The Basic Structure

What has been held to be part of the structure

Only a court deciding a case says what the structure contains, so the list grows. Shelat and Grover JJ.: these elements "cannot be catalogued but can only be illustrated".

• Kesavananda, per Sikri C.J.: "Supremacy of the Constitution"; "Republican and Democratic form of Government"; "Secular character of the Constitution"; "Separation of powers between the Legislature, the executive and the judiciary"; "Federal character of the Constitution", all resting on "the dignity and freedom of the individual". Shelat and Grover JJ. added "the unity and the integrity of the nation".
• Minerva Mills: the limited nature of the amending power, and "This harmony and balance between fundamental rights and directive principles is an essential feature of the basic structure of the Constitution."
• I.R. Coelho: article 21 read with articles 14 and 19, "and the principles underlying them".
• Not in it, per Khanna J.: "Right to property does not pertain to basic structure or framework of the Constitution." See [The Right to Property].

The doctrine is laid down

Facts. Kesavananda Bharati v. State of Kerala, AIR 1973 SC 1461, (1973) 4 SCC 225, decided 24 April 1973. The head of a mutt challenged the Kerala Land Reforms Act 1963 as amended and, by amendment of his petition, the Twenty-fourth, Twenty-fifth and Twenty-ninth Amendments. Waman Rao records that seven of the thirteen judges formed the majority.

Held. The amending power reaches every provision, but not to any effect. Khanna J.: "The power of amendment under Article 368 does not include power to abrogate the Constitution nor does it include the power to alter the basic structure or framework of the Constitution." Sikri C.J.: every provision may be amended "provided in the result the basic foundation and structure of the Constitution remains the same". Shelat and Grover JJ. met the objection of vagueness head on: "The basic structure of the Constitution is not a vague concept and the apprehensions expressed on behalf of the respondents that neither the citizen nor the Parliament would be able to understand it are unfounded."

Why it matters. Its form matters as much as its content. There is no single majority judgment: the doctrine comes from separate opinions agreeing on a proposition and differing on its application, which is why the later cases settle what it means rather than whether it exists.

The doctrine bites

Facts. Minerva Mills Ltd v. Union of India, AIR 1980 SC 1789, 1981 SCR (1) 206, decided 31 July 1980. A textile company whose undertaking had been nationalised challenged the nationalisation and, with it, sections 4 and 55 of the Constitution (Forty-second Amendment) Act 1976. Section 55 had inserted clauses (4) and (5) into article 368; section 4 had amended article 31C so that a law giving effect to any directive principle was immune from challenge under articles 14 and 19.

Held. Both sections are void. Clause (5) "transgresses the limitations on the amending power": "Since the Constitution had conferred a limited amending power on the Parliament, the Parliament cannot under the exercise of that limited power enlarge that very power into an absolute power. Indeed, a limited amending power is one of the basic features of Indian Constitution." In the Court's phrase, "The power to destroy is not a power to amend." Clause (4) falls with it: if courts cannot pronounce on an amendment then "Article 13 of Constitution will then become a dead letter". On article 31C, "to destroy the guarantees given by Part III in order purportedly to achieve the goals of Part IV is plainly to subvert the Constitution by destroying its basic structure", and it had "removed two sides of that golden triangle", articles 14 and 19.

Why it matters. It is the proof that the doctrine is not decorative. Parliament had legislated to abolish the limit and to abolish review of the abolition, and both failed on one reasoning: a donee of a limited power cannot use that power to make itself unlimited.

The doctrine gets a date

Facts. Waman Rao v. Union of India, AIR 1981 SC 271, 1981 SCR (2) 1, decided 9 May 1980. Landholders challenged the Maharashtra Agricultural Lands (Ceiling on Holdings) Act 1961 as amended, and with it articles 31A, 31B and the unamended 31C. The Acts were in the Ninth Schedule, so the real question was whether an amendment putting a law there could itself be attacked.

Held. The First and Fourth Amendments do not damage the basic structure, being made "so closely on the heels of the Constitution" as to be "a contemporary practical exposition of the Constitution". Then the line was drawn. Amendments to the Ninth Schedule made before 24 April 1973 are valid; those made "on or after April 24, 1973" are open to challenge as damaging the basic structure.

Why it matters. For the date, which is the day Kesavananda was decided. From that day Parliament was on notice that the power was limited. It is a frankly pragmatic line, drawn to avoid unsettling three decades of land holding, and saying so is part of a good answer.

The doctrine gets a test

Facts. I.R. Coelho v. State of Tamil Nadu, decided on 11 January 2007 by a Bench of nine. The question was whether a law placed in the Ninth Schedule, and so protected by article 31B, may nevertheless be attacked as damaging the basic structure. The report carries no equivalent citations line, so no report reference is printed for it here.

Held. A law abridging Part III rights may or may not violate the basic structure; if it does it must be invalidated, whether it got there by amending Part III or by insertion in the Ninth Schedule. Validity is judged "on its own merits", by "the actual effect and impact of the law on the rights guaranteed under Part III". Post-1973 Ninth Schedule entries are tested "on the touchstone of the basic or essential features of the Constitution as reflected in Article 21 read with Article 14, Article 19, and the principles underlying them".

Why it matters. It converts a doctrine about amendments into a test about effects. The question is not what an amendment is called but what it does: the impact test.

The criticism, at its strongest

The doctrine has no textual home, and that is not a small objection. Article 368(1) says "any provision of this Constitution". The proviso lists the articles that get extra protection, and it is a closed list Parliament could have extended and did not. The Constituent Assembly wrote no unamendable clause. So the Court has read into article 368 a restriction that is not in it, and used it to override an amendment carried by both Houses and, where the proviso applies, by half the States.

The vagueness objection is sharper, and its strongest version comes from inside the Court. Chelameswar J., in the judgment on the Ninety-ninth Amendment quoted in Janhit Abhiyan, decided on 7 November 2022, summarised the case law and included these propositions: "None of the cases referred to above specified or declared what is the basic structure of the Constitution", and "There is no universally applicable test vis-a-vis all the basic features."

Put the two together and the objection is complete. Unelected judges may strike down what elected legislatures have enacted, by a standard nowhere written and never exhaustively defined.

The best answer to it

First, the textual home is the word "amend". A power to amend is not a power to make a new constitution. Sikri C.J. reasoned from "the concept of amendment within the contours of the Preamble and the Constitution", Khanna J. from limitations that "inhere and are implicit in the word 'amendment'". The Court added no restriction; it declined to read a word as its own opposite.

Second, the alternative was tried, and its consequence is unanswerable. Clause (5) said in terms that there shall be "no limitation whatever", and clause (4) said no court could look. That is the literal position with nothing hidden, and Minerva Mills states where it ends: a law made under such an amendment escapes review entirely, "because it will receive the protection of the constitutional amendment which the courts will be powerless to strike down".

Third, vagueness is a real cost and not a defeat. Nobody can list the structure in advance, and the Court has not pretended otherwise. But a rule can work without being exhaustive: negligence and reasonableness are no better defined and are applied daily. The doctrine trades certainty for security, deliberately.

What it does NOT mean

It does not mean an amendment is void merely because it touches a fundamental right. I.R. Coelho is express: a law abridging Part III "may violate the basic structure doctrine or it may not". The enquiry is into effect, not subject matter.

It does not mean the doctrine applies to ordinary legislation. It limits the constituent power under article 368. An ordinary Act is tested under article 13 and under the legislative lists, not against the basic structure. And there is no fixed list: every feature named above is one a court held to be part of the structure in a decided case.

A worked example

Parliament amends the Constitution so that disputes about the validity of central laws go to a tribunal whose members the Union Government appoints, and the jurisdiction of the Supreme Court and the High Courts is excluded. Seventeen State legislatures ratify it. Ashok challenges it.

Step one, does article 368 permit it in terms? Yes. Clause (1) reaches "any provision", and the ratification requirement was met.

Step two, what is Ashok left with? Only the basic structure. He cannot say the amendment is "law" under article 13(2): see [The Amending Power].

Step three, what feature does he name? Judicial review, and the separation of powers, which is in Sikri C.J.'s list. The amendment does to the Supreme Court what clause (4) of article 368 attempted, and rights without remedies "are as writ in water".

Step four, apply the impact test. The question is the actual effect. If review survives in substance because the tribunal is independent and an appeal lies to a court, the amendment may stand. If review is gone, form cannot save it: "the form of an amendment is not the relevant factor, but the consequence thereof". See [Judicial Review] and [Tribunals].

Quick revision

• Article 368 gives a power to amend, not to destroy; no amendment may damage or destroy the basic structure. It binds constituent power, so State ratification cures no breach.
• Kesavananda (1973), thirteen judges, seven to six: laid down in separate opinions, with no single majority judgment.
• Sikri C.J.'s five: supremacy of the Constitution, republican and democratic government, secular character, separation of powers, federal character; founded on the dignity and freedom of the individual.
• Minerva Mills (1980): clauses (4) and (5) and the amended article 31C void. "The power to destroy is not a power to amend."
• Waman Rao (1980): Ninth Schedule entries made on or after 24 April 1973 are open to challenge.
• I.R. Coelho (2007), nine judges: the impact test, through article 21 read with articles 14 and 19.
• Not in the structure: the right to property, per Khanna J.
• The criticism: no textual home and no exhaustive definition. The answer: "amend" is the text, and the alternative is an unlimited power over the Constitution.

Test yourself

1. State the doctrine in one sentence, and the word in article 368 it is built on.

2. What did Minerva Mills strike down, and what reasoning disposed of clause (5)?

3. Why does 24 April 1973 matter, and which case drew the line there?

4. Explain the impact test, and say why an amendment's form is not decisive.

5. "The doctrine has no textual home." Put that criticism at its strongest, then answer it.

6. Name three things held to be part of the structure and one held not to be.