legislatures, and came after years of complaint — much of it from the legal profession itself — about the opacity of the collegium. A five-judge bench struck both down on the ground that the Commission's composition impaired the primacy of the judiciary in appointments and therefore violated judicial independence as a basic feature.⁸ The collegium, an extra-textual judicial creation of the Three Judges Cases, was restored.

I take no position here on whether the NJAC was a good idea or a bad one. The relevant fact is structural. The Ninety-ninth Amendment was the most procedurally pristine constitutional change in memory. It was passed by Parliament with the supermajority Article 368 demands, ratified by the States, and aimed at a specific institutional defect that had been the subject of public criticism for two decades. It was struck down because the Court read judicial primacy in appointments — itself an interpretive gloss the Court had given to the original constitutional text — into the basic structure. As a matter of constitutional grammar, the Court was not just policing Parliament. It was policing Parliament's attempt to revise the Court's own earlier reading of the Constitution. That is the moment the basic-structure doctrine completed its arc: it moved from a tool by which the Court checked the legislature to a tool by which the Court protected its own institutional preferences against the legislature's revision.

Robinson's account of the Court as a polyvocal institution, sitting routinely in two-judge benches and producing sometimes-contradictory precedent, makes the picture less reassuring still.¹⁰ The doctrine's content is not authored by a single deliberative bench. It accretes through panels of varying composition and seniority. There is no single body in the Court that speaks for the doctrine. The doctrine is what the relevant five-judge bench, on the relevant day, says it is. Robinson notes that constitutional benches of five or more judges, which were once heard at the rate of about a hundred a year, dipped to roughly ten a year after the Emergency and have not recovered. Most of the Court's day-to-day work is now done in two-judge benches whose pronouncements on basic-structure matters bind nobody on a future bench but do shape the rhetorical climate in which the next basic-structure case arrives.

The NJAC ruling is also where the institutional defence of the doctrine becomes hardest to maintain on its own terms. The collegium system, restored as the price of striking the Ninety-ninth Amendment down, is itself an extra-textual judicial creation, with no anchor in Article 124 or 217 as drafted. It came into being through the Three Judges Cases as a defensive move against executive interference, and like the basic-structure doctrine, it has expanded with no internal limits since. To strike down a parliamentary attempt to revise that judicial creation, on the ground that the judicial creation is now basic structure, is to convert a contestable interpretive choice into an unalterable feature of the Constitution. The doctrine, in this configuration, does not protect the Constitution against Parliament. It protects the Court's prior interpretation of the Constitution against Parliament. Those are different things, and the difference is precisely what the basic-structure doctrine, in its narrow 1973 form, was meant to keep clear.

The case for the doctrine, addressed honestly

I owe the reader the strongest version of the position I am arguing against, because the position has real force.

It runs like this. India does not have a strong second chamber. The Rajya Sabha is structured to mirror the Lok Sabha's politics rather than to push back on it; the Vice-President's office has no independent political base; the President's reserve powers are conventionally exercised on ministerial advice. India does not have a culture of executive self-restraint. Article 356 has been used promiscuously; ordinances have been issued in lieu of legislation as a matter of routine; the writ of the Speaker has been bent to majority interest. India does not have the embedded political checks that allow other constitutional orders to manage with a thinner judicial review. The federalism is asymmetric and the financial commands of the Union are heavy. In an environment of this shape, the argument runs, an expansive basic-structure doctrine is not a luxury. It is the structural reality that prevents constitutional decay. The Court cannot be a thin-bladed reviewer of amendments when no other check on the political branches actually bites.

The empirical case here is stronger than its critics admit. The Thirty-ninth Amendment shows what an unrestrained Parliament will do when the political incentive is present. The Emergency shows what an unrestrained executive will do. The Forty-second Amendment shows what a Parliament with a working two-thirds majority and a compliant judicial mood will do. A Court that lacks the doctrinal tool to push back on those moments is a Court that has failed in the function for which it exists. The institutional defence is not a fig leaf for judicial vanity. It is a serious account of what an Indian constitutional order, in its present condition, requires.

I take the defence seriously. I think it proves too much. If the basic-structure doctrine exists because no other check works, then we have not designed a constitutional order — we have appointed a guardian. A guardian, however wise, is not a structure. The Constitution as drafted contemplated mutual checks: Parliament checking the executive through the confidence vote and budgetary control; the executive checking Parliament through the President's assent and reserve powers; the judiciary checking both through ordinary review of statutes and executive action. The basic-structure doctrine, in its expanded form, has gradually taken over the work of the first two. We now have a system in which Parliament is checked by the Court, the executive is checked by the Court, and the Court is checked by the Court. The institutional defence accepts this as a permanent settlement. I do not. A Constitution in which one institution holds the only meaningful check on every other one is not a constitutional order. It is a delegation of constitutional self-government to that one institution, conducted in the language of doctrine.

There is a second problem with the institutional defence, less often stated. The defence assumes that the Court's expansive role is the cause of constitutional stability, when it is at least equally plausible that the relationship runs the other way. The political branches have not built the culture of self-restraint the Constitution presupposed because they do not have to: the Court will catch the worst excesses, and so the political class has no incentive to police itself. If we want a constitutional order in which the political branches act with restraint, we need the political branches to face the consequences of acting without it. A Court that absorbs every shock the political branches throw at the Constitution is a Court that has insulated those branches from the discipline of constitutional accountability. The institutional defence describes a steady state. It does not describe a healthy one.

What an honest reading would do

An honest reading of the basic-structure doctrine would do three things.

It would limit the doctrine to constitutional amendments, where its founders meant it to live. L. Chandra Kumar's line — that Parliament cannot abrogate basic structure through any law — would be retained as a dictum, not as a doctrine. Statutes that violate the rights chapter or the federal scheme would be struck down on the rights or federalism analysis they violate, not under a freestanding basic-structure head. This is not a loss in protection; the rights chapter and the federal scheme do most of the work the doctrine claims to do, and they do it with the discipline of textual anchoring.

It would require the Court, when it does deploy the doctrine on amendments, to articulate in advance the specific element of the Constitution it claims is basic, the textual or structural ground on which the claim rests, and the limits of the principle being applied. The current practice — discovering basic features case by case, from a list that grows as the docket grows — is a practice no other constitutional doctrine is permitted. We do not let due process expand by accretion in this way; we should not let basic-structure review do so.

It would acknowledge that some of the questions the Court answers under the doctrine today are questions on which the Constitution has assigned the answer to Parliament. The judiciary's primacy in its own appointments was not an assignment the Constitution made; the Court read it in. Parliament's attempt to revise that reading was the kind of constitutional revision Article 368 exists to enable. To strike that revision down on the ground that the Court's own reading is now basic structure is to confuse a judicial gloss with the constitutional text it glosses. The two are not the same.

None of this is a programme for reform. The Court will not adopt it; the political class has no interest in challenging it; the legal profession has the largest stake of any constituency in the doctrine remaining as expansive as it is. What I am arguing for is something more modest. We should be willing to say, when we celebrate the basic-structure doctrine, that the doctrine we are celebrating in 2026 is not the doctrine Kesavananda built. The 1973 holding was necessary, defensible, and contained. The doctrine that has grown around it is none of those things, and the difference matters.

Where this leaves us

I do not think the basic-structure doctrine should be discarded. The narrow holding of Kesavananda — that Parliament may amend the Constitution but may not abrogate it — is essential, and it has demonstrably done the work it was built to do. The Forty-second Amendment was the test case the doctrine was made for, and the doctrine met the test. What has happened since is different. The doctrine has been turned into a general-purpose theory of Indian constitutional law, capable of policing any state action, including the Court's own structural reforms. That generalisation is not justified by the original holding, and it is not contained by anything internal to the doctrine. It is held in place only by the Court's continued willingness to exercise it with restraint, which is a thin reed on which to rest the basic distribution of constitutional authority in a country of India's size.

What I want, plainly: a Court that retains Kesavananda's holding, applies it to the cases Kesavananda was about, and reverts to ordinary constitutional analysis for everything else. The basic-structure doctrine was built to police constitutional amendments. It should go back to doing that, and only that. The expansion has been long enough; the cost of carrying it further is no longer worth what it was when the doctrine was new. The 1973 holding is a real constitutional achievement, and it deserves to be defended in the form it was actually delivered. The expansionary doctrine that bears its name today is a different proposition, and we should be willing to retire it without pretending that doing so retires the achievement.

Get the week's biggest decisions, distilled.

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