Federalism in India: Constitutional Provisions and the Nature of the Indian Model

Every federation answers one practical question: how do territorially distinct communities pool enough power to be secure and prosperous together without dissolving into one another? India answered it during a partition, while absorbing over five hundred princely states, using machinery inherited from a unitary colonial regime.

The document that emerged calls itself a Union of States rather than a federation, keeps the residue of legislative power at the centre and lets Parliament redraw the map of the units — and then gives those units an entrenched legislative sphere, a chamber of their own, and a court to defend both. Whether that combination is a federation at all is the oldest live argument in Indian political science, and it is not a semantic one: what India is called determines what the courts will protect.

Federalism as Government by Contract

• Federalism derives from the Latin foedus — a covenant, compact or contract, the root that also gives English “confederate” and “fidelity”.
• A federation is therefore government by contract: two or more orders of government, each drawing authority from a common written instrument rather than from each other. Neither level is the creature of the other, and a province whose powers can be withdrawn by ordinary central legislation is not a federal unit but a devolved administrative region.
• A.V. Dicey supplied the phrase that still does most of the work: federalism is a political contrivance for those who desire union but not unity — unity without uniformity — an artificial device adopted because the alternatives, separation or forced homogenisation, are worse.

Why communities agree to federate

• Security. Units realise they cannot defend themselves separately, and that separate establishments duplicate cost while multiplying vulnerability — the American colonies against British reconquest, the German states against French pressure.
• Economic advantage. Federation creates a single market: a seamless internal flow of goods, capital, services and persons behind a common external tariff. India devotes an entire Part of the Constitution — Articles 301 to 307 — to guaranteeing that internal freedom.
• Management of diversity. Where linguistic, religious or national communities are territorially concentrated, federalism lets each govern what is closest to its identity while sharing what requires scale. This motive dominates in India, Canada, Belgium, Switzerland and Spain, and produces a different institution from the security-and-market variety.

The essential features of a federation

• A written constitution, because a contract either party may rewrite unilaterally is not a contract.
• A division of powers between two orders, each legislating on its subjects in its own right rather than by permission. India does this through the Seventh Schedule; the United States enumerates federal powers and reserves the rest to the states; Canada enumerates both and gives the residue to the centre.
• An independent judiciary as arbiter, because the contract will be disputed and one contracting party cannot judge its own cause.
• Bicameralism with a chamber representing the units, expressing the states as states rather than as one undifferentiated electorate.
• Rigidity of amendment for the federal provisions. India’s expression of this is the proviso to Article 368, requiring ratification by at least half the state legislatures for amendments touching the distribution of powers, the representation of states in Parliament, the Seventh Schedule, the election of the President and the extent of the executive power.

“The Indian Union is a unitary State with subsidiary federal features rather than a federal State with subsidiary unitary features.” — K.C. Wheare

Federation, confederation and unitary system

Feature	Unitary	Federation	Confederation
Source of sub-national power	Delegated by the centre, withdrawable by ordinary law	Constitutional — both levels draw from one document	Retained by the units; the centre’s power is their grant
Sovereignty	Single, at the centre	Divided by subject; units lose external sovereignty permanently	Units keep sovereignty, including external
Right of exit	Not applicable	None — the union is indestructible	Units may secede
Judicial umpire	Not structurally required	Essential	Absent or weak
Amendment	By the central legislature	Unit concurrence for federal provisions	Usually unanimity
Examples	United Kingdom, France, Japan	United States, Germany, Australia, India	European Union, ASEAN

• The decisive line was drawn by the American Supreme Court in Texas v. White, describing the United States as an indestructible Union composed of indestructible States. A federation cannot be dismantled by its members; a confederation can.
• Units in a federation lose external sovereignty once and for all, with two consequences visible in the Indian text.
  • No right of secession — which is what Article 1’s language was chosen to secure.
  • No independent foreign policy. Article 253 lets Parliament legislate for the whole or any part of India to implement any treaty or international agreement, notwithstanding the Seventh Schedule. The federal division simply does not apply to the external sphere.
• The Soviet Union was confederal in design: its republics held a formal right to secede and a formal capacity for external relations. India accordingly maintained diplomatic dealings with Central Asian republics such as Tajikistan while they were still Soviet units — something structurally impossible with a unit of a genuine federation, and a design that failed along its constituent seams in 1991.

Models of federal practice

• Dual federalism — the American original. The two governments operate as watertight compartments, each supreme within its own sphere, with minimal overlap of function, personnel or finance.
• Cooperative federalism — Canadian in origin, elaborated in the United States from the 1930s welfare state. The two levels are interdependent rather than independent, their functions interlocking rather than separated.
  • M.P. Singh makes the chronological point that matters: by the time India drafted in 1946–49, cooperative federalism was already established in federal theory, so the framers were not deviating from an orthodoxy but adopting a live alternative.
• Executive or administrative federalism — the division of legislative power does not track the division of administrative power. One level legislates; the other implements, as in Germany and, for most Union law, in India.

How federations are classified

Axis	Categories	India
Origin	Coming-together, holding-together, putting-together	Holding-together
Direction of force	Centrifugal vs centripetal	Centripetal
Status of units	Symmetrical vs asymmetrical	Asymmetrical, de jure and de facto
Relation between levels	Dual vs cooperative	Cooperative
Locus of administration	Legislative vs executive	Executive

• Alfred Stepan’s three-fold typology of federal origins is the most useful single classificatory tool available.
  • Coming-together federations are formed by previously sovereign units bargaining into a union they could have declined — the United States, Switzerland, Australia. Because the units set the terms, they extract equal representation in the upper house, residuary powers and a demanding amendment rule.
  • Holding-together federations are formed by a previously unitary state devolving power to hold a diverse polity together — India, Belgium, Spain. The centre sets the terms, so the guarantees run the other way, and asymmetry is normal because accommodations are negotiated one region at a time. Putting-together federations are assembled by coercion around a dominant core, the Soviet Union being the paradigm.

The theorists whose vocabulary the debate uses

• K.C. Wheare supplied the discipline’s working definition: a federal government exists where the general and regional governments are each, within a sphere, co-ordinate and independent.
• William Riker treated federation as the outcome of a bargain rather than a philosophy. Leaders accept federation when they want territorial expansion without conquest and will trade central authority for the units’ consent; the terms reflect the parties’ relative strength at the founding.
  • The implication for India is clarifying: because the units had almost no bargaining power in 1947–50 — the princely states were being absorbed, the provinces were colonial creations — the bargain was struck heavily in the centre’s favour.
• William Livingston shifted the ground with a sociological account: federalism is a property of societies, not of constitutions, and federal institutions are merely the instruments through which territorially grouped diversities express themselves. On that test India is unambiguously federal whatever Article 1 says.
• Daniel Elazar returned the concept to its root, treating federalism as covenant — partners bound by mutual obligation rather than hierarchy — and formulated it as the combination of self-rule and shared rule.
  • Units govern themselves in their own sphere and also participate in governing the whole. The formulation names the half most often forgotten: a unit left alone but excluded from national decision-making is not a federal partner.
  • India’s institutions map onto both. The State List and Articles 245–246 supply self-rule; the Rajya Sabha, the Article 368 ratification proviso, the GST Council, the Finance Commission and the Inter-State Council supply shared rule. The persistent Indian grievance is not about self-rule but about shared rule — the institutions of joint decision are consultative rather than co-decisional.

“India, That Is Bharat, Shall Be a Union of States”

• Article 1 does not use the word “federation”, and the Drafting Committee’s rejection of “Federation of States” was deliberate.
• B.R. Ambedkar gave two structural reasons.
  • The Union is not the result of an agreement between the states. No unit negotiated its way in, so none can negotiate its way out or claim the terms were extracted from it.
  • No state has a right to secede. The division of the country into states is a convenience of administration, not a recognition of prior sovereignties.

“The Federation is a Union because it is indestructible. Though the country and the people may be divided into different States for convenience of administration the country is one integral whole, its people a single people living under a single imperium derived from a single source.” — B.R. Ambedkar

• India’s federal character is therefore a grant of the Constitution, not a residue of prior sovereignty. The states did not create the Union; the Constitution created both.
• Ambedkar’s own summary is the most honest description of the design available: the Constitution can be both unitary and federal according to the requirements of time and circumstances. In normal times it works as a federation; in an emergency it converts to something close to a unitary state, and then converts back.
• The corresponding formula — an indestructible Union of destructible States — captures the asymmetry exactly. The Union cannot be dissolved; any state can be merged, divided, renamed or extinguished by ordinary majority under Article 3.
• Article 1 also settles what the states are: the First Schedule lists them, and the territory of India comprises the states, the union territories and any territory acquired. There are at present 28 states and 8 union territories.

Is India a Federation? The Classificatory Debate

Wheare’s quasi-federal thesis

• K.C. Wheare compared the Indian Constitution against the American model and called the result quasi-federal — in the phrase that stuck, federal in form but unitary in spirit, carrying subsidiary federal features and prominent unitary ones.
• The features he identified are real ones: legislative power tilted to the centre, with the longer List and the residue both belonging to the Union; Parliament’s power to legislate on State List subjects without state consent; Article 356; the absence of dual citizenship, a dual judiciary or separate state constitutions; a Governor appointed by the centre who may reserve state legislation; and the states’ dependence on central transfers for much of their expenditure.

The critique of Wheare

• The method is legal-constitutional, and therefore static. Wheare read statutes and institutions; he did not read the working of the system, which is where federalism lives.
  • M.P. Singh presses this directly: federalism is a dynamic concept to be understood in its socio-cultural and political context, not from statutes alone.
  • The proof is internal to India. The same unamended provisions produced near-unitary behaviour under single-party dominance and markedly federal behaviour under coalition government. A method that cannot register that difference is measuring the wrong thing.
• The standard is ethnocentric. Wheare took the American constitution as the ideal type and graded every other federation by its distance from it. Granville Austin’s reply is decisive: every federation is sui generis, the product of its own circumstances, and India was never modelled on the United States.
• Even the American benchmark has not held still. Franklin Roosevelt’s New Deal produced what was itself labelled “new federalism” — a centralising turn driven by the welfare state, conditional grants-in-aid and an expansive commerce power. A model that no longer describes the United States cannot be the fixed standard by which India is called quasi-federal.
• The classification proves too much. By Wheare’s criteria Canada — residuary power at the centre, Lieutenant-Governors, powers of reservation and disallowance — would also be quasi-federal, yet nobody denies Canada is a federation.

A definition that classifies every federation but one as defective has stopped describing and started prescribing.

The rival characterisations

Observer	Characterisation	The point being made
K.C. Wheare	Quasi-federal	Prominent unitary features; subsidiary federal ones
Ivor Jennings	A federation with a strong centralising tendency	It is a federation; the tendency is degree, not kind
Paul Appleby	Extremely federal	Seen from administrative practice, India is more federal than its text suggests
W.H. Morris-Jones	Bargaining federalism	The relationship is conducted by continuous negotiation, not command
Granville Austin	Cooperative federalism, and sui generis	A strong general government need not mean weak provincial governments
Sarkaria Commission	A functional arrangement for cooperative action rather than a static institutional concept	A strong centre is essential; over-centralisation is avoidable
Balveer Arora	From quasi-federal toward quasi-confederal in the coalition decades	The states’ leverage rose until the centre’s autonomy became conditional
Louise Tillin	An asymmetric federation with an unsettled internal map	The units are a live political variable, not a fixed given
Rekha Saxena	Multi-level federalism	Two orders is incomplete; the third tier and the states’ external engagement belong in the account
Suhas Palshikar	A centralising system after 2014	The party system, not the text, sets the operative degree of federalism

“Cooperative federalism produces a strong central, or general government, yet it does not necessarily result in weak provincial governments that are largely administrative agencies for central policies.” — Granville Austin

• Austin’s two halves are separable, and the whole force of his claim lies in the second. A strong general government is compatible with provincial governments that remain governments.
  • The evidence for the second half is substantial. States have refused to implement central schemes, litigated central legislation to a standstill, legislated against central preference, withdrawn general consent to central investigating agencies, and — in the farm laws episode of 2020–21 — brought about the repeal of a central statute.
• Arora’s quasi-confederal reading describes a period rather than a permanent condition: between 1989 and 2014 the centre’s survival depended on regional parties whose price was policy concession. Palshikar’s account of the post-2014 order completes the argument from the other end — a single party with an independent majority restores the centralising tendency without a single provision being altered.

The Supreme Court’s own answers

• State of West Bengal v. Union of India (1963) — the first and most unitary answer. Upholding Parliament’s power under the Coal Bearing Areas (Acquisition and Development) Act, 1957 to acquire land owned by a state, Sinha CJ held that the Constitution is not truly federal, that the states are not sovereign in the American sense, and that no compact between independent units underlies it.
  • Subba Rao J dissented, holding that Union and states draw coordinate authority from the same instrument. The dissent is the ancestor of everything the Court later held.
• State of Rajasthan v. Union of India (1977) — arising from the dissolution of nine assemblies, the Court declined to interfere with the Article 356 proclamation and repeated that Indian federalism is not of the classical kind, while accepting that the President’s satisfaction was not wholly beyond scrutiny.
• S.R. Bommai v. Union of India (1994) — the turning point, reversing the 1963 position in principle. Federalism is a basic feature of the Constitution.
  • The consequences were operative: an Article 356 proclamation is justiciable; the material must bear a rational nexus to the action; majority is tested on the floor of the House; the assembly may be suspended but not dissolved until both Houses approve; and a dismissed government may be restored. After Bommai the Court supplies a floor below which centralisation cannot legally go.
• ITC Ltd. v. Agricultural Produce Market Committee (2002) — a Constitution Bench reading of legislative entries affirming that the essence of the arrangement is the division of powers, that entries are to be construed generously, and that Union supremacy operates only where fields genuinely overlap, not as a general presumption.
• Kuldip Nayar v. Union of India (2006) — a pull the other way. Upholding deletion of the domicile requirement for Rajya Sabha membership and the open ballot, the Court held that residence is not an essential feature of the federal structure, treating the chamber’s states-representing character as formal rather than substantive.
• The NCT of Delhi line minted the Court’s contemporary vocabulary. Government of NCT of Delhi v. Union of India (2018) described the relationship as collaborative federalism and pragmatic federalism and held the Lieutenant Governor bound by ministerial advice outside the three reserved subjects; Government of NCT of Delhi v. Union of India (2023) gave the Delhi government control over services other than public order, police and land, reasoning explicitly from asymmetric federalism. Parliament responded with an amending Act creating a National Capital Civil Service Authority, whose challenge continues.
• Mineral Area Development Authority v. Steel Authority of India (2024) and State of Uttar Pradesh v. Lalta Prasad Vaish (2024) — two nine-judge benches months apart, both by 8:1, both expanding state competence.
  • The former held that royalty under the Mines and Minerals (Development and Regulation) Act, 1957 is not a tax, and that the states’ power to tax mineral rights under Entry 50 of the State List cannot be cut down by implied limitations read into a Union entry.
  • The latter held that “intoxicating liquor” in Entry 8 covers industrial alcohol, overruling Synthetics and Chemicals (1990), and that a parliamentary declaration under Entry 52 of the Union List cannot swallow a specific state entry.
  • Read together they are the most substantial judicial correction in favour of the states since Bommai, and the method is the significant part: entries take their widest meaning, and the Union’s field is not enlarged by implication.

Since 1994 the Supreme Court has been the states’ most reliable ally, and since 2024 their most effective one.

The Federal and Unitary Features of the Constitution

Federal features	Unitary features
Written and supreme Constitution binding both levels	Strong centre: the longer Union List and residuary powers under Article 248 and Entry 97
Division of powers in the Seventh Schedule	Single citizenship — no separate state citizenship
Bicameral Parliament with the Rajya Sabha as the chamber of states	Single integrated judiciary enforcing both bodies of law
Independent judiciary as umpire, with Article 131 original jurisdiction over federal disputes	All India Services under Article 312, appointed and disciplined by the Union
Rigid amendment — the proviso to Article 368 requires ratification by half the states	Governor appointed by the centre, holding office during the President’s pleasure
Article 263 Inter-State Council and Article 280 Finance Commission as institutions of shared rule	Emergency provisions — Articles 352, 356 and 360 — converting the system to a unitary one
Article 246 giving state legislatures exclusive competence over List II	Article 3 — state areas, boundaries and names alterable without consent
Articles 301–307 guaranteeing a common economic union	Articles 200 and 201 — reservation of state bills, with no time limit prescribed
Separate state executives with independent electoral mandates	Articles 256, 257 and 365 — Union directions, non-compliance a ground for President’s Rule
Entrenched removal protection for the State Election Commissioner, on the terms applying to a High Court judge	Single Election Commission and a single Comptroller and Auditor General auditing both levels

• The two columns are not evenly weighted, and pretending otherwise is the commonest error in describing India. The unitary features operate on the federal features rather than beside them: Article 356 does not coexist with state autonomy, it suspends it; Article 3 does not coexist with the states, it constitutes and can deconstitute them.

Why the unitary features were included

• Partition was in progress while the Constitution was drafted. The Union Powers Committee had begun with a much weaker centre in contemplation and revised its position explicitly in response.
• Over five hundred princely states had to be integrated, many with no administrative link to British India and some holding out. A design giving units a veto over their own absorption would have made integration impossible.
• The “dangerous decades” thesis. Contemporary observers expected India to fragment, and the provisions that look most anti-federal — Article 3, Article 356, the All India Services, the emergency chapter — were designed for state failure and secession.
• The colonial inheritance. The Government of India Act, 1935 supplied the three-Lists structure, the Governor’s discretion, the emergency provisions and the reservation of bills. India federalised a unitary administrative state rather than assembling a federation from sovereign parts.
• The honest assessment is that these were insurance premiums, paid every year whether or not the risk materialised. Article 356 was included as a provision Ambedkar hoped would remain a dead letter; it has been invoked well over a hundred times.

The Distribution of Legislative Powers

The architecture: Articles 245 to 255

• Article 245 fixes territorial extent: Parliament legislates for the whole or any part of India, a state legislature for its own territory. No law of Parliament is invalid for extra-territorial operation, while a state law must satisfy a territorial nexus — a real and not illusory connection between the state and the object of the law.
• Article 246 allocates subject matter and states the order of precedence: List I prevails over List III, and List III over List II. Union supremacy is built into the allocation rule, not merely into the repugnancy rule.
• Article 246A, inserted by the 101st Amendment, is the structural exception: it confers simultaneous power on Parliament and state legislatures for the Goods and Services Tax, outside the List scheme, with Parliament exclusive on inter-state supply.

The three Lists and how they have moved

List	Original	Current	Character
Union List	97	98	Exclusive to Parliament — defence, foreign affairs, currency, banking, communications, inter-state trade
State List	66	59	Exclusive to state legislatures — public order, police, public health, agriculture, land, local government
Concurrent List	47	52	Both may legislate — criminal law, civil procedure, marriage, contracts, economic and social planning, education, forests

• The direction of movement has been one-way. Entries have travelled from List II to List III and from List III to List I; nothing significant has travelled back.
• The 42nd Amendment (1976) is the largest single transfer, moving five subjects from the State List to the Concurrent List: education — which is why the NEET dispute with Tamil Nadu is a Concurrent List quarrel rather than an invasion of exclusive state competence — forests, protection of wild animals and birds, weights and measures other than the establishment of standards, and the administration of justice with the constitution and organisation of all courts except the Supreme Court and High Courts. The same amendment added Entry 2A to the Union List, on deployment of Union armed forces in aid of the civil power of a state.
• The Sarkaria Commission’s recommendation that residuary subjects other than taxation move to the Concurrent List remains unimplemented, as does the Rajamannar Committee’s more radical proposal to move subjects back into List II.

The doctrines that decide competence

• Pith and substance. Where a law touches a subject in the other legislature’s List, the court asks what the law is in its true nature and character; if that falls within the enacting legislature’s competence it survives despite incidental encroachment.
• Colourable legislation. What cannot be done directly cannot be done indirectly. A law that is in substance about a subject outside the legislature’s competence, disguised as something within it, is void. The test is substance, not motive.
• Harmonious construction. Where entries appear to conflict, the court seeks a reading that gives effect to both, treating one as controlling only when reconciliation is impossible — the principle that carried the day in Lalta Prasad Vaish (2024), where a general Union entry on industries was not allowed to hollow out a specific state entry.

Repugnancy: Article 254

• On a Concurrent subject, where a state law is repugnant to a law of Parliament, the parliamentary law prevails and the state law is void to the extent of the repugnancy — which is why the Concurrent List is often described as a Union List in waiting.
• Article 254(2) is the states’ one genuine legislative shield: a repugnant state law prevails within that state if it has been reserved for the consideration of the President and received his assent. It has been used, most visibly in state amendments to central agrarian, labour and criminal legislation, but its value depends on whether the President, acting on Union ministerial advice, will assent — which converts a legislative protection into a political one.
• The shield is defeasible. The proviso to Article 254(2) preserves Parliament’s power to enact a later law adding to, amending, varying or repealing the state law. The state’s victory is always provisional.
• The doctrine of occupied field supplements this: where Parliament has legislated exhaustively on a Concurrent subject, evincing an intention to cover the whole field, a state law fails even without direct textual conflict.

The five routes into the State List

Article	Trigger	Duration and effect
249	Rajya Sabha resolution in the national interest, by two-thirds of members present and voting	One year, renewable; the law lapses six months after the resolution expires
250	National Emergency in force	For the emergency and six months after
252	Two or more state legislatures request it	Permanent transfer for consenting states; amendable or repealable only by Parliament
253	Implementation of a treaty, agreement or convention with another country	No time limit, no state consent; overrides the Seventh Schedule entirely
356	President’s Rule	Parliament exercises the state legislature’s power and may delegate it to the President

• Article 249 is the most federal of the five, because the resolution must come from the states’ own chamber and be renewed annually. It has been used sparingly.
• Article 252 is the most consensual and the most irreversible: a state that opts in cannot opt out by its own legislation but must persuade Parliament. Urban land ceiling, transplantation of human organs and water pollution legislation took this route.
• Article 253 is the most contested. It requires no consultation, and its reach has grown with the reach of international agreements themselves.
  • When treaties concerned defence, currency and diplomacy, Article 253 touched nothing in List II. Contemporary agreements reach agriculture, water, intellectual property, labour standards, public health, investment protection and climate obligations, all of which run into state subjects and state budgets.
  • The states’ demand is not for a veto but for consultation — a demand for shared rule in Elazar’s sense, since they are asked to implement obligations in whose formation they had no part. Para-diplomacy, the involvement of state governments in external policy touching their interests, is the modest institutional answer, practised informally in border and river negotiations and state investment missions without ever being institutionalised.
• Residuary power belongs to the Union. Article 248 with Entry 97 gives Parliament exclusive power over any matter not enumerated in Lists II and III, including any residuary tax — the Canadian rather than the American or Australian model. The practical weight is larger than it looks, because everything the framers could not foresee, from data to biotechnology to the digital economy, arrives in the Union’s lap by default.

The Governor’s assent: Articles 200 and 201

• Article 200 gives the Governor four courses on a bill passed by the state legislature: assent; withhold assent; return it (if not a money bill) for reconsideration, after which he shall not withhold assent if it is passed again; or reserve it for the President. Reservation is mandatory only where the bill would derogate from the powers of the High Court.
• Article 201 gives the President three: assent, withhold assent, or direct the Governor to return the bill for reconsideration within six months. The Constitution prescribes no time limit at either stage.
• The absence of a time limit is where the provision does its real work. Indefinite inaction is functionally a veto the text never confers, and because it is inaction it leaves nothing to challenge — near-identical organised-crime statutes in two states met opposite fates, one assented to promptly and the other waiting years, until a change of government at the centre.
• State of Tamil Nadu v. Governor of Tamil Nadu (April 2025) held that the Governor has no absolute veto and no pocket veto, prescribed timelines under Article 200, and — invoking Article 142 — treated ten long-pending Tamil Nadu bills as having received deemed assent. The President then made a Reference under Article 143 on 13 May 2025, placing fourteen questions on Articles 200 and 201 before the Court.
• In Re: Assent, Withholding or Reservation of Bills (20 November 2025), a five-judge bench headed by CJI B.R. Gavai, reversed the operative core of the April judgment.
  • Courts cannot prescribe timelines, because the Constitution prescribes none.
  • “Deemed assent” is contrary to the Constitution, and Article 142 cannot supply a remedy the text excludes.
  • The discharge of functions under Articles 200 and 201 is not justiciable as such — but “prolonged, unexplained and indefinite inaction” frustrating the legislative process remains open to limited judicial intervention. Two of the fourteen questions were left unanswered.
• The consequence is uncomfortable for the states. The constitutional route to compelling gubernatorial action has narrowed to a residual remedy against extreme inaction, leaving ordinary discipline to convention and political pressure — the two things that failed in the first place. A federal guarantee that depends on the good faith of a central appointee is not a guarantee.

The Economic Union: Articles 301 to 307

• Part XIII exists because a federation with internal customs barriers is a common market in name only.
• Article 301 declares that trade, commerce and intercourse throughout the territory of India shall be free.
• Article 302 lets Parliament impose restrictions in the public interest. Article 303 forbids both Parliament and state legislatures from giving preference to one state over another or discriminating between them.
• Article 304 is the states’ operative provision. A state may (a) tax goods imported from other states as similar goods produced within the state are taxed, so as not to discriminate; and (b) impose reasonable restrictions in the public interest, but only with the previous sanction of the President.

The judicial history

• Atiabari Tea Co. v. State of Assam (1961) held that a tax which directly and immediately restricts the movement of goods engages Article 301 and must satisfy Article 304(b). The freedom was read as freedom from restrictive taxation, not merely from physical barriers.
• Automobile Transport (Rajasthan) v. State of Rajasthan (1962) qualified this with the compensatory tax doctrine: a levy that is essentially a charge for facilities actually provided is not a restriction on trade at all.
  • The doctrine was attractive and unworkable. It required courts to audit whether a tax’s yield matched the facilities provided, a test for which no judicial method exists.
• Jindal Stainless Ltd. v. State of Haryana (2016) — a nine-judge bench, by 7:2, resolved five decades of confusion.
  • “Free” in Article 301 does not mean free from taxation. What is guaranteed is freedom from discriminatory and restrictive impositions, not immunity from the states’ ordinary fiscal power.
  • The compensatory tax doctrine has no constitutional basis and was discarded, overruling Automobile Transport and the earlier Jindal Stainless reference to that extent.
  • A levy violating Article 304(a) cannot be saved by complying with Article 304(b) — non-discrimination is not a procedural requirement that presidential sanction can cure.Entry tax was accordingly upheld where similar goods produced within the taxing state bear a comparable burden.
• The federal significance of Jindal Stainless is generally understated. By reading Article 301 down to a non-discrimination principle, the Court restored a wide field of state taxing power that four decades of compensatory-tax jurisprudence had held hostage to judicial audit — the same direction of travel as the 2024 nine-judge decisions.

Article 3 and the Territorial Design

• Article 2 empowers Parliament to admit into the Union, or establish, new States on such terms as it thinks fit.
• Article 3 empowers Parliament by law to form a new State by separation or union of territory, increase or diminish the area of any State, alter its boundaries, or alter its name.
• Two procedural conditions attach, and both are weaker than they look.
  • The bill may be introduced only on the recommendation of the President, which in practice means the Union Council of Ministers.
  • The President must refer the bill to the legislature of the affected State for its views within a specified period, and is not bound by those views.
• Article 4 completes the design: such a law may make consequential amendments to the First and Fourth Schedules and shall not be deemed an amendment of the Constitution for the purposes of Article 368.
  • The entire territorial map of India can therefore be redrawn by ordinary parliamentary majority, without the special majority or state ratification Article 368 would otherwise require.

The federalism objection

• No other major federation lets its centre reconstitute its units unilaterally.
  • United States: no new state may be formed within the jurisdiction of an existing one, or by junction of states, without the consent of the legislatures concerned as well as Congress.
  • Australia: alteration of a state’s limits requires that state’s parliamentary consent.
  • Canada: amendments affecting provincial boundaries need the concurrence of the province concerned.
  • Germany: reorganisation of the Länder is subject to confirmation by referendum in the affected territories.
• The objection states itself. If a federation is a contract between the general government and territorially defined units, and the general government may unilaterally redefine or abolish the units, the identity of the contracting party is at the centre’s disposal — which is hard to reconcile with the notion of a contract at all.

The defence

• The sui generis argument. Granville Austin’s position is that each federation is designed for the problem it faces, and India’s problem in 1947 was that its internal map was an artefact of conquest.
  • The provinces of British India were the residue of a century of annexation, bearing no relation to any linguistic, cultural or economic logic, and the princely states were the residue of who had made terms with the British and when.
  • A federation that entrenched colonial boundaries would have entrenched the injustice they encoded.
• The Ambedkar formulation. India is an indestructible Union of destructible States, and the destructibility of the states is the price of the indestructibility of the Union.
• Gandhi’s own position matters here: he advocated reorganisation of provinces on linguistic lines from the 1920s, and the Congress had reorganised its provincial committees on that basis long before independence — territorial flexibility was a nationalist commitment, not a centralist imposition.
• Louise Tillin’s argument in Remapping India is that the internal territorial map remains unsettled. Demands for new states have never stopped, and an order that could not respond to them would have to suppress them.
• The counterfactual is the strongest defence. Had Article 3 been rigid, the linguistic reorganisation of 1956 would have required either the consent of the states being dismembered or a constitutional amendment for each — which is to say it would not have happened.
  • What that reorganisation accomplished is the counter-intuitive point: conceding language as the organising principle of the states removed language from the national agenda, converting a potential secessionist grievance into ordinary state politics. The contrast with Pakistan, where the refusal to accommodate Bengali contributed directly to 1971, is the clearest natural experiment available.
• The concession the defence must make is that flexibility designed for accommodation can be used for coercion, and that a reference to a legislature whose views need not be heeded is no safeguard. Article 3 is defensible in what it enabled and indefensible in what it fails to constrain.

Article 3 in contemporary practice

• Andhra Pradesh, 2014. The Andhra Pradesh Reorganisation Act creating Telangana was passed after the state assembly rejected the draft bill by resolution. The reference was made, the views disregarded, the state bifurcated — the cleanest illustration of what Article 3’s consultation is worth.
• Jammu and Kashmir, 2019. The Jammu and Kashmir Reorganisation Act, 2019 downgraded a state into two union territories while it was under President’s Rule with its assembly dissolved.
  • The views required by the proviso were sought from Parliament itself, then exercising the state legislature’s powers under Article 356. The safeguard was satisfied by the very body whose decision it was meant to constrain, which shows that Article 3’s consultation can be reduced to a formality by the prior use of Article 356.
  • In Re: Article 370 (December 2023) upheld the abrogation, held Article 370 a temporary provision, and directed restoration of statehood at the earliest with elections by September 2024. Elections were held in September–October 2024; statehood has still not been restored, and petitions seeking it remain pending.
• The pattern across both episodes is the same. Article 3 has been used not to accommodate a regional demand against central preference, but to execute a central decision over regional objection — a reversal of the purpose the defence relies on.

Asymmetrical Federalism

• Asymmetrical federalism means the units of a federation do not all stand in the same relationship to the centre: some hold powers, protections or representation that others do not.
• De jure asymmetry is written into the constitution or statute — Article 371A’s protection of Naga customary law, the Sixth Schedule councils, Article 239AA for Delhi. The unit’s legal status differs.
• De facto asymmetry arises from differences in size, wealth, population and political weight that produce unequal influence within formally equal institutions. Uttar Pradesh and Sikkim have identical constitutional status and utterly different capacity to shape national outcomes.
• Charles Tarlton’s original distinction between symmetrical and asymmetrical federal systems made the point that no real federation is symmetrical, and that what matters is how a system manages the asymmetry it necessarily contains — by ignoring it, compensating for it, or constitutionalising it. India constitutionalises it.

Asymmetry of representation

• The Rajya Sabha allocates seats broadly in proportion to population, not equally. Uttar Pradesh returns 31 members; Sikkim, Nagaland, Mizoram, Tripura, Manipur, Goa and Arunachal Pradesh return one each.
• The contrast with the United States Senate is exact and deliberate. Every American state returns two senators, so Wyoming’s voters carry roughly seventy times the per-capita Senate weight of California’s.
• Alfred Stepan’s analysis of that contrast is the argument that most needs understanding.
  • A demos-constraining federation systematically over-weights small units against the national majority — through equal representation in a powerful upper house, supermajority amendment rules turning on a minority of the population, and a large policy field removed from national reach. The United States is his leading example.
  • A demos-enabling federation preserves the principle that a citizen’s vote carries the same weight wherever she lives, while protecting territorial minorities by other means. India is his leading example.
  • Stepan’s conclusion is that the Indian model suits multinational developing societies better than the American — societies that must accommodate territorial diversity while pursuing large redistributive projects a demos-constraining design would block.
• The argument has a live edge. Because Rajya Sabha and Lok Sabha strength both rest on population, the coming delimitation — the freeze on seat allocation having been extended by the 84th Amendment to the first census after 2026 — threatens to convert demographic performance into political penalty, which is why a Joint Action Committee of southern and allied states formed in March 2025.

Article 370 and the special case that ended

• Article 370 was the deepest asymmetry the Constitution contained: Parliament’s legislative power over Jammu and Kashmir was confined to defence, external affairs and communications, with the state’s concurrence required for anything more, and the state had its own constitution.
• Article 35A, inserted by presidential order in 1954, empowered the state to define permanent residents and reserve rights in employment, property and settlement to them.
• Both were rendered inoperative in August 2019 by a presidential order under Article 370(1) applying the whole Constitution to the state, together with the Reorganisation Act; In Re: Article 370 (2023) upheld the route taken.
• What matters analytically is what the change substituted: an asymmetric federal accommodation was replaced by a centralised administrative arrangement, and the open question is whether ordinary federal politics can now do the integrative work special status was designed to do.

The Article 371 family

Article	State	Provision
371	Maharashtra and Gujarat	Governor’s special responsibility for development boards for Vidarbha, Marathwada, Saurashtra and Kutch, with equitable allocation of funds and equitable arrangements for technical education and employment
371A	Nagaland	No Act of Parliament applies to Naga religious or social practices, Naga customary law and procedure, administration of civil and criminal justice involving Naga customary law, or ownership and transfer of land, unless the Assembly so decides
371B	Assam	The President may provide for a committee of the Assembly of members elected from the tribal areas
371C	Manipur	A similar committee for the hill areas; the Governor reports annually to the President on their administration
371D	Andhra Pradesh and Telangana	Equitable opportunities in public employment and education across regions; local cadres; an Administrative Tribunal for service disputes
371E	Andhra Pradesh	Parliament may establish a central university in the state
371F	Sikkim	Assembly of at least 30 members, with seats reservable for sections of the population; pre-merger matters immunised from challenge
371G	Mizoram	Assembly of at least 40 members; no Act of Parliament applies to Mizo religious or social practices, customary law, justice according to it, or land, unless the Assembly so decides
371H	Arunachal Pradesh	Assembly of at least 30 members; Governor’s special responsibility for law and order in his individual judgment
371I	Goa	Assembly of at least 30 members
371J	Karnataka	Hyderabad-Karnataka region — development board, equitable allocation of funds, and reservation in education and state posts for persons of the region

• Article 371A is the strongest asymmetry now in the Constitution. On four enumerated subjects an Act of the sovereign Parliament of India does not apply to Nagaland unless the Nagaland Assembly resolves that it shall — a state veto over central legislation, which classical accounts say Indian federalism does not contain.
• These provisions were not enacted together. They accumulated between 1956 and 2012, each answering a specific regional grievance or accession settlement, which is exactly what a holding-together federation looks like from the inside.

The Fifth and Sixth Schedules

• The Fifth Schedule governs Scheduled Areas and Scheduled Tribes in states other than Assam, Meghalaya, Tripura and Mizoram.
  • The Governor may by notification direct that an Act of Parliament or of the state legislature shall not apply to a Scheduled Area, or shall apply with exceptions and modifications; a Tribes Advisory Council advises on tribal welfare; the Governor reports annually to the President.
  • The Panchayats (Extension to the Scheduled Areas) Act, 1996 extends panchayat governance into these areas with adaptations recognising the Gram Sabha’s authority over minor forest produce, land alienation and local plans.
• The Sixth Schedule governs the Tribal Areas of Assam, Meghalaya, Tripura and Mizoram through Autonomous District and Regional Councils.
  • They hold genuine legislative power — land other than reserved forest, management of forests, shifting cultivation, village and town administration, inheritance, marriage and divorce, and social custom — subject to the Governor’s assent.
• The limit is real and is the standard criticism. Where a District Council’s law conflicts with the state legislature’s, the state law prevails, and the councils’ financial dependence is near-total. Ten such councils now operate in the North-East, and Ladakh’s principal constitutional demand is Sixth Schedule inclusion alongside statehood.

Union territories and the case of Delhi

• Union territories are administered by the President through an Administrator and are not states at all.
• Article 239AA creates a unique intermediate category for the National Capital Territory of Delhi: a Legislative Assembly competent over the State and Concurrent Lists except public order, police and land, a Council of Ministers responsible to it, and a Lieutenant Governor otherwise bound by their advice.
  • The arrangement has generated the largest body of federalism litigation of the last decade, running from the 2018 judgment on the elected government’s primacy through the 2019 split verdict on services to the 2023 Constitution Bench, and then to the parliamentary response creating the National Capital Civil Service Authority, now itself under challenge.

Fiscal asymmetry

• Special Category Status was an asymmetry of transfers rather than of powers.
  • It originated in 1969 with the Gadgil formula approved by the National Development Council, and applied to states meeting criteria of hilly terrain, low population density or substantial tribal population, strategic border location, backwardness and non-viable finances.
  • Eleven states held it — the North-Eastern states with Himachal Pradesh, Uttarakhand and the erstwhile state of Jammu and Kashmir —with a far more favourable grant-to-loan ratio in central assistance.
  • It effectively lapsed after the Fourteenth Finance Commission, which treated the category as unsustainable and raised devolution to all states to 42% instead: a shift from categorical to formulaic redistribution.
  • The demand persists — from Bihar, from Andhra Pradesh on the basis of assurances given at bifurcation, and from Odisha — — formulaic devolution distributes but does not target.
• Asymmetry survives inside the Finance Commission’s horizontal formula, which is redistributive by design. Income Distance transfers from richer to poorer states; Area, Forest and Ecology and Demographic Change each compensate a different structural disadvantage.
  • The Sixteenth Finance Commission, reporting for 2026–31, held vertical devolution at 41% against eighteen states demanding 50%, and weighted Income Distance 42.5%, Population (2011) 17.5%, Demographic Performance 10%, Forest and Ecology 10%, Area 10%, adding a new criterion — the State’s contribution to GDP, 10%. That last is the first efficiency criterion of its kind, and a partial retreat from pure equalisation.

Asymmetry as an integrative device

• Asymmetry is not a defect in Indian federalism but its principal method. Every major accommodation — Nagaland in 1963, Sikkim in 1975, Mizoram after 1986, the Sixth Schedule councils, Article 371D for the Telangana region within undivided Andhra Pradesh, 371J for Hyderabad-Karnataka — took the form of a bespoke arrangement rather than a general rule.
• A federation that must treat all units identically can only respond to a regional demand by conceding it to everyone or refusing it to the claimant. India can concede it to one.
• The cost of the method is coherence. An order composed of exceptions is harder to administer, invites the complaint that concessions go to whoever presses hardest,.
• The benefit is that the union has held. The Soviet Union, Yugoslavia, Czechoslovakia and Pakistan all fractured; India, with more languages and more religious diversity than any of them, has lost no territory to secession.

The more the constitution bends, the less it breaks.

India and the United States: A Structural Comparison

Dimension	United States	India
Origin	Coming-together — sovereign states federating by bargain	Holding-together — a unitary state devolving to accommodate diversity
Direction of force	Centrifugal	Centripetal
Relation between levels	Dual federalism — watertight compartments	Cooperative federalism — interlocking
Locus of administration	Legislative — each level administers its own laws	Executive — states administer much Union law
Status of units	Symmetrical	Asymmetrical
Constitutions	Two sets — federal and fifty state constitutions	One (Jammu and Kashmir’s, now gone, was the exception)
Citizenship	Dual	Single
Judiciary	Dual court hierarchies	Integrated — one hierarchy for both bodies of law
Residuary power	With the states (Tenth Amendment)	With the Union (Article 248, Entry 97) — the Canadian model
Upper house	Senate — two per state, regardless of population	Rajya Sabha — seats broadly proportionate to population
Stability of units	Indestructible Union of indestructible States	Indestructible Union of destructible States
Amendment	Two-thirds of both Houses or a convention, then ratification by three-fourths of the states	Special majority in Parliament, with ratification by half the states only for enumerated federal provisions
Emergency displacement of units	No equivalent to Article 356	Articles 352, 356 and 360
Territorial alteration	Requires the consent of the legislatures concerned	Article 3 — views sought, not binding
Stepan’s category	Demos-constraining	Demos-enabling

• Canada is the more illuminating comparison, and the neglected one: residuary power at the centre through the “Peace, Order and Good Government” clause, Lieutenant-Governors appointed by the federal government with powers of reservation and disallowance over provincial legislation — the direct ancestor of Articles 200 and 201 — and cooperative federalism as the self-description.
  • The difference is that Canada’s disallowance power has fallen into desuetude by convention while India’s equivalent remains in continuous use, which is the clearest illustration that identical text does different work in different political cultures.
• Australia is a coming-together federation with an equal-state Senate and residuary powers with the states, yet its High Court’s expansive reading of the Commonwealth’s taxation and corporations powers has produced a fiscal centralisation as pronounced as India’s — evidence that centralisation is not a function of constitutional design alone.
• Germany supplies administrative federalism at its purest and the institution India most conspicuously lacks: a Bundesrat in which the governments of the Länder are directly represented and hold a veto over legislation affecting them.
  • The Rajya Sabha is elected by state legislators but its members owe no duty to the state government and vote on party lines. India has a chamber named for the states and no chamber of state governments — the gap the Inter-State Council and the GST Council partially and imperfectly fill.

Cooperative, Competitive and Bargaining Federalism

Cooperative federalism as design

• Cooperative federalism rests on interdependence rather than independence, and the difference from the dual model is architectural. In the dual model each government is complete in itself; in the cooperative model each is deliberately made incomplete, so that neither can achieve its purposes without the other. The interlocking is the safeguard.
• In India the interdependence runs both ways, which is the point most often missed.
  • States depend on the centre financially, raising roughly a third of general government revenue and spending roughly two-thirds of it, the gap bridged by devolution and grants.
  • The centre depends on the states administratively. It has no district administration of its own; Union laws on health, agriculture, food, education, elections, disaster management and policing are executed by state machinery.
  • This is why India is an executive or administrative federation rather than a legislative one like the United States. The state’s constitutional weight shows up in implementation more than in law-making, and a state that will not implement can obstruct effectively without winning a legal argument.
• The institutional expressions are the Concurrent List itself, Article 263 and the Inter-State Council, the Zonal Councils, Article 280 and the Finance Commission, the GST Council under Article 279A, and the NITI Aayog’s Governing Council.
• The standing criticism is that these institutions are consultative rather than co-decisional. The Inter-State Council has met roughly eleven times in thirty-five years, with a full decade between its tenth and eleventh meetings; the Zonal Councils have never had a water dispute referred to them; the NITI Aayog’s Governing Council advises without deciding.
  • The GST Council is the exception, and the only genuinely co-decisional federal institution India has: the Union holds one-third of the weighted votes, all states together two-thirds, and three-quarters is required to pass — so neither side can act alone, though the arithmetic gives the Union an effective veto that no combination of states possesses.
  • Union of India v. Mohit Minerals (2022) held the Council’s recommendations persuasive and not binding, and Parliament’s and the state legislatures’ power on GST simultaneous — restoring state autonomy in principle while leaving the fiscal reality untouched.

Bargaining federalism as practice

• M.P. Singh’s formulation is the sharpest summary of the gap between design and operation: cooperative federalism is the idea; bargaining federalism is the practice. The distinction concerns the posture of the centre, not the volume of interaction.

	Cooperative federalism	Bargaining federalism
Basis of the relationship	Trust and consultation	Relative bargaining strength
The centre’s role	Friend, philosopher and guide	Bullying elder brother or patriarch
What determines outcomes	The merits of the question	Who needs whom more
The states’ position	Partners in a joint enterprise	Supplicants with variable leverage
Institutional form	Deliberation in shared bodies	Negotiation, often bilateral and informal

• What raised the states’ bargaining power was the regionalisation of the party system, not any constitutional change. Between 1989 and 2014 no party held a majority of its own at the centre, and regional parties whose base was a single state held the balance; the price of their support was policy.
• M.P. Singh’s two-axis model explains the oscillation without invoking any change in the text. The parliamentary axis denotes a strong centre and dominates when a single party holds an independent majority; the federal axis denotes strong states and dominates under coalition government.
  • The same Constitution therefore behaves unitarily and federally in different decades, which is the decisive refutation of any purely textual classification, Wheare’s included.
• The corollary is uncomfortable and true: in India there are fewer centre–state disputes than party-to-party disputes. Friction is largely a function of whether the same party governs at both levels.

Competitive federalism

• Competitive federalism is a neo-liberal reframing built on “minimum government, maximum governance”. As the state’s developmental functions are rolled back, private investment fills the vacuum; investment is mobile; so states must compete on policy, regulation and infrastructure to attract it, and that competition disciplines state governments more effectively than central direction did.
• Its instruments are rankings and incentives rather than transfers — ease-of-doing-business rankings of states, the Aspirational Districts Programme, sectoral indices, and reform-linked additions to state borrowing limits.
• Balveer Arora’s objection is the necessary corrective: competition between unequally endowed units is not fair competition.
  • States begin from radically different positions in infrastructure, human capital, institutional capacity and geography. A coastal state with a port and a literate workforce will win an investment contest against a landlocked state without them, however well the latter governs.
  • Competitive federalism therefore cannot substitute for cooperative federalism; it can only sit on top of it, and it needs a redistributive floor to be legitimate. That floor is what the Finance Commission’s equity criteria supply and what the drift toward efficiency criteria erodes.
• A further objection concerns discretion. Replacing formula-based allocation with discretionary and scheme-linked transfers enlarges the centre’s room to reward and punish, converting competition between states into competition for central favour — the opposite of what the concept promises.

Collaborative and pragmatic federalism

• The NCT of Delhi line contributed two further terms, and they are doctrinal rather than decorative.
  • Collaborative federalism describes a relationship in which the two levels must work toward a shared constitutional purpose, so that a functionary of one may not simply obstruct the other’s functioning.
  • Pragmatic federalism describes a method of interpretation: where the text is ambiguous, the reading that permits both governments to function is preferred over the reading that produces deadlock.
• The value of these terms is that they are justiciable in a way “cooperative federalism” never was.

Criticisms and Defence of the Design

The case against

• The design is centralising and the text admits it: the longer List and the residue with the Union, five routes into the State List, Article 254 giving the Union the last word on the Concurrent List, and Article 253 removing the external sphere from the division entirely.
• The units are not secure in their own existence, since Article 3 permits Parliament to redraw or abolish them by simple majority with a consultation requirement that binds nobody.
• An elected state government can be dismissed by the centre. Article 356 has been invoked well over a hundred times, overwhelmingly against governments opposed to the party in office at the centre, and the Bommai discipline arrived only after four decades.
• The Governor is a structural problem, not a personnel problem — a central appointee holding office during the President’s pleasure, with untimed discretion over the state’s legislation and a role in forming its government. The November 2025 advisory opinion narrowed rather than widened the remedy against inaction.
• Fiscal dependence undercuts every formal autonomy. The growth of cesses and surcharges outside the divisible pool has shrunk the shareable base to roughly 81% of gross tax revenue, from 89.1% in 2014–15, so the notional 41% devolution delivers materially less than it states.
• The instruments of shared rule are weak. The Inter-State Council meets rarely; the Zonal Councils have decided nothing; the Rajya Sabha does not represent state governments, and after Kuldip Nayar need not represent state residents.
• The pressure has migrated rather than disappeared. As the constitutional route narrowed after Bommai, the instruments shifted to fiscal design, centrally sponsored schemes with matching requirements, central agencies, and the Governor’s office.

The case for

• The design was fitted to the problem. A constitution drafted during Partition, amid the integration of five hundred princely states and a widely predicted fragmentation, could not have been written on the American template and survived.
• The record is the strongest argument. Every comparable multinational state that entrenched unit sovereignty dissolved; Pakistan lost its eastern wing within twenty-four years. India has held.
• Accommodation worked where force alone did not. Linguistic reorganisation defused language as a national grievance; the Mizo Accord of 1986 ended an insurgency military action had not; Article 371A gave Nagaland a shield stronger than anything an American state possesses.
• The states are not administrative agencies, which is Austin’s precise claim. They legislate independently over policing, land, agriculture and health, command the machinery through which most Union policy must pass, have compelled the repeal of central legislation, and litigate successfully — as the 2024 nine-judge decisions demonstrate.
• The judiciary supplies a floor. Since Bommai, federalism is a basic feature, so the centralising tendency has a legal limit no parliamentary majority can lower.
• Asymmetry has done what uniformity could not. The Article 371 family, the Fifth and Sixth Schedules and Article 239AA are accommodations a purely centralising account of the Constitution cannot explain at all.

The Indian Constitution does not guarantee federalism; it makes federalism possible and leaves the rest to politics.

Conclusion

• India is a federation, and Wheare was wrong — not because his observations were inaccurate but because his method could not accommodate what it observed. He measured a holding-together federation against a coming-together template and reported the difference as a deficiency.
• The accurate description is a centralised federation whose operative degree of federalism is set politically, within limits the judiciary has fixed since 1994. The design is centralised, the practice is political, the floor is judicial.
• The classification matters because the answer is operative. After Bommai, “federalism is a basic feature” is not a description but a limit — it is why an Article 356 proclamation is reviewable, why an assembly cannot be dissolved before parliamentary approval, and why the argument for extending justiciability to assent timelines and to the divisible pool has constitutional purchase.
• The most instructive recent development runs against the received account: two nine-judge benches in 2024 expanded state legislative and taxing competence, while the November 2025 advisory opinion narrowed the states’ remedy against the Governor. The Court is neither uniformly centralising nor uniformly federalising — it is enforcing a text that points both ways.
• Where the design is genuinely open to attack is not its strong centre but its weak institutions of shared rule. Self-rule in India is constitutionally secure and judicially defended; shared rule rests on a chamber that does not represent state governments, a council that meets once every three years, and a fiscal architecture the states have no vote in designing. The GST Council is the one exception, and also the one forum where states have extracted real concessions.
• The centralised design was justified by the circumstances of its adoption and vindicated by the union’s survival — but the justification has expired for the parts that were insurance against fragmentation, and the unfinished business is not the transfer of subjects from one List to another. It is the construction of institutions in which the states decide with the centre rather than being consulted by it.