The basic structure doctrine is the most consequential piece of Indian constitutional law that appears nowhere in the Constitution. Nothing in Article 368 says that the amending power is limited by substance. Nothing in Part III says that some rights are more fundamental than others. The words basic structure do not occur in the constitutional text at all. Yet the doctrine has invalidated constitutional amendments, restructured the relationship between Parliament and the judiciary, and become the argument of first resort whenever a government proposes a change that its opponents consider existential.

That gap between text and doctrine is why the subject is argued about with such heat, and why so much of what is said about it is wrong. The single most repeated claim, that Kesavananda Bharati held fundamental rights cannot be amended, is the opposite of what the case decided. The second most repeated claim, that there is a definitive list of basic features, describes something the courts have consistently refused to produce. This article treats the doctrine as what it is in practice, a test that a court applies to a specific amendment on specific facts, and traces how that test was built decision by decision.

The procedural mechanics of amendment, the three routes, the arithmetic of the special majority, and the ratification proviso, are set out separately in this series’ guide to how the Constitution of India is amended, and this article assumes them. What is analysed here is the substantive limit that the courts constructed on top of that procedure, and the twenty-two years of litigation that produced it.

The ambiguity the constitutional text left open

The problem begins with two provisions that were drafted to do different jobs and were never reconciled with each other.

Article 13 provides that the state shall not make any law which takes away or abridges the rights conferred by Part III, and that a law made in contravention is void to the extent of the contravention. It also supplies a definition of law that is inclusive rather than exhaustive, listing ordinances, orders, bye-laws, rules, regulations and notifications.

Article 368, as originally enacted, set out a procedure for amendment. It said how a bill is introduced, what majority is required, when state ratification is needed, and that the Constitution shall stand amended once the bill receives assent. It said nothing about whether there is anything Parliament may not amend.

Put those two provisions side by side and a question presents itself immediately. If Parliament passes a constitutional amendment that abridges a fundamental right, has it made a law within the meaning of Article 13? If yes, the amendment is void, and the fundamental rights are effectively unamendable. If no, the amending power is subject to no substantive limit at all, and every guarantee in Part III exists at the sufferance of a two-thirds majority.

Neither answer is comfortable, and the framers had not chosen between them. That is not a criticism of the drafting; the question is genuinely hard, and the Constituent Assembly had reason to think it would not arise urgently. It arose within sixteen months, when the Provisional Parliament amended the free speech, equality and property guarantees in a single bill. The story of how that amendment came about is told in this series’ account of the First Amendment and the judgments that provoked it, and its challenge is where the doctrinal sequence begins.

Why could the constitutional text not settle the question itself?

Because Article 13 and Article 368 were drafted to do different jobs and were never reconciled. One voids laws abridging fundamental rights without saying whether an amendment counts as a law. The other sets out a procedure for amendment without saying whether anything is beyond it. The Constitution simply does not choose.

The first answer: Shankari Prasad and Sajjan Singh

The challenge to the First Amendment reached the Supreme Court as Shankari Prasad Singh Deo v Union of India, decided in 1951, and the Court gave the second of the two possible answers.

It held that the word law in Article 13 means ordinary legislation made in exercise of legislative power, and does not extend to a constitutional amendment made in exercise of constituent power. The distinction is between Parliament acting as a legislature, bound by the Constitution, and Parliament acting as the amending authority, exercising a power the Constitution itself confers for the purpose of altering it. On that reading, Article 13 has nothing to say about amendments, and the amending power is unlimited in substance.

The reasoning is defensible and it is also constructive rather than declaratory, which matters for the argument that comes later. The Constitution does not use the phrase constituent power in Article 368 as originally enacted, and it does not tell the reader that amendments are excluded from the definition of law. The Court supplied both. Anyone who thinks the basic structure doctrine is illegitimate because it is not in the text has to explain why the distinction between legislative and constituent power, which is equally not in the text and which was invented to reach the opposite result, is not open to the same objection.

The question returned in 1965 in Sajjan Singh v State of Rajasthan, which concerned the Seventeenth Amendment and a further expansion of the Ninth Schedule. The majority followed Shankari Prasad. What made the decision significant was not the outcome but the doubts recorded within it. Hidayatullah and Mudholkar JJ each expressed reservations about the proposition that the fundamental rights were entirely at the mercy of the amending power, and Mudholkar J raised, in terms, the question of whether the Constitution possessed basic features that a mere amendment could not alter.

That is the earliest recognisable ancestor of the doctrine, and it appears in a concurring or separate opinion rather than in a holding. It travelled from there into argument, then into a dissent, then into a majority.

Golak Nath and the overcorrection

In I C Golak Nath v State of Punjab, decided in 1967, an eleven-judge bench reversed the position by six to five.

The majority held that a constitutional amendment is law within the meaning of Article 13, and therefore that Parliament has no power to abridge the fundamental rights at all. It coupled that holding with prospective overruling, an unusual device by which the Court declared that its decision would operate only for the future, so that the amendments already made, including the First, Fourth and Seventeenth, would remain valid and the land reform statutes already protected would not be disturbed.

Two things about Golak Nath are usually missed and both matter.

The first is that its holding was absolute rather than calibrated. It did not say that Parliament may amend the fundamental rights but not destroy them. It said Parliament may not abridge them, full stop. That is a far stronger limitation than the basic structure doctrine that replaced it, and it is the reason Golak Nath was unsustainable. A constitution in which the rights chapter is permanently frozen, immune to correction even where a right is found to be badly drafted or where circumstances change, is not a workable constitution over a long horizon, and the Court had left itself no room to distinguish a trivial amendment from a destructive one.

The second is that prospective overruling made the decision look like a political accommodation, which weakened its authority from the moment it was delivered. The Court had held the amendments unconstitutional in principle and allowed them to stand in practice. Whatever the pragmatic merits of that, it invited the response that the Court was making policy rather than deciding law.

What made the Golak Nath rule unworkable?

Its absoluteness. It barred any abridgement of fundamental rights by amendment rather than distinguishing amendment from destruction, which left no room for legitimate constitutional revision and no principle for identifying which changes actually threatened the constitutional order. Kesavananda replaced the flat bar with a calibrated test.

The government’s answer: four amendments in three years

Golak Nath produced a legislative response that is the clearest example in Indian history of the reactive amendment pattern, and understanding the response is essential to understanding what the Court did next, because Kesavananda is not a case about an abstract question. It is a case about these amendments.

The Twenty-fourth Amendment answered Golak Nath directly and comprehensively. It amended Article 13 to provide that nothing in that article applies to an amendment made under Article 368. It amended Article 368 itself to declare that Parliament may, in exercise of its constituent power, amend by way of addition, variation or repeal any provision of the Constitution. It changed the presidential assent provision so that assent to an amendment bill became mandatory rather than discretionary. And it altered the marginal note of the article to describe the power as one of amending the Constitution rather than merely a procedure for amendment.

That is not a subtle piece of drafting. It is an instruction to the judiciary, written into the Constitution, that the amending power is unlimited and that Article 13 is not available as a constraint.

The Twenty-fifth Amendment did two things. It replaced the word compensation in the property acquisition provision with the word amount, and provided that the adequacy of that amount would not be justiciable, which was directed at the bank nationalisation litigation. And it inserted Article 31C, which had two limbs. The first gave primacy to laws giving effect to the directive principles concerning the distribution of material resources and the prevention of concentration of wealth, so that such laws could not be challenged as inconsistent with the equality and freedom guarantees. The second limb provided that a law containing a declaration that it was for those purposes should not be called in question in any court on the ground that it did not give effect to that policy.

The second limb is the crucial one. It is not a substantive provision about rights at all; it is an ouster of judicial review. A legislature could immunise a statute by declaring its own purpose, and no court could examine whether the declaration was true.

The Twenty-sixth Amendment abolished the privy purses and the recognition of rulers, following a Supreme Court decision that had struck down the executive attempt to do the same thing. The Twenty-ninth Amendment placed two Kerala land reform statutes in the Ninth Schedule, and it is that amendment which gave the Kesavananda petitioner his standing.

The political backdrop matters. The government had nationalised the major commercial banks and had seen the manner of it struck down. It had attempted to derecognise the former rulers and had seen that struck down. It had been told by the Court in Golak Nath that the rights chapter was beyond amendment. It had then won a decisive general election on a redistributive platform, and it possessed the numbers to amend at will. The four amendments were the exercise of that mandate against a judiciary the government regarded as obstructive, and the judiciary understood them exactly that way.

Kesavananda Bharati: the case, not the slogan

Kesavananda Bharati v State of Kerala was decided on 24 April 1973 by a bench of thirteen judges, the largest ever constituted in the Supreme Court of India, presided over by Chief Justice S M Sikri. The hearing ran for months. The judgments run to hundreds of pages and there are eleven of them.

The petitioner was the head of a mutt in Kerala whose property was affected by state land reform legislation, and his petition began as a challenge under the religious freedom and property provisions. It became the vehicle for the constitutional question because the Kerala statutes had been placed in the Ninth Schedule by the Twenty-ninth Amendment, so the validity of the amendment had to be decided before the statutes could be examined. The Twenty-fourth and Twenty-fifth Amendments were before the Court in the same proceedings.

What the Court decided, stated as narrowly as the record permits, was this.

The Twenty-fourth Amendment was valid. Parliament does have the power to amend any provision of the Constitution, including Part III, and Golak Nath was wrong to hold otherwise. This is the half of the decision that supporters of parliamentary supremacy consistently forget, and it is a substantial victory for the government.

But that power is not unlimited. The word amend does not extend to a change that damages or destroys the basic structure or basic features of the Constitution. An amendment that does so is beyond the power conferred by Article 368, and a court may hold it invalid on that ground.

The first limb of Article 31C was upheld. The second limb, which ousted judicial review by making a legislative declaration conclusive, was struck down.

The majority for the basic structure limitation was seven to six. The margin is a single judge, and no honest account of the doctrine can suppress that.

What did Kesavananda Bharati actually hold about fundamental rights?

That Parliament can amend them. The case expressly overruled Golak Nath on this point and upheld the Twenty-fourth Amendment. What it added was a limit of a different kind: no amendment, whether it touches Part III or any other Part, may damage or destroy the basic structure of the Constitution.

The popular summary, that Kesavananda protected fundamental rights from amendment, gets the case exactly backwards. It is a summary of Golak Nath, which Kesavananda overruled. The confusion is understandable, because the practical effect of the basic structure doctrine has often been to protect rights, but the legal proposition is different and the difference matters enormously in argument. A litigant who tells a court that fundamental rights cannot be amended will lose. A litigant who says that this particular amendment damages a basic feature has stated the actual test.

The View of the Majority and the argument about what was signed

At the close of the hearing a document titled the View of the Majority was drawn up, summarising the propositions on which a majority of the bench had agreed, and it was signed by nine of the thirteen judges. It is that document, rather than any single judgment, which is conventionally treated as stating the ratio of the case.

The document has been contested ever since, and the contest is a serious one rather than a debating point.

The objection is that a bench of thirteen delivering eleven separate judgments does not produce a single ratio simply because someone drafts a summary at the end and circulates it for signature. Some of the signatories had written judgments that did not obviously support every proposition in the summary. Four judges did not sign. A document produced in that manner, the argument runs, is an administrative convenience that has been elevated into binding law, and the foundation of the most powerful doctrine in Indian constitutional law is correspondingly unstable.

The answer to that objection is partly practical and partly legal. Practically, there was no other way to extract a usable holding from eleven judgments running to hundreds of pages, and the alternative was years of litigation about what the case had decided. Legally, the doctrine’s authority no longer rests on the summary document at all. It rests on the fact that later benches, including benches of nine and larger, have applied it, refined it, and treated it as settled for decades. A doctrine that has been applied to strike down amendments, has survived every attempt to displace it, and has been accepted as settled by successive constitutional benches, does not depend for its authority on the drafting history of a summary sheet.

That answer is sound, but it should be given honestly rather than triumphantly. The doctrine acquired its foundation retrospectively, through acceptance and application, rather than possessing a clean one at the moment of its creation. That is a real feature of how it came to be, and it is why the critique deserves to be stated at full strength rather than waved away.

The supersession and the political consequence

Two days after the judgment, three judges who had been in the majority on the basic structure limitation, and who were senior to the judge appointed, were passed over for the office of Chief Justice. All three resigned.

This belongs in an article about the doctrine rather than in a political history, because it is the single clearest evidence of what was actually at stake. The government understood the decision as a defeat and responded by breaking the convention of seniority in appointing the head of the judiciary, which had been observed since the Court was constituted. The message to the bench was unmistakable.

The consequence was the opposite of the one intended. The supersession converted a seven to six decision, which might have been narrowed or distinguished by later benches, into an institutional commitment. A judiciary that has just watched three of its senior members resign over a doctrine does not abandon that doctrine. The episode also fed directly into the constitutional confrontation that followed, and into the judicial appointments litigation of later decades, where the memory of 1973 shaped the Court’s determination to control the process by which judges are chosen.

Readers who want to see how the same government’s constitutional project developed after this point should read this series’ treatment of the 42nd Amendment and what it attempted to entrench, which contains the most aggressive attempt ever made to abolish the limitation Kesavananda had created.

Indira Nehru Gandhi v Raj Narain: the doctrine used for the first time

The doctrine’s first application came in 1975, and it came in a case about an election.

The Allahabad High Court had set aside the Prime Minister’s election on grounds of electoral malpractice. While the appeal was pending in the Supreme Court, and after the Emergency had been proclaimed, Parliament enacted the Thirty-ninth Amendment. Clause 4 of that amendment provided, in substance, that the election of a person holding the office of Prime Minister or Speaker could not be called in question in any court, that the pending proceedings would abate, and that the election would be deemed valid.

The Supreme Court struck that clause down as violating the basic structure. The reasoning varied among the judges but the core of it is that free and fair elections and the judicial resolution of election disputes are part of the democratic structure the Constitution establishes, and that a constitutional amendment which decides a pending case in favour of one party, retrospectively and by name, is not an amendment at all but an exercise of judicial power by a legislature.

Three features of this decision deserve emphasis.

It demonstrated that the doctrine was operative rather than theoretical. Kesavananda had announced a limit without applying it to invalidate anything, and a limitation that is never used is not really a limitation. Within two years it was used.

It identified the kind of amendment most likely to fail the test. The clause struck down was not an ordinary policy amendment. It was an amendment that removed a specific dispute from judicial determination and decided it. The doctrine has always been at its strongest against amendments of that character, and at its weakest against amendments of general legislative policy.

And it was decided in the middle of the Emergency, by a Court under extraordinary pressure, which is a fact usually cited on the other side of the ledger given the same Court’s decision in the habeas corpus case the following year. Both things are true, and the tension between them is part of the honest institutional record.

Minerva Mills: the attempt to abolish the doctrine and its failure

The Forty-second Amendment attempted to place the amending power beyond judicial reach in the most direct manner available. Two new clauses were added to Article 368. One provided that no amendment shall be called in question in any court on any ground. The other declared that there shall be no limitation whatever on the constituent power of Parliament to amend the Constitution. The same amendment expanded Article 31C so that a law giving effect to any of the directive principles, rather than only the two originally specified, would be immune from challenge under the equality and freedom guarantees.

In Minerva Mills v Union of India, decided in 1980, the Supreme Court struck down both.

The reasoning on the Article 368 clauses is the most important sentence of doctrine after Kesavananda itself. A limited amending power is a basic feature of the Constitution. Parliament cannot use a limited power to enlarge that very power into an unlimited one. If it could, the limitation would be meaningless, because any government with the numbers could remove the limit first and then do as it pleased. The clauses therefore failed by the very test they were designed to abolish.

The reasoning on the expanded Article 31C identified a second basic feature: the balance between the fundamental rights in Part III and the directive principles in Part IV. The Court held that the Constitution is founded on that balance, that neither part is subordinate to the other, and that an amendment giving blanket primacy to the whole of Part IV over the freedom and equality guarantees destroys it.

That second holding is more contested than the first, and it should be flagged as such. The proposition that a particular equilibrium between two chapters is itself constitutionally entrenched is a strong claim, and it is the clearest instance of the doctrine being used to protect a judicial reading of the Constitution’s design rather than an identifiable institutional feature. Serious critics have argued that this is where the doctrine overreached, and the argument is not frivolous.

Waman Rao and the line drawn through the Ninth Schedule

The doctrine created an immediate practical problem that had nothing to do with the amendments it had struck down. If an amendment can be invalidated for damaging the basic structure, what happens to the Ninth Schedule, which by 1980 held a large and growing collection of statutes placed there by amendments passed over three decades?

The problem is not academic. Each addition to the schedule is made by a constitutional amendment. If those amendments are testable, then every statute in the schedule becomes challengeable, and land titles settled decades earlier on the strength of protected statutes reopen. If they are not testable, the schedule becomes a hole in the doctrine through which any statute can be passed.

In Waman Rao v Union of India, decided in 1981, the Supreme Court chose a line rather than a principle. Amendments adding statutes to the Ninth Schedule before 24 April 1973, the date Kesavananda was decided, would not be reopened. Amendments adding statutes after that date would be open to challenge on the ground that they damage the basic structure.

The choice of date is candidly pragmatic and the Court did not pretend otherwise. There is no doctrinal reason why an amendment made on 23 April 1973 should be immune while one made on 25 April is not. The justification is entirely about settled expectations: before Kesavananda nobody had any reason to think the schedule was vulnerable, transactions had been completed and land redistributed on that footing, and reopening them would produce chaos disproportionate to any constitutional benefit.

This is prospective operation of a doctrine by another route, and it is the same technique Golak Nath had used and been criticised for. The difference is that here it was doing genuinely necessary work, and the Court was open about the fact that it was drawing a line for administrative reasons rather than deriving one from principle.

I R Coelho and the test for post-1973 additions

Waman Rao settled that post-1973 additions to the Ninth Schedule were testable. It did not settle what test applied to them, and that question took a quarter of a century to answer definitively.

The difficulty was structural. Article 31B says that a listed statute shall not be deemed void on the ground that it is inconsistent with any of the fundamental rights. If a statute in the schedule is challenged, and the challenge is that the statute violates a fundamental right, Article 31B appears to answer it. Saying that the amendment which added the statute must satisfy the basic structure test does not obviously help, because inserting a statute into a schedule looks like an ordinary exercise of an established constitutional technique rather than an assault on the constitutional core.

I R Coelho v State of Tamil Nadu, decided on 11 January 2007 by a nine-judge bench headed by Chief Justice Y K Sabharwal, resolved it.

The Court held that Ninth Schedule statutes added after 24 April 1973 are open to challenge, and that the test is not whether the amendment adding them formally complied with Article 368 but whether the statute itself, examined on its own terms, violates or abrogates the fundamental rights that form part of the basic structure. In practice the examination proceeds against the guarantees of equality, freedom and life and personal liberty read together, the combination that Indian constitutional argument calls the golden triangle, and asks whether the effect of the impugned statute is to abrogate or damage those guarantees rather than merely to abridge them in some respect.

Three consequences follow and they are often stated too broadly.

The Ninth Schedule is not a nullity. Statutes added before the dividing line retain the protection Article 31B confers, and later additions are not automatically bad; they are testable.

The test is a high one. Coelho does not permit a court to strike down a scheduled statute simply because it restricts a fundamental right, since restriction is the ordinary consequence of most regulatory legislation. The question is whether the right is abrogated or its essence destroyed. That distinction is doing an enormous amount of work and is not easy to apply.

And the doctrine, having begun as a limit on amendments, now reaches through an amendment to the statute it protects. This is a substantial extension. The object of scrutiny is no longer the constitutional amendment as such but a piece of ordinary legislation that an amendment purported to immunise, which is a further step away from the text of Article 368 than Kesavananda took.

What test applies to Ninth Schedule entries made after 1973?

The statute itself is examined rather than the amendment that listed it. It is testable against the fundamental rights that form part of the basic structure, principally the guarantees of equality, freedom and life and personal liberty read together, and the question is whether it abrogates them rather than merely restricts them. Earlier entries retain their protection.

The basic structure timeline

Set out in sequence, the doctrine is visibly a construction rather than a discovery, and each decision responds to a specific problem left by the one before it.

Decision	Year	Bench and split	Question before the Court	What it settled	What it added to the doctrine
Shankari Prasad Singh Deo v Union of India	1951	Five judges, unanimous	Is a constitutional amendment a law under Article 13?	It is not; constituent power differs from legislative power	Establishes an unlimited amending power and the legislative versus constituent distinction
Sajjan Singh v State of Rajasthan	1965	Five judges, majority	Same question, on the Seventeenth Amendment	Shankari Prasad followed	Doubts recorded by Hidayatullah and Mudholkar JJ, including the first judicial suggestion of unalterable basic features
I C Golak Nath v State of Punjab	1967	Eleven judges, six to five	Same question	Reversed: an amendment is law under Article 13 and cannot abridge Part III	An absolute bar, applied prospectively, which proved unworkable
Kesavananda Bharati v State of Kerala	1973	Thirteen judges, seven to six	Validity of the Twenty-fourth, Twenty-fifth and Twenty-ninth Amendments	Parliament may amend any Part including Part III, but may not damage or destroy the basic structure	The doctrine itself; Golak Nath overruled; the second limb of Article 31C struck down
Indira Nehru Gandhi v Raj Narain	1975	Five judges	Validity of clause 4 of the Thirty-ninth Amendment	Clause struck down	First application of the doctrine to invalidate an amendment; free and fair elections and judicial determination of disputes identified as core
Minerva Mills v Union of India	1980	Five judges	Validity of the Article 368 ouster clauses and the expanded Article 31C inserted by the Forty-second Amendment	Both struck down	Limited amending power is itself a basic feature; so is the balance between Parts III and IV
Waman Rao v Union of India	1981	Five judges	Whether Ninth Schedule additions are testable	Additions after 24 April 1973 are testable; earlier ones are not reopened	Fixes the dividing line and protects settled transactions
I R Coelho v State of Tamil Nadu	2007	Nine judges, unanimous	What test applies to post-1973 Ninth Schedule additions	The statute itself is examined for abrogation of core rights	The rights test; the doctrine reaches through the amendment to the protected statute

Read as a sequence, the pattern is clear. Every decision after 1967 is a response to an attempt to circumvent the previous one. Golak Nath answered an amending power that had been used freely for sixteen years. The Twenty-fourth Amendment answered Golak Nath. Kesavananda answered the Twenty-fourth. The Forty-second Amendment answered Kesavananda. Minerva Mills answered the Forty-second. Coelho answered the Ninth Schedule route around all of it.

The open-list design

The doctrine’s most criticised feature is that it does not define what the basic structure is. No judgment has produced an exhaustive list, and the Court has repeatedly declined invitations to produce one.

Features that have been judicially recognised at one point or another include the supremacy of the Constitution, the republican and democratic form of government, the secular character of the Constitution, the separation of powers between the legislature, the executive and the judiciary, the federal character of the distribution of powers, judicial review, free and fair elections, the rule of law, the independence of the judiciary, the limited nature of the amending power, and the balance between fundamental rights and directive principles. Different judgments phrase these differently and some appear in a single opinion rather than in a holding.

The claim this article advances is that the refusal to close the list is a design feature rather than a defect, and the argument for it is short and hard to escape.

A closed list would be written into a judgment. A judgment can be answered by a constitutional amendment. The Twenty-fourth Amendment answered Golak Nath, the Thirty-ninth answered an election judgment, and the Forty-second attempted to answer Kesavananda itself. If the Supreme Court had published a definitive schedule of basic features, that schedule would have become the target, and the amending power would have been used to address each entry in turn, or to insert a provision declaring that the enumerated features are not part of the basic structure. A list is a fixed object and fixed objects can be attacked.

An open standard cannot be attacked the same way, because there is nothing specific to aim at. The cost is uncertainty, and the cost is real. A government contemplating a constitutional amendment cannot know in advance whether it will survive, which is a genuine rule-of-law objection and the strongest thing the critics have. The benefit is durability. The doctrine has survived every attempt to displace it, including a direct constitutional assault at the height of the Emergency, and it has survived precisely because there was no fixed text to repeal.

That is the trade the Court made, consciously or otherwise, and it is worth naming plainly: the doctrine purchased its survival with its own indeterminacy.

How the test is actually applied

Because the doctrine is discussed as a slogan, it is worth being concrete about what a court actually does with it, since the practical test is much narrower than the rhetoric suggests.

The first question is whether the challenge is to a constitutional amendment at all. The basic structure doctrine applies to amendments. Ordinary legislation is tested against the constitutional provisions themselves, and a statute cannot be struck down for violating the basic structure as such. Litigants advance the argument against ordinary statutes with some regularity, and it does not succeed on that footing. The exception created by Coelho is narrow and concerns statutes given constitutional immunity through the Ninth Schedule.

The second question is identification: which feature is said to be damaged, and what is the evidence that it forms part of the constitutional core? A pleading that asserts damage to the basic structure without identifying a feature is not an argument, and courts treat it accordingly.

The third question is the one that decides most cases: does the amendment damage or destroy the feature, or does it merely affect, alter, or restrict it? This is the threshold that most challenges fail. Almost every constitutional amendment affects something that could be described as a basic feature at some level of generality, because federalism, democracy and judicial review are implicated by a very large proportion of constitutional change. The doctrine is not a rule that constitutional features may not be touched. It is a rule that they may not be destroyed, and the difference between an alteration and a destruction is the whole of the argument in a real case.

The fourth question is the effect test. Courts examine what the amendment actually does rather than how it is characterised, and an amendment which is innocuous in form but which in operation removes a category of dispute from judicial determination, or which entrenches a government against electoral accountability, is far more vulnerable than one which makes a substantial change of policy openly.

The practical consequence for anyone advising on a constitutional amendment is that the risk concentrates in a small class of measures: those that oust judicial review, those that alter the process by which power is transferred or contested, those that entrench the amending authority against being checked, and those that operate retrospectively on a pending dispute. Amendments of policy, however large, are rarely struck down. Because that assessment turns on the precise terms of the instrument and on how a court characterises its effect, the general position set out here is not a substitute for advice on a specific measure.

The doctrine used against judicial appointments

The most significant modern application of the doctrine came when Parliament amended the Constitution to replace the collegium system for appointing judges to the Supreme Court and the High Courts with a National Judicial Appointments Commission, established by a constitutional amendment and given detail by an accompanying statute. The commission was to include the Chief Justice of India and two senior judges, the Union Minister of Law and Justice, and two eminent persons selected by a committee, with a mechanism permitting a small minority of members to veto a recommendation.

The Supreme Court struck down both the amendment and the statute, holding by a majority that primacy of the judiciary in the appointment of judges is an aspect of judicial independence, that judicial independence forms part of the basic structure, and that the composition and veto structure of the commission compromised that primacy.

This decision is where the argument about the doctrine becomes sharpest, and it is worth setting out honestly because it is the case that turned many previously sympathetic observers into critics.

The case for the decision is that judicial independence is on any view a basic feature, that a body in which the executive holds a seat and in which two members can block an appointment gives the executive effective influence over who judges it, and that a judiciary whose composition is controlled by the government it is supposed to check is not independent in any meaningful sense. The amendment had passed both Houses with very large majorities and had been ratified by a substantial number of state legislatures, but the doctrine exists precisely to constrain measures that command large majorities, since a measure without a large majority does not need constraining.

The case against is more uncomfortable. The system the amendment sought to replace, in which senior judges select judges, is itself a judicial construction rather than a constitutional text, built through a line of decisions that read the requirement of consultation with the Chief Justice as conferring primacy. When the Court struck down the amendment, it protected a mechanism of its own creation, using a doctrine of its own creation, against a change made by the elected branches through the most demanding procedure the Constitution provides. Whatever the merits, the structural objection that the judiciary was the judge in its own cause is not answerable by pointing to the importance of judicial independence, because that is precisely what was in dispute.

Both positions are held by serious people. This article does not resolve the argument, because it is finally an argument about who should be trusted with the composition of the judiciary, which is a question of political value rather than of constitutional record. What the record does show is that the doctrine is not a dead letter, that it applies to amendments passed by large cross-party majorities, and that its most consequential recent application concerned the institution that applies it.

Where the doctrine is weakest: federalism and policy

The doctrine’s reach is often overstated because its successes are memorable and its failures are invisible. In practice it has been applied sparingly, and there are whole categories of amendment where it has never bitten.

Federalism is the clearest example. The federal character of the Constitution is routinely listed among the recognised basic features, and constitutional amendments have repeatedly altered the balance between the Union and the states, transferring legislative subjects, restructuring taxation, and reorganising the relationship between the two levels of government. None of these has been struck down for damaging federalism. The reason is the threshold: an amendment that shifts a subject from one legislative list to another alters the federal arrangement without destroying it, and the doctrine addresses destruction rather than alteration.

The same is true of policy amendments generally. Amendments that have redistributed property rights, altered the reservation framework, restructured local government, and created new institutions have all been upheld or have gone unchallenged. Where the doctrine has succeeded, the amendments under challenge shared a family resemblance: they excluded courts, decided pending disputes, entrenched an incumbent, or removed the limit on the amending power itself.

That pattern is the best predictor available of how a future challenge will fare, and it is more useful than any list of basic features. An amendment that changes what the state may do is unlikely to fail. An amendment that changes who may review what the state does is where the doctrine lives.

Reservation is the field where these questions have been most persistently argued, because successive amendments have altered the constitutional footing of affirmative action while the courts have developed a body of doctrine about its limits. The interaction between the two, and the question of whether a ceiling on reservation is itself a constitutional requirement or a judicially set proportion, is treated in this series’ account of Indra Sawhney and the fifty percent ceiling, which is where the argument properly belongs.

What remains unsettled

Four questions are genuinely open, and describing them structurally is more honest than pretending the doctrine is fully worked out.

The first is whether the doctrine applies to state legislation or to ordinary central legislation at all. The settled position is that it does not, and that ordinary statutes are tested against constitutional provisions rather than against the basic structure. Arguments to the contrary continue to be advanced and continue to be rejected, but the Coelho extension, which permits a scheduled statute to be examined for abrogation of core rights, sits somewhat awkwardly with the clean version of the rule, and the boundary is not perfectly crisp.

The second is the status of the Preamble. The question recurs constantly, and it is often asked in a confused form. The Preamble has been held amendable, and it has in fact been amended. It has also been treated as part of the Constitution and as evidence of the objectives against which the constitutional scheme is read, and several of the recognised basic features are drawn from ideas the Preamble expresses. The two propositions coexist: the Preamble can be amended, and an amendment which used the Preamble to destroy a feature the Preamble describes would face the same test as any other amendment.

The third is how the doctrine interacts with constitutional provisions that are themselves in tension, where protecting one recognised feature requires diminishing another. Federalism and the integrity of the union, judicial review and parliamentary sovereignty over its own procedures, secularism and religious freedom in the personal law field: each pairing has produced litigation in which the doctrine could be invoked on both sides, and there is no meta-rule for ranking basic features against one another.

The fourth is the extent to which the doctrine has migrated into contexts it was not built for. Arguments framed in basic structure terms are made against ordinary statutes, executive action, and state legislation with some frequency, and courts have occasionally used the vocabulary in those contexts even while denying that the doctrine applies. That drift is worth watching, because a doctrine that means everything eventually constrains nothing.

Does the basic structure doctrine apply to ordinary laws?

No. It is a limit on the power to amend the Constitution, and an ordinary statute is tested against the constitutional provisions themselves rather than against the basic structure. The narrow exception is a statute placed in the Ninth Schedule after 24 April 1973, which Coelho permits a court to examine for abrogation of core fundamental rights.

The case against the doctrine, stated at full strength

The criticism of the basic structure doctrine is frequently dismissed as the complaint of governments that lost, which is intellectually lazy. The serious version of the case has five limbs and each of them is difficult to answer.

The textual objection comes first. The Constitution confers a power to amend and specifies the procedure for its exercise. It does not say that the power is substantively limited, and it does not identify anything as unamendable, although the framers plainly knew how to entrench provisions when they wished, since the ratification proviso demonstrates that they distinguished between categories of provision and imposed different requirements. Where a constitution wishes to place matters beyond amendment, it can and often does say so in terms. The Indian Constitution does not. A court that reads in a limitation the framers considered and declined to write has not interpreted the text; it has supplemented it.

The democratic objection follows. A constitutional amendment in India requires a majority of the total membership of each House and two-thirds of those present and voting, and for a substantial category of provisions it additionally requires ratification by half the state legislatures. That is the most demanding procedure the Indian system contains, more demanding than the process for passing a statute, changing a government, or declaring an emergency. When a court sets aside a measure that has cleared it, an unelected body of judges is overriding the most broadly assembled expression of political will the Constitution provides for. The countermajoritarian difficulty exists in ordinary judicial review too, but there it is answered by saying that the court is enforcing the Constitution against a mere legislature. Here the court is enforcing something against the constituent authority itself, which is a different and much stronger claim.

The indeterminacy objection is the most practically damaging. A limitation whose content is not specified in advance, which the Court has expressly declined to enumerate, and which is applied by asking whether a feature has been damaged or merely altered, does not give a government contemplating constitutional change any way to know whether its measure is lawful. Legal certainty is not a technicality; it is a component of the rule of law that the doctrine itself claims to protect. A doctrine that protects the rule of law by being unruly is in an awkward position.

The foundational objection concerns how the doctrine came into being. It rests on a seven to six decision from a bench of thirteen delivering eleven separate judgments, whose ratio is conventionally taken from a summary document that four judges did not sign and whose accuracy has been questioned. Doctrines of comparable importance in other systems have cleaner origins. This one does not, and the fact that it has been applied for decades since answers the practical question of its authority without answering the question of whether its foundation was sound.

The institutional objection is the sharpest and it arises from the appointments litigation. When the doctrine is used to strike down an amendment concerning the composition of the judiciary, the judiciary is deciding a question in which it has a direct institutional interest, using a limitation it created, to protect an arrangement it had itself constructed. No principle of adjudication tolerates that combination in any other setting. The answer that there is nobody else who could decide the question is true and is not a complete answer, because it establishes only that the problem is unavoidable rather than that it is not a problem.

The case for the doctrine, stated at full strength

The defence is not a rebuttal of those five points so much as an argument that the alternative is worse, and it has four limbs.

The first is the self-destruction argument, and it is the strongest. If the amending power is unlimited, then Parliament with a two-thirds majority may abolish elections, dissolve the federal structure, remove judicial review, extend its own term indefinitely, or replace the Constitution entirely, and do all of it lawfully. Every one of those is an amendment on the unlimited view. Constitutions that are silent about substantive limits are not silent because their framers wished to permit their own destruction; they are silent because framers do not usually contemplate it. The Indian framers had watched constitutional orders in Europe dismantled through lawful procedures within living memory, and the argument that they intended to leave that route open is implausible.

The second is that the abstract worry became concrete. Within twenty-five years of the Constitution coming into force, a government used the amending power to place its own election beyond judicial determination, to declare that the amending power was unlimited, and to bar any court from examining an amendment on any ground. These are not hypothetical scenarios advanced by anxious theorists; they are things that were actually attempted and that a doctrine actually stopped. The strongest evidence for the doctrine is the record of what was tried while it existed.

The third answers the democratic objection on its own ground. A supermajority in Parliament is not the same thing as the constituent power of the people. The Constitution was framed by a body constituted for the purpose, and Article 368 confers on Parliament a power to amend that Constitution, not a power to replace it with another. On this reading the doctrine does not override the constituent power; it distinguishes between the constituent power, which lies outside Parliament, and the derived amending power, which is what Parliament actually holds. That distinction is contestable, but it is not incoherent, and it is the reasoning Minerva Mills adopted when it held that a limited power cannot be used to make itself unlimited.

The fourth is comparative and modest. Every constitutional system with a written constitution and a rights chapter has to answer the question of what stops a determined majority from dismantling it, and the available answers are few: an unamendable clause written into the text, an extraordinary procedure such as a referendum or a constitutional convention, or a judicially enforced substantive limit. India’s text supplies neither of the first two. The judicially enforced limit is what was left.

Naming the disagreement precisely

The argument about the basic structure doctrine is usually conducted as though it were a dispute about constitutional interpretation. It is not, or not mainly. It is a dispute about risk allocation, and stating it that way makes both positions intelligible.

Every constitutional system faces two risks. The first is that an elected majority uses lawful procedures to entrench itself and dismantle the constraints on it. The second is that an unelected judiciary uses an indeterminate doctrine to block changes that a large and legitimate majority genuinely wants. The two risks cannot both be minimised, because the institutional arrangements that reduce one increase the other.

Supporters of the doctrine think the first risk is graver, and they point to the Emergency-era amendments as proof that it is real in India specifically rather than merely conceivable. Critics think the second risk is graver, and they point to the appointments litigation as proof that judicial self-protection is real rather than merely conceivable. Both are pointing at genuine episodes. Neither is being unreasonable.

What follows from this is a discipline for reading anything written about the doctrine. An account that treats the critics as apologists for authoritarian government is not engaging. An account that treats the defenders as naive about judicial power is not engaging either. The evidence on both sides is real, the choice between the risks is a choice of value, and the honest position is to say which risk you weigh more heavily and why, rather than to pretend that the constitutional text settles it.

Four misconceptions worth clearing

The first has already been stated and is worth repeating because it is so pervasive: Kesavananda did not hold that fundamental rights cannot be amended. It held the opposite, upheld the amendment that reversed Golak Nath, and substituted a test that applies to every Part of the Constitution equally. A reader who takes away only one correction from this article should take away this one.

The second is the confusion of the doctrine with judicial review generally. Judicial review is the power of a court to test state action against the Constitution, and it existed in India from the day the Constitution came into force. The basic structure doctrine is a limit on constitutional amendment specifically. When a court strikes down a statute or an executive order, it is exercising ordinary judicial review, not applying the basic structure doctrine, and describing every judicial invalidation as an application of the doctrine drains the term of meaning.

The third is the belief that there is an authoritative numbered list of basic features. There is not. Various judgments have identified features, sometimes in a holding and sometimes in a separate opinion, and study material commonly compiles these into a list of a convenient length. Treating such a compilation as authoritative is a mistake in an examination answer and a serious mistake in a pleading, because the compilation has no source and no court has adopted it.

The fourth concerns the Preamble, which generates a perennial question in a form that conceals the real issue. The Preamble is part of the Constitution and can be amended, and it has been. Several recognised basic features draw on values the Preamble articulates. Neither of those propositions contradicts the other, because the doctrine does not protect provisions as such; it protects features, and a feature may be reflected in several provisions or in none of them exclusively. The correct answer to whether the Preamble can be amended is that it can, subject to the same limit that applies to every other amendment.

What the doctrine has actually done

A useful corrective to the rhetoric on both sides is to look at how often the doctrine has been used to invalidate a constitutional amendment. The answer is: rarely.

The amendments that have been struck down or partly struck down on this ground share the family resemblance noted earlier. A clause removing the election of specified office holders from judicial determination and validating a particular election retrospectively. Clauses declaring the amending power unlimited and barring any court from examining an amendment. An expansion of an immunity clause giving blanket primacy to one Part of the Constitution over the freedom and equality guarantees. A restructuring of judicial appointments that transferred effective control over the composition of the judiciary.

Against that, the list of amendments upheld or unchallenged runs to the overwhelming majority of all amendments made. Constitutional changes reorganising states, altering the legislative lists, restructuring taxation, creating new institutions, extending reservation, restructuring local government, and adjusting the machinery of government have all proceeded. The doctrine has not, in practice, made the Constitution difficult to amend. Amendments continue to be made at a rate that is high by comparative standards.

That fact cuts against both caricatures. It disposes of the claim that the judiciary has arrogated a veto over constitutional change, because the record shows the veto is exercised in a narrow class of cases. It also disposes of the claim that the doctrine is the only thing standing between India and constitutional collapse, because a doctrine used this sparingly is not carrying that much weight on a day-to-day basis. What the doctrine does is set an outer boundary that has been tested about half a dozen times in five decades, and its practical significance may lie as much in the amendments never attempted as in those struck down.

That last proposition cannot be proved, and it should be offered as an inference rather than a finding. A government’s law officers advising on a proposed amendment will assess the risk of a basic structure challenge, and measures that would clearly fail are unlikely to reach the drafting stage. The deterrent effect of a doctrine is invisible by nature, which is why arguments about its importance tend to be unfalsifiable in both directions.

What the separate opinions add

Kesavananda is usually taught through the summary, which is efficient and which loses most of what makes the case interesting. Reading the separate opinions changes the picture in three ways worth knowing even for a reader who will never read the judgments themselves.

The judges who accepted a substantive limit did not agree on its source. Some located it in the meaning of the word amend, arguing that to amend is to change while preserving identity, so that a change destroying identity is not an amendment at all and falls outside the power conferred. Some located it in the structure of the Constitution as a whole, arguing that certain features are presupposed by the document rather than created by any single provision. Some relied on the Preamble as evidence of the objectives the constitutional scheme was designed to serve. These are different arguments with different consequences, and the divergence explains why the doctrine’s content has never been settled: the majority agreed on the conclusion more than on the reasoning.

The judges in the minority were not indifferent to the risk of constitutional destruction. Several of them accepted that the amending power could not be used to abrogate the Constitution or to establish a wholly different order, while declining to accept a judicially enforceable limitation of the kind the majority adopted. The distance between the two camps was narrower than the seven to six split suggests, and the disagreement was substantially about whether the limit should be enforced by courts rather than about whether a limit exists.

And the judgments contain extensive treatment of the property and directive principles questions that the summary largely omits, because those were the issues the parties had actually come to argue. The constitutional question that made the case famous was, in the structure of the litigation, a preliminary point.

Why the doctrine survived the Emergency

The most surprising fact about the doctrine is that it existed for barely two years before being tested by the most determined attempt any Indian government has made to abolish a judicial limitation, and that it emerged intact.

Three things explain the survival.

The first is the drafting error at the heart of the attempt. The Forty-second Amendment did not try to abolish the doctrine by identifying and protecting specific measures. It tried to abolish judicial review of amendments altogether, by declaring the constituent power unlimited and barring any court from examining an amendment on any ground. That is the one form of attack the doctrine is best equipped to defeat, because a clause removing the limit is the paradigm case of using a limited power to enlarge itself. A more careful government would have made a narrower move.

The second is timing. The clauses were struck down in 1980, after the Emergency had ended, after the government that enacted them had lost an election, and after a successor Parliament had already reversed much of the Emergency architecture. The Court was not confronting a government at the height of its power; it was tidying up a repudiated project. Courts find it easier to be brave about the recent past than about the present, and it would be dishonest to attribute the outcome purely to institutional courage.

The third is that the doctrine had by then acquired constituencies outside the judiciary. Opposition parties, the bar, the press and a substantial part of public opinion had come to regard it as a protection rather than as an obstruction, largely because of what the Emergency had demonstrated. A doctrine with political support is harder to remove than a doctrine that only judges care about, and the doctrine acquired that support because of the way it was attacked.

The lesson generalises beyond India and is worth stating as a proposition rather than as a story about one amendment: a judicial limitation on constituent power survives not because of the quality of its reasoning but because of whether an attempt to remove it is perceived as legitimate. The Forty-second Amendment failed on perception before it failed in court. Anyone assessing the doctrine’s durability against a future challenge should be asking about legitimacy and timing rather than about the strength of the 1973 reasoning.

The counterfactual: an India without the doctrine

The alternative history is worth thinking through, because both sides of the argument tend to assume its answer rather than examine it.

Suppose Kesavananda had gone the other way by the same single vote. The Twenty-fourth Amendment would have been upheld in full, the amending power would have been unlimited in substance, and the position would have been that of Shankari Prasad restored and constitutionalised.

The immediate consequence would have been that the Thirty-ninth Amendment’s election clause would have stood. The Prime Minister’s election would have been validated by constitutional amendment and removed from judicial determination, and the Emergency’s principal legal justification would have been unnecessary in that respect.

The medium-term consequence is harder to predict and the honest answer is that it depends on a judgment about Indian politics rather than about law. One view is that the Forty-second Amendment’s entrenchment of executive power would have stood, that the six-year terms would have persisted, and that the restoration of 1977 to 1979 would have been considerably harder because the incoming government would have needed the same supermajorities to undo what had been done. On this view the doctrine bought the Republic a margin during the period when it was most needed.

The other view is that the electoral defeat of 1977 was decisive irrespective of doctrine, that a government thrown out by voters loses its constitutional project regardless of what any court has held, and that the real protection came from the electorate rather than from the bench. On this view the doctrine’s contribution has been symbolic rather than load-bearing, and the Republic’s actual defence against constitutional capture is that Indian voters have repeatedly removed governments that overreached.

Both views are held by informed people and neither can be tested. What can be said is narrower. The doctrine did strike down specific measures that a majority had enacted, those measures did concern the accountability of the executive to courts and to voters, and no other institution was in a position to strike them down. Whether that made the difference or merely accompanied the difference is a question the record cannot settle.

The verdict

The basic structure doctrine is a judicial construction with a thin textual foundation, an uncertain content, and a contested origin, and it is also the most important thing the Supreme Court of India has done.

Both halves of that sentence should be held together, because most writing on the subject picks one. The doctrine is not derived from the constitutional text and no amount of interpretive ingenuity makes it so. It has never been defined and it will not be. It was created by a single vote and its ratio was extracted from a document some of the bench declined to sign. Its most consequential recent application concerned the composition of the institution that applies it. Every one of those objections is accurate.

And a constitution with a rights chapter, a federal structure, a system of judicial review, and an amending procedure requiring a supermajority, but with no substantive limit on what a supermajority may do, is a constitution that can be lawfully dismantled by anyone who wins two consecutive elections decisively enough. India’s Parliament attempted something close to that within twenty-five years of the Constitution coming into force, in the open, using the ordinary procedure, and the only thing that stopped it was a doctrine that no framer had written. That is not a theoretical justification; it is the historical record.

The considered position this article takes is that the doctrine is a necessary invention that has been used with more restraint than its critics acknowledge and less predictability than its defenders admit, and that the strongest criticism of it is not that it exists but that its content is determined case by case by the institution with the largest stake in the outcome. That criticism cannot be answered within the doctrine. It can only be managed, by courts applying it sparingly, identifying the feature at issue precisely, and distinguishing genuinely between alteration and destruction rather than reaching for the doctrine whenever a change is unwelcome.

For readers building a working knowledge of this cluster, the practical advice is to organise the material by decision rather than by feature, because the doctrine only makes sense as a sequence of responses, and to keep the bench strengths, the splits and the specific holdings in a single place where they can be revised, which is exactly the kind of private, offline research file that VaultBook exists to hold. Aspirants should note that this is the single most heavily examined topic in Indian polity, that questions almost always test the distinction between what Kesavananda held and what it is popularly said to have held, and that working through how the question has actually been framed is more productive than memorising a list of features, which is why it repays working through previous-year questions on the ReportMedic explorer and keeping up a daily practice habit on ReportMedic.

The vocabulary problem: damage, destroy, abrogate, alter

Much of the confusion around the doctrine is caused by imprecise use of four words that carry different weights, and a reader who can distinguish them will follow the argument in any judgment.

Alter is neutral. Every amendment alters something, and alteration is what the amending power is for. An amendment that alters a basic feature has not, without more, damaged it, and a challenge framed as an objection to alteration is not a basic structure challenge at all.

Abridge is the vocabulary of Golak Nath and of Article 13. To abridge a right is to reduce it. Golak Nath held that no abridgement of Part III was permissible by amendment, and that holding is gone. Summaries that still reach for the word import a test that was overruled, and they produce a wrong answer.

Damage and destroy are the Kesavananda formulation, and the pairing is significant. Destruction is the clear case: a feature is removed or rendered inoperative. Damage is the harder one, and it is where the real argument happens, because it accepts that a feature can be impaired short of removal in a way that still fails the test. What courts have generally required is that the impairment go to the essence of the feature rather than to its incidents.

Abrogate is the Coelho vocabulary, applied to a scheduled statute and to the rights that form part of the basic structure. Abrogation is stronger than restriction. A statute that restricts a right in some respect does not abrogate it; a statute that empties the right of content in the field it occupies does.

The practical use of this vocabulary is diagnostic. When a commentator writes that an amendment violates the basic structure, ask which word is being used and whether the argument supports it. Most public argument about proposed amendments asserts destruction and establishes, at best, alteration, and the gap between the two is where nearly every failed challenge has fallen.

What the doctrine did not stop between 1973 and 1980

A doctrine is judged partly by what it prevented and partly by what it did not, and the honest record of the seven years after Kesavananda is mixed enough to be worth setting out.

The Emergency was proclaimed, and the proclamation itself was not invalidated. Preventive detention on a very large scale proceeded. The Thirty-eighth Amendment insulated the proclamation and related orders from judicial review, and the Thirty-ninth and Forty-second Amendments went further. Elections were postponed by extending the term of the House. Press censorship operated. The Supreme Court, in the habeas corpus litigation, held that the right to move a court for enforcement of the right to life and personal liberty stood suspended during the Emergency, which is the decision most often cited against the institution’s claim to have defended constitutional government.

The doctrine did not prevent any of that at the time. What it did was survive, and what it did afterwards was strike down the entrenchment clauses once the political weather had changed.

Two conclusions follow. The first is that a judicial limitation on constituent power is not a real-time protection against an executive with a supermajority and emergency powers; it operates on a slower cycle, and it operates best after the fact. Anyone who expects the doctrine to stop a determined government in the moment is expecting something it has never delivered.

The second is that operating after the fact is not worthless. The clauses declaring the amending power unlimited would otherwise have remained in the Constitution, available to any future government, and they were removed by a court rather than by a Parliament. The restoration project that a subsequent Parliament carried out left significant parts of the Emergency architecture standing, and the parts the courts removed were among the most dangerous. That division of labour, with Parliament reversing the machinery and the judiciary removing the entrenchment, is examined in this series’ account of the 42nd Amendment and what survived it.

Two different limits on the amending power

A recurring confusion in argument, and a common error in written answers, is to treat the basic structure doctrine as the only limit on constitutional amendment. There are two, they operate independently, and an amendment can fail either.

The first is procedural and textual. An amendment must be introduced correctly, must obtain the special majority in each House, and where it affects the specified federal provisions it must additionally be ratified by the legislatures of at least half the states. An amendment that misses any of these fails, and it fails without any court needing to consider its substance. That has happened, and the mechanics of it are set out in this series’ guide to the amendment procedure and the ratification proviso.

The second is substantive and judicial. Even an amendment that has cleared every procedural requirement may be invalidated for damaging the basic structure.

Keeping these apart matters for three reasons. They are decided by different questions, one about counting and consultation and the other about constitutional identity. They produce different remedies, since a procedural failure can be cured by doing it properly while a substantive failure cannot be cured at all. And they attract different criticisms: the procedural limit is uncontroversial because it is written down, while all the argument attaches to the substantive one. A reader who says that a court struck down an amendment should always be asked which limit was applied.

The doctrine beyond amendment review

Although the doctrine is formally a limit on the amending power, the values it identifies have been used in a second and less discussed way: as a standard for reviewing the exercise of certain constitutional powers by the executive.

The clearest instance concerns the power to dismiss a state government and impose central rule. In the litigation that settled the modern law on that subject, the Supreme Court held that secularism is a basic feature of the Constitution and that a state government acting in a manner that subverts it may furnish grounds for action, while simultaneously holding that the proclamation itself is subject to judicial review and that the material on which it is based can be examined. That decision imported the vocabulary of basic features into the review of executive action, and it is one of the reasons the doctrine’s boundaries feel less crisp in practice than the formal rule suggests.

The proper way to understand this is that the features identified through the amendment cases have become an accepted description of the Constitution’s fundamental commitments, and that description is then available in other contexts as an aid to interpretation. That is not the same as applying the doctrine, and courts have been careful, mostly, to distinguish the two. But the traffic between the two uses is real, and it is the mechanism by which basic structure language has spread into arguments where the doctrine strictly does not apply.

For a reader assessing whether an argument is sound, the test is simple. If the challenge is to a constitutional amendment, the doctrine applies directly. If the challenge is to anything else, the reference to basic features is doing interpretive work rather than supplying a ground of invalidity, and an argument that treats it as a ground is overstated.

The afterlife of Article 31C

One consequence of Minerva Mills is genuinely unresolved and it illustrates how untidy constitutional adjudication becomes when an amendment is struck down rather than repealed.

Article 31C was inserted by the Twenty-fifth Amendment with a limited scope, covering laws giving effect to two specified directive principles. The Forty-second Amendment widened it to cover laws giving effect to any directive principle. Minerva Mills struck down the widening.

The question that follows is what remains. One view is that striking down the amending words revives the narrower provision as it stood before, on the reasoning that an invalid amendment never validly displaced the earlier text, so the original Article 31C continues to operate in its limited form. The other view is that the provision as amended was a single new provision, that striking it down leaves nothing, and that a court cannot resurrect repealed text by invalidating the instrument that replaced it.

The practical position that has generally been proceeded upon is that the narrower version survives, but the question has been the subject of continuing argument and has attracted judicial attention more than once. A reader should treat it as an open question described structurally rather than as a settled rule, and should check the current position before relying on it in any matter where the point is material.

The general lesson is worth extracting because it recurs across this series. Striking down an amendment is not the same as repealing it, and the two produce different consequences for the text that the amendment had displaced. Where a repeal is enacted, Parliament can and usually does say what happens to the earlier position. Where a court invalidates, nobody has legislated for the aftermath, and the revival question has to be answered by doctrine that was not designed for it. Any account of the basic structure doctrine that stops at the moment of invalidation has skipped the part that practitioners actually have to deal with.

How the doctrine changed the way amendments are drafted

The least visible effect of the doctrine is on drafting practice, and it is probably the most pervasive.

Constitutional amendments after 1980 are noticeably more careful about judicial review than those before it. The technique of inserting a clause barring courts from examining a measure, which appears repeatedly in the amendments of the seventies, has largely disappeared, because it is now understood to be the single most reliable way to attract a successful challenge. Where governments have wanted to limit judicial scrutiny, the drafting has shifted to narrower devices: specifying the grounds on which a matter may be examined, channelling disputes to a designated forum, or defining the subject matter tightly enough that the reviewable question is small.

Amendments touching the federal structure have similarly become more elaborate, providing for consultation mechanisms and shared institutional arrangements rather than transferring competence outright, which reduces the argument that federalism has been damaged rather than merely reorganised.

And the ratification requirement is now taken seriously in a way it was not always taken seriously, because a procedural failure is a cleaner ground of invalidity than a substantive one and governments have learned that the proviso has teeth.

None of this can be proved to be the doctrine’s doing, since drafting practice responds to many pressures. But the correlation is strong, the mechanism is obvious, and law officers advising on constitutional amendments have every incentive to draft around the risk. If the doctrine has a routine day-to-day effect on Indian constitutional government, this is where it is: not in the handful of amendments struck down, but in the shape of the amendments that were written to survive.

The two things Kesavananda upheld

Accounts of the case concentrate on what it limited, which produces a distorted impression of the outcome. The government won substantially on the day, and understanding what it won is necessary to understanding the shape of the doctrine.

It won on the amending power itself. The Twenty-fourth Amendment was upheld in its entirety. Parliament’s competence to amend any provision of the Constitution, including the fundamental rights, was confirmed, and the contrary holding in Golak Nath was set aside. Everything the government had wanted from that amendment, other than immunity from substantive review, it received.

It won on the first limb of Article 31C. Laws giving effect to the directive principles concerning the distribution of material resources and the prevention of concentration of wealth were held immune from challenge on equality and freedom grounds. That is a substantial constitutional preference for a category of redistributive legislation, and the Court upheld it.

What the government lost was narrow and specific: the clause making a legislative declaration of purpose conclusive and unexaminable by any court. Strip away the rhetoric and the pattern is the one that recurs throughout the doctrine’s history. The Court did not object to the redistributive project, did not object to the immunity conferred on it, and did not object to Parliament amending the rights chapter to achieve it. It objected to being told that it could not examine whether the declaration was true.

That is the doctrine’s centre of gravity and it has stayed there. Substantive policy, however radical, is generally safe. Removing the capacity to review is generally not. A reader who wants a single sentence to carry away from the case should carry that one rather than any formulation about basic features.

Implied limitations and where they come from

The doctrine is properly described as an implied limitation, and the phrase repays a moment’s attention because it locates the argument accurately.

An express limitation is one the text states. The requirement of ratification by half the states for specified provisions is an express limitation on the amending power, written into the proviso, and nobody disputes it.

An implied limitation is one a court derives from the structure or purpose of an instrument rather than from its words. Courts derive implied limitations routinely in other contexts: a power conferred for a purpose cannot be used for a different purpose, a delegate cannot sub-delegate without authority, a body cannot use a procedural power to enlarge its own jurisdiction. None of these rules is generally written down, and none of them is regarded as judicial invention.

The Kesavananda reasoning applies that ordinary technique to the constituent power, and its core move is about the meaning of a single word. To amend an instrument is to change it while it remains recognisably the same instrument. A change that produces a different constitution is not an amendment of the existing one, whatever it is called, so it falls outside the power that Article 368 confers rather than being an unlawful exercise of that power. On this reading the Court is not limiting the amending power at all; it is describing its boundary.

The objection to that move is not that implied limitations are illegitimate in general. It is that the identity of a constitution is not a fact that can be determined the way the scope of a statutory power can, that reasonable people will disagree about when a change destroys identity, and that vesting the determination in a court makes the court the final judge of what the Constitution essentially is. That objection is sound as far as it goes. Where it stops short is that the alternative, treating the amending power as capable of producing any constitution at all, requires accepting that a power conferred by an instrument can be used to abolish the instrument, which is the proposition ordinary legal reasoning refuses everywhere else.

Both sides of that exchange are worth holding in mind, because the argument about the doctrine is often conducted as though the Court had done something unprecedented in kind. It did something familiar in kind and unprecedented in consequence, and those are different criticisms.

What a future challenge would look like

Because the doctrine is applied to specific instruments rather than to categories, the most useful thing an article can offer about its future is a structural account of where the risk sits, framed durably rather than tied to whatever amendment happens to be under discussion.

An amendment faces serious basic structure risk when it does one of four things.

It removes a category of dispute from judicial determination, whether by ousting jurisdiction outright, by making an executive or legislative determination conclusive, or by validating something retrospectively notwithstanding contrary judgments. This is the pattern in every amendment the doctrine has actually defeated, and it is defeated because the doctrine’s own survival depends on courts retaining the capacity to apply it.

It alters the mechanism by which power is contested or transferred. Amendments touching the conduct of elections, the term of a legislature, the disqualification of members, or the resolution of election disputes are examined with far more suspicion than amendments of policy, because they affect the process through which a majority becomes a majority.

It changes the composition or independence of the institution that would review it. The appointments litigation established this and did so in the most contested way possible, but the principle predates it: an amendment affecting who judges, on what terms, and with what security of tenure, is examined against the independence of the judiciary as a recognised feature.

And it enlarges the amending power itself, directly or by implication. Minerva Mills settled that this fails automatically, and the reasoning has not been doubted since.

Conversely, an amendment is at low risk when it changes what the state may do rather than who reviews it: transfers of legislative competence, new taxation arrangements, new institutions, new categories of beneficiary, restructured local government, altered service conditions. These have been made repeatedly and have survived, and the pattern is stable enough to be predictive.

Two cautions belong with that framework. It describes risk rather than outcome, since the doctrine turns on characterisation and characterisation is contested in every serious case. And it is a description of how courts have behaved rather than a rule they have announced, since no judgment has laid down categories of this kind. It is offered as a working model for reading a proposed amendment, not as a statement of law.

A note on citing the doctrine accurately

Because Kesavananda produced eleven judgments and a separately drawn summary, citing it carelessly is easy and it is the commonest way a written argument on this subject signals that its author has not read the case.

Three errors recur. The first is attributing a proposition to the Court when it appears only in one judge’s opinion. Several of the most quotable formulations about constitutional identity and the meaning of amendment come from individual judgments rather than from anything a majority adopted, and presenting them as the holding overstates them. The second is citing the case for the proposition that fundamental rights cannot be amended, which reverses the actual holding and which appears with depressing regularity even in otherwise careful writing. The third is treating a compiled list of basic features as though the Court had enumerated it.

The safe practice is to state the holding in the narrowest form the record supports, to identify which later decision established any refinement being relied on, and to name the specific feature said to be damaged rather than gesturing at the doctrine generally. A submission that says an amendment damages judicial review because it removes a defined category of dispute from every court, and cites the decision that established judicial review as a basic feature, is arguable. A submission that says an amendment violates the basic structure without more is not an argument at all, and courts say so.

There is a further reason to be careful, and it is about audience rather than accuracy. The doctrine is invoked constantly in political argument by people who have no intention of litigating anything, and the vocabulary has been worn smooth by that use. When a term is available to describe any constitutional change one dislikes, its precision decays, and the decay eventually reaches the courtroom, where an advocate arrives with a submission built from newspaper usage rather than from the decisions. Keeping the citation tight is therefore not pedantry; it is the mechanism by which a doctrine with no textual anchor retains enough definition to be applied at all.

The same discipline serves a reader assessing commentary. An account that names the decision, the feature, and the specific holding is doing the work. An account that invokes the doctrine as a general objection to a change it dislikes is doing something else, and the distinction is usually visible within a paragraph.

Frequently Asked Questions

Q: What is the basic structure doctrine and why does it matter?

It is the rule that Parliament may amend any provision of the Constitution but may not, through amendment, damage or destroy the Constitution’s basic structure. It matters because it is the only substantive limit on constituent power in a system where the amending procedure requires nothing beyond a special majority in each House and, for certain provisions, ratification by half the states. Without it, a government with a two-thirds majority could lawfully remove judicial review, extend its own term, dissolve the federal arrangement, or set aside elections, since each of those would be an amendment. The doctrine has been used sparingly, mainly against measures that ousted courts, entrenched incumbents, or purported to make the amending power unlimited.

Q: What did Kesavananda Bharati decide by a seven to six majority?

That the amending power in Article 368 is subject to an implied substantive limitation: an amendment may not damage or destroy the basic structure of the Constitution. The same decision upheld the Twenty-fourth Amendment and overruled Golak Nath, confirming that Parliament can amend any Part of the Constitution including the fundamental rights. It also struck down the second limb of Article 31C, which had made a legislative declaration of purpose conclusive and unexaminable by courts. The case was heard by thirteen judges, the largest bench ever constituted in the Supreme Court of India, presided over by Chief Justice S M Sikri, and produced eleven separate judgments, from which a summary titled the View of the Majority was drawn and signed by nine.

Q: Which features have been recognised as part of the basic structure?

Features identified across various judgments include the supremacy of the Constitution, the republican and democratic form of government, secularism, the separation of powers, the federal character of the distribution of powers, judicial review, free and fair elections, the rule of law, the independence of the judiciary, the limited nature of the amending power, and the balance between fundamental rights and directive principles. The critical point is that no court has produced an exhaustive list and the Supreme Court has declined invitations to do so. Compilations circulating in study material have no authoritative source. Treating any such list as closed is an error, and a pleading that relies on one rather than identifying the specific feature at issue and showing damage to it is not a properly framed argument.

Q: Why was Golak Nath overruled?

Because the rule it laid down was absolute and therefore unworkable. Golak Nath held that a constitutional amendment is law within the meaning of Article 13 and that Parliament consequently has no power to abridge the fundamental rights at all. That froze the entire rights chapter against correction, drew no distinction between a trivial amendment and a destructive one, and offered no principle for identifying which changes genuinely threatened the constitutional order. It also applied prospectively, allowing the amendments it condemned to stand, which made it look like a political accommodation rather than a legal holding. Kesavananda replaced the flat bar with a calibrated test that permits amendment of any Part while forbidding damage to the constitutional core.

Q: Can a law placed in the Ninth Schedule still be challenged in court?

If it was added after 24 April 1973, yes. Waman Rao fixed that date, the day Kesavananda was decided, as the dividing line, holding that earlier additions would not be reopened because transactions had been completed on the strength of them. I R Coelho then settled the test for later additions: the court examines the scheduled statute itself and asks whether it violates or abrogates the fundamental rights that form part of the basic structure, principally the guarantees of equality, freedom and life and personal liberty read together. The threshold is high, because ordinary regulatory restriction of a right is not abrogation. Entries made before the dividing line retain the protection Article 31B confers.

Q: What did Minerva Mills add to the doctrine?

Two propositions, one of which is foundational. It held that a limited amending power is itself part of the basic structure, so Parliament cannot use the amending power to declare that power unlimited, and struck down the clauses inserted by the Forty-second Amendment which had declared the constituent power unlimited and barred any court from examining an amendment on any ground. It also held that the balance between the fundamental rights in Part III and the directive principles in Part IV is a basic feature, and struck down the expansion of Article 31C which had given blanket primacy to the whole of Part IV over the equality and freedom guarantees. The first holding is what makes the doctrine self-protecting; the second is more contested.

Q: Does the Constitution itself mention the basic structure doctrine?

No. The words do not appear anywhere in the constitutional text, and Article 368 as originally enacted set out only a procedure for amendment without stating any substantive limit. The doctrine is entirely judicial in origin, derived from reasoning about what the word amend can mean, about the structure the Constitution presupposes, and about the objectives expressed in the Preamble. Critics regard that absence as fatal to its legitimacy. Defenders point out that the competing rule, the distinction between legislative power and constituent power adopted in Shankari Prasad to reach the opposite result, is equally absent from the text, so the argument from silence does not favour either side as cleanly as it appears to.

Q: Why does 24 April 1973 matter for Ninth Schedule laws?

It is the date Kesavananda Bharati was decided, and Waman Rao adopted it as the line dividing Ninth Schedule entries into two classes. Statutes added to the schedule before that date are not reopened, on the reasoning that nobody had any basis before Kesavananda for supposing the schedule was vulnerable, land had been redistributed and transactions completed on the strength of the protection, and unsettling them would cause disproportionate disruption. Statutes added after that date are open to challenge on the ground that they damage the basic structure. The Court was candid that the choice of date rests on settled expectations rather than on principle, since no doctrinal reason distinguishes an addition made a day before from one made a day after.

Q: How many judges heard Kesavananda Bharati and why was the bench so large?

Thirteen, which remains the largest bench ever constituted in the Supreme Court of India. The size followed from what was at stake procedurally. Golak Nath had been decided by eleven judges, and a bench seeking to reconsider that holding had to be larger than the one that gave it. The questions before the Court concerned the validity of three constitutional amendments and, through them, the scope of the amending power itself, so the Court assembled the largest bench available. The consequence was eleven separate judgments running to hundreds of pages, a hearing lasting months, and the practical impossibility of extracting a ratio without a summary document, which is the source of much of the later argument about the case.

Q: What is the golden triangle in Indian constitutional law?

It is the shorthand for the three fundamental rights guarantees read together: equality before the law and equal protection of the laws, the freedoms including speech, assembly, association, movement, residence and occupation, and the protection of life and personal liberty. The phrase describes the proposition that these guarantees are not independent silos but a connected core, so that state action must satisfy all three rather than any one. It matters for the basic structure doctrine because the test settled in I R Coelho for statutes added to the Ninth Schedule after the dividing line asks whether the statute abrogates the rights forming part of the basic structure, and that examination proceeds principally against these three guarantees taken together.

Q: Has the basic structure doctrine ever been used to strike down an ordinary statute?

Not on that ground, and arguments seeking to extend it that way have consistently failed. The doctrine is a limit on the power to amend the Constitution. An ordinary statute is tested against the constitutional provisions themselves, which is the ordinary judicial review that has existed since the Constitution came into force. The one qualification is the Coelho position, which permits examination of a statute placed in the Ninth Schedule after 24 April 1973, but that is a consequence of the statute having been given constitutional immunity by an amendment rather than an application of the doctrine to ordinary legislation. Litigants nonetheless raise the argument regularly, and courts continue to reject it.

Q: Did the 42nd Amendment succeed in abolishing the basic structure doctrine?

No, and the attempt failed by the very test it was designed to remove. The amendment added two clauses to Article 368, one declaring that there shall be no limitation whatever on the constituent power of Parliament and the other barring any court from calling an amendment in question on any ground. Minerva Mills struck both down in 1980, holding that a limited amending power is itself a basic feature and that Parliament cannot use a limited power to convert it into an unlimited one, because if it could the limitation would be meaningless. The episode is the clearest demonstration that the doctrine is self-protecting against direct assault, though it should be noted that the clauses were struck down after the government that enacted them had lost office.

Q: Is the Preamble part of the basic structure?

The question conflates two separate issues. The Preamble is part of the Constitution and is amendable, and it has in fact been amended. Separately, several features recognised as basic draw on values the Preamble expresses, including the republican and democratic form of government, secularism and the sovereignty of the Republic. Both propositions hold simultaneously, because the doctrine protects features rather than provisions, and a feature may be reflected in several provisions or in none exclusively. The correct answer is therefore that the Preamble can be amended, that an amendment to it is subject to the same limit as any other amendment, and that the Preamble is treated as evidence of the objectives against which the constitutional scheme is read.

Q: Why is the View of the Majority document in Kesavananda controversial?

Because it is the conventional source of the case’s ratio and its status is contested. Thirteen judges delivered eleven separate judgments, and at the close of the hearing a summary of the propositions on which a majority had agreed was drawn up and signed by nine of them. Four did not sign, and it has been argued that some signatories had written judgments not obviously supporting every proposition in the summary. The objection is that a doctrine of this consequence should not rest on a document produced as an administrative convenience. The answer generally accepted is that the doctrine’s authority no longer depends on that document at all, since later benches, including one of nine judges, have applied and refined it for decades.

Q: Did the Supreme Court use the doctrine to strike down the judicial appointments amendment?

Yes. Parliament amended the Constitution to replace the collegium system with a National Judicial Appointments Commission whose membership included the Union Law Minister and two eminent persons alongside the Chief Justice of India and two senior judges, with a mechanism allowing a small minority to block a recommendation. The Supreme Court struck down the amendment and the accompanying statute, holding by a majority that judicial primacy in appointments is an aspect of judicial independence and that independence forms part of the basic structure. The decision is the most contested modern application of the doctrine, because the Court protected an arrangement of its own creation, using a limitation of its own creation, in a matter concerning its own composition.

Q: What is the difference between damaging the basic structure and merely altering it?

It is the difference that decides almost every case. Every constitutional amendment alters something, and alteration is exactly what the amending power exists to permit, so a challenge that establishes only alteration establishes nothing. The Kesavananda formulation requires damage or destruction, which means the impairment must reach the essence of the feature rather than its incidents. In practice this is why amendments transferring legislative subjects, restructuring taxation, reorganising states or extending reservation survive despite plainly affecting features described as basic, while amendments that oust judicial review, decide a pending dispute, or entrench the amending power against being checked do not. Most public argument asserts destruction and demonstrates alteration.