On 11 December 2023 a Constitution Bench of five judges of the Supreme Court delivered judgment in In Re: Article 370 of the Constitution, reported at [2023] 16 SCR 1 and cited as 2023 INSC 1058. The Article 370 Supreme Court judgment answered a question the constitutional text had carried unanswered for over seven decades: whether the provision that had governed the relationship between India and the State of Jammu and Kashmir since 1950 could be brought to an end by the very mechanism the provision itself supplied, and whether the manner of its ending in August 2019 survived scrutiny. The Bench was unanimous in result. It produced three opinions, one for the majority of the Bench and two separate concurrences, and it decided some questions with great precision while expressly declining to decide two others.

The interpretive difficulty at the centre of the case was genuine, and it is worth stating plainly before any holding is described. Article 370 contained its own ending. Its third clause empowered the President, by public notification, to declare that the article would cease to be operative, but the proviso to that clause required the recommendation of the Constituent Assembly of the State before such a notification issued. That Assembly had framed the Constitution of Jammu and Kashmir, adopted it on 17 November 1956, and dissolved on 25 January 1957. No body answering its description had existed for over six decades when, in August 2019, the President issued two Constitutional Orders that applied the entire Constitution of India to the State and then declared Article 370 inoperative in its existing form. The petitioners said the proviso had frozen the provision in place when the Assembly disappeared, because the one consent the text demanded could never again be given. The Union said the power survived, because a recommendation is not a veto and a temporary body cannot entrench a temporary provision against the authority the clause itself names. The text alone did not settle the contest, because the text had been written for a Constituent Assembly expected to complete the State’s constitutional settlement, not for a dispute arising sixty-six years after that Assembly went home.

This article reports the judgment. It states each holding separately, in the order of the questions the Court framed, and it keeps the three opinions distinct, because on one question, whether the State retained an element of sovereignty after accession, the lead opinion and one concurrence reasoned to the same destination by opposite roads. It states the one thing the Court held beyond power, a single paragraph of the first Constitutional Order, with the same exactness, because secondary accounts divide into two camps on this point and both camps are wrong: the first Order was neither upheld in its entirety nor struck down in its entirety. It records the two questions the Court expressly declined to decide, because those silences are part of the judgment’s legal content and will govern how the decision is used in future reorganisation disputes. And it presents the criticisms of the judgment at full strength, in the terms in which named critics have made them, alongside the answers the judgment itself gives, adjudicating neither side.

The provision and the ambiguity in it

Article 370 stood in Part XXI of the Constitution, the part headed “Temporary, Transitional and Special Provisions”, and its own marginal heading described it as a temporary provision with respect to the State of Jammu and Kashmir. It had been moved in the Constituent Assembly of India on 17 October 1949 by N. Gopalaswami Ayyangar as draft Article 306A, in the circumstances of a ceasefire, a reference of the Kashmir question to the United Nations, and a commitment that the people of the State would in due course express their will through a Constituent Assembly of their own. Ayyangar told the Assembly that the arrangement was an interim one, made necessary by the abnormal conditions in the State, and that the will of the people, expressed through their Constituent Assembly, would determine the State’s final constitutional position. The provision that emerged from that debate had three moving parts, and the litigation of 2023 turned on the relationship among all three.

The first clause distributed legislative and constitutional authority. Under clause (1)(a), the other provisions of Article 370 itself applied notwithstanding anything else in the Constitution. Under clause (1)(b), Parliament’s power to legislate for the State was confined, in the first instance, to the matters declared in the Instrument of Accession, defence, external affairs, communications and ancillary subjects, on which the President acted in consultation with the State Government, and to such other matters in the Union and Concurrent Lists as the President might specify with the concurrence of the State Government. Under clause (1)(d), the remaining provisions of the Constitution applied to the State with such exceptions and modifications as the President might by order specify, with the same division of procedure: consultation for orders touching the accession subjects, concurrence for orders touching anything else. The second clause dealt with the situation in which concurrence was given before the State’s Constituent Assembly convened, requiring such concurrence to be placed before that Assembly. The third clause was the termination mechanism described above, and its proviso made the recommendation of the Constituent Assembly of the State necessary before the President issued a notification declaring the article inoperative or operative only with exceptions and modifications.

What is the difference between consultation and concurrence in Article 370?

Consultation under clause (1)(b) required the President to seek the views of the State Government but not to obtain its agreement, while concurrence required the State Government’s actual consent before the President could act. The distinction structured the whole provision: the accession subjects could be reached after consultation, and everything beyond them needed concurrence.

Article 367, the interpretation clause of the Constitution, supplied the second textual element. It provides, in substance, that references in the Constitution are to be construed in accordance with the General Clauses Act, 1897, adapted for the purpose, and it had been adapted for Jammu and Kashmir repeatedly across the decades: it was through modifications of the kind clause (1)(d) authorised that references to the Sadr-i-Riyasat of the State came to be read as references to the Governor, upheld by the Court in Mohd. Maqbool Damnoo v. State of Jammu and Kashmir in 1972. The question that reached the Bench in 2023 was whether that interpretive machinery could be used, not to construe a reference, but to rewrite the proviso to clause (3) itself, substituting the Legislative Assembly of the State for the Constituent Assembly of the State as the body whose recommendation the proviso required. Every other question in the case stood downstream of the historical ones, but this one was a pure question about the permissible uses of an interpretation clause, and it is the question on which the Court’s answer has the longest doctrinal reach beyond the facts.

The third textual element was Article 3, which empowers Parliament by law to form new states and to alter the areas, boundaries and names of existing states, and whose Explanation I provides that the word “State” in clauses (a) to (e) of that article includes a Union territory. The State of Jammu and Kashmir had been converted, by the Jammu and Kashmir Reorganisation Act, 2019, into two Union territories, one bearing the State’s name and equipped with a legislature, the other, Ladakh, without one. Whether Article 3 permits the conversion of a state into Union territories at all, as distinct from the formation of a Union territory out of a state’s separated territory, was placed before the Bench in terms. Article 356 completed the set of provisions in issue: the August 2019 Orders were issued while a Proclamation under that article was in force, the Legislative Assembly stood dissolved, and Parliament was exercising the powers of the State Legislature under Article 357. The validity of what was done therefore depended in part on the permissible scope of action during President’s Rule, a body of doctrine the Court had built in S. R. Bommai v. Union of India and Rameshwar Prasad v. Union of India.

The drafting of Article 370 and the settlement it recorded

The Bench treated the drafting record as evidence, and the record repays attention because both sides built their cases on it. When Ayyangar moved draft Article 306A on 17 October 1949, he explained to the Constituent Assembly why the State’s position could not be assimilated to that of the other acceding states. A war with Pakistan over the State’s territory had ended in a ceasefire supervised by the United Nations, the State’s government functioned under abnormal conditions, and the Indian commitment, recorded in correspondence with the United Nations and in public statements, was that the final shape of the State’s relationship with India would be settled by the will of its people expressed through a Constituent Assembly of the State. The draft article, he said, was an interim arrangement to give constitutional form to the accession until that settlement occurred. Two features of the speech mattered to the 2023 argument. Ayyangar described the provision’s subject matter as transitional in character, which the Union later read as the framers’ own classification of the article. And he described the role of the State’s Constituent Assembly in terms that stopped short of sovereignty: the Assembly would determine the extent to which the Constitution of India would apply and the form of the State’s own Constitution, within a relationship the accession had already created.

The petitioners’ drafting argument ran in the opposite direction from the same material. The Constituent Assembly of India, they submitted, had deliberately withheld from the President and from Parliament any unilateral power to end the arrangement, reserving the initiative to the State’s Constituent Assembly precisely because the settlement was a compact whose terms only the State’s constituent body could release. On this submission, the proviso to clause (3) is not a procedural detail but the settlement’s core security, and a court should be slow to read a security of that kind out of existence on the strength of the article’s placement among transitional provisions. The lead opinion’s answer, developed across its historical chapters, was that the debates show a provision made for a situation expected to resolve itself, drafted by men who assumed the State’s Constituent Assembly would meet, decide, and disperse, and who cannot be taken to have intended that body’s silence to govern the relationship for ever. The opinion read the drafting record together with the events of the following decade: the Delhi Agreement of 1952 between the Union and the State Government, under which the hereditary rulership was abolished and the elected Sadr-i-Riyasat substituted, the State Constituent Assembly’s adoption of a Constitution declaring the State an integral part of the Union, and that Assembly’s dissolution without any step under clause (3). Integration, on the Court’s reading of the record, was the direction of travel the settlement itself contemplated, and clause (3) was the instrument by which the travel would formally conclude.

One further drafting point recurs in the judgment and is worth isolating, because it explains why the sovereignty argument failed in the lead opinion despite the accession’s limited terms. The Instrument of Accession was in the standard form executed by the rulers of the acceding states; it was not a treaty negotiated between two continuing sovereigns, and clause 8 of the Instrument, which preserved the ruler’s sovereignty and powers except as the Instrument provided, preserved them within the constitutional order the accession joined, not against it. The proclamation of 25 November 1949 then directed that the Constitution of India would govern the relationship, and the State’s own Constituent Assembly, the body the petitioners treated as the repository of the State’s constituent authority, framed a Constitution whose third section declared the State to be an integral part of the Union of India and whose scheme accepted the supremacy of the Constitution of India in the fields it covered. The lead opinion treated this sequence as a completed constitutional settlement rather than a standing negotiation, and its sovereignty holding follows from that characterisation.

The road from accession to the Constitution Bench

The State’s accession was executed by Maharaja Hari Singh on 26 October 1947 and accepted the following day, and it covered defence, external affairs and communications, with ancillary matters. On 25 November 1949 the Yuvraj, Karan Singh, issued a proclamation directing that the Constitution of India, when adopted, would govern the constitutional relationship between the State and the Union, and that the Government of India Act, 1935 would stand repealed in its application to the State. The lead opinion in the 2023 judgment treated that proclamation, with the Instrument of Accession, as the point at which any argument for retained sovereignty had to be tested, and much of its historical reasoning is a reading of the sequence from accession to proclamation to the adoption of the State’s own Constitution seven years later.

The first Presidential Order under Article 370, the Constitution (Application to Jammu and Kashmir) Order, 1950, applied a limited set of constitutional provisions to the State. The Order of 1954, issued with the concurrence of the State Government after the Delhi Agreement of 1952, extended the application of the Constitution very substantially, inserted Article 35A into the constitutional scheme as it applied to the State, and defined the permanent resident regime that survived, with additions and challenges, until 2019. Across the following decades the President issued a long series of orders under clause (1) applying further provisions of the Constitution to the State, a process the petitioners described as erosion of the original settlement and the Union described as progressive integration, and which the lead opinion later characterised, in a phrase that carries its fifth holding, as a gradual process of constitutional integration whose culmination the 2019 declaration was.

The State’s Constituent Assembly, elected in 1951, framed the Constitution of Jammu and Kashmir, adopted it on 17 November 1956, and dissolved on 25 January 1957, the day before that Constitution came into force. Before dissolving, it had made no recommendation under the proviso to Article 370(3) that the article cease to operate. It had, on the contrary, proceeded on the footing that the State’s constitutional relationship with the Union would continue to be governed by Article 370, and the State Constitution it produced declared, in its Section 3, that the State is and shall be an integral part of the Union of India, while Section 147 of that Constitution barred amendments affecting the relationship. The rival readings of this history defined the parties’ positions in 2023. For the petitioners, the Constituent Assembly’s silence was the point: the one body authorised to recommend termination had dispersed without recommending it, and the provision had thereby become permanent by exhaustion of its own termination route. For the Union, the Assembly’s silence was legally beside the point, because a power expressed to depend on a recommendation is not extinguished when the recommending body, itself temporary, ceases to exist without making one.

The immediate sequence began on 19 December 2018, when the President issued a Proclamation under Article 356 in relation to the State, following a period of Governor’s rule under Section 92 of the State Constitution from 20 June 2018 and the Governor’s dissolution of the Legislative Assembly on 21 November 2018. The Proclamation was approved by Parliament and extended in July 2019. On 5 August 2019 the President issued the Constitution (Application to Jammu and Kashmir) Order, 2019, numbered Constitutional Order 272, in exercise of the power under clause (1)(d) and with the concurrence of the Governor, who was then discharging the functions of the State Government under the Proclamation. That Order superseded the 1954 Order, applied all the provisions of the Constitution of India to the State, and, by its Paragraph 2, added a clause (4) to Article 367 in its application to the State, whose operative effect for the litigation was to substitute the Legislative Assembly of the State for the Constituent Assembly of the State in the proviso to Article 370(3). On the same day and the day following, Parliament, exercising the powers of the State Legislature under Article 357, adopted resolutions recommending that the President issue the declaration under clause (3). On 6 August 2019 the President issued Constitutional Order 273, declaring that all the clauses of Article 370 would cease to be operative except clause (1), itself restated so that all provisions of the Constitution, as amended from time to time, would apply to the State. Parliament then enacted the reorganisation statute, which received assent on 9 August 2019 and took effect on the appointed day of 31 October 2019; the statute itself, its schedules and its machinery are the subject of the article on the Jammu and Kashmir Reorganisation Act 2019, which owns that statute’s questions in this series.

The challenge was brought by writ petitions filed by political parties, former legislators, civil society organisations and individuals, and a preliminary question occupied a Constitution Bench in March 2020: whether the earlier decisions in Prem Nath Kaul v. State of Jammu and Kashmir and Sampat Prakash v. State of Jammu and Kashmir pulled in different directions and required a reference to a bench of seven judges. The Bench of five, in Dr Shah Faesal v. Union of India, held that no such reference was needed, and the merits were heard by the Bench that eventually decided the case, differently composed in part, across sixteen days of argument in August and September 2023. Judgment was reserved on 5 September 2023 and delivered on 11 December 2023.

Why did an earlier bench refuse to send the Article 370 questions to a larger bench?

The Constitution Bench in Dr Shah Faesal held in March 2020 that Prem Nath Kaul and Sampat Prakash did not conflict in a manner requiring a reference to seven judges, because the two decisions addressed different aspects of the provision and the questions in the challenges could be decided by a bench of five.

The Bench that decided the case comprised the Chief Justice, Dr D. Y. Chandrachud, and Justices Sanjay Kishan Kaul, Sanjiv Khanna, B. R. Gavai and Surya Kant. The Chief Justice wrote the lead opinion for himself and for Justices Gavai and Surya Kant. Justice Kaul wrote a separate opinion concurring in the result, and Justice Khanna wrote a short separate opinion concurring with both of his colleagues, adding particulars of his own. The result was unanimous. The reasoning was not uniform, and the differences matter at two points: on sovereignty, where Justice Kaul’s route runs through ground the lead opinion expressly rejects, and on the conversion of a state into Union territories, where Justice Khanna recorded a caution in terms the lead opinion did not use.

Which lawyers argued the Article 370 case for the petitioners and for the Union?

The petitioners were led by senior advocates including Kapil Sibal, Gopal Subramanium, Rajeev Dhavan and Dushyant Dave, arguing in support of the challenges across the sixteen days of hearing. The Union was led by the Solicitor General, Tushar Mehta, with the Attorney General, R. Venkataramani, also appearing in support of the Orders and the reorganisation.

The questions the Court framed

The Court organised its task into eight questions, and the architecture of the judgment follows them, so a reader who holds the eight questions in mind can place every holding. The first asked whether the provisions of Article 370 were temporary in nature or had acquired a status of permanence. The second asked whether the amendment of Article 367 in exercise of the power under clause (1)(d), substituting the Legislative Assembly for the Constituent Assembly in clause (3), was constitutionally valid. The third asked whether the entire Constitution of India could have been applied to the State in exercise of the power under clause (1)(d). The fourth asked whether the abrogation of Article 370 by the President under clause (3) was invalid for want of the Constituent Assembly’s recommendation. The fifth and sixth concerned the Governor’s proclamation of 20 June 2018 under Section 92 of the State Constitution, the dissolution of the Legislative Assembly under Section 53(2) of that Constitution on 21 November 2018, and the Proclamation under Article 356 of 19 December 2018 with its extensions. The seventh concerned the Reorganisation Act, tested against the two provisos to Article 3. The eighth asked whether, during a Proclamation under Article 356 and with the Assembly dissolved, the conversion of the State into Union territories was a valid exercise of power. The judgment’s disposition of these questions is uneven by design: five holdings of substance, one holding of invalidity embedded inside a question whose larger answer favoured the Union, and two questions the Court in the end did not answer.

How the Article 370 Supreme Court judgment is organised

A word on the document itself will save researchers time. The lead opinion is long and its conclusions are gathered in a single passage, at paragraph 514, which states the holdings seriatim and is the passage from which the headnote in the Supreme Court Reports is drawn; a reader who wants the Court’s answers without its reasoning can begin there, though this article has followed the reasoning because the silences are invisible in the conclusions alone. Justice Kaul’s concurrence states its own conclusions at paragraph 112 and contains the epilogue and the recommendation at its close, around paragraph 120. Justice Khanna’s opinion is a few pages in length and states its points in numbered paragraphs, the operative summary standing at its paragraph 2 and the observations on conversion at paragraph 6. The report of the case in the Supreme Court Reports, [2023] 16 SCR 1, prints a headnote that distinguishes the three opinions holding by holding, and for this judgment the headnote is unusually reliable, because the case turns on differences of reasoning that the headnote preserves rather than merges.

The arguments, stated as the Court recorded them

The petitioners’ case was constructed in five layers, and the judgment’s structure is a reply to each. The first layer was constituent: the accession was limited, the State retained internal sovereignty, and Article 370 was a compact whose termination the compact itself reserved to the State’s Constituent Assembly. The second layer was textual and temporal: whatever the position in 1950, the Constituent Assembly’s dissolution without a recommendation had closed the clause (3) route, and the provision had acquired permanence by the exhaustion of its own machinery. The third layer was the colourable device: Constitutional Order 272’s amendment of Article 367 achieved by interpretation what only clause (3) or a constitutional amendment could achieve, and was therefore void. The fourth layer was federal: the consents and recommendations that punctuate the 2019 sequence were supplied by the Governor and by Parliament, so that the Union in substance consented to itself, during a Proclamation whose own validity and timing were open to question, and irreversible changes were made in a period of suspended democracy in the State. The fifth layer was the reorganisation: Article 3 empowers Parliament to form states and alter their boundaries, but not to extinguish a state’s statehood by converting it into Union territories, still less to do so on the strength of views expressed by Parliament in the State Legislature’s place.

The Union’s case, led by the Solicitor General, answered layer for layer. The accession was complete and unconditional in law, the proclamation of November 1949 adopted the Constitution of India as the governing instrument, and the State’s Constituent Assembly itself declared the State an integral part of the Union, so no sovereignty survived to be compacted away. Article 370 was temporary by text, by placement and by the declared intention of its mover. The recommendation in the proviso was recommendatory, the Assembly was a temporary body, and Sampat Prakash had settled the survival of the President’s powers two generations earlier. Constitutional Order 272 applied the Constitution, which clause (1)(d) plainly permitted, and the application of all provisions in one order was the exercise of a discretion as to extent, not an abuse. During President’s Rule, Articles 356 and 357 place the State Legislature’s powers, legislative and otherwise, in Parliament, and the petitioners identified no mala fides in the sense the doctrine requires. On reorganisation, the Union pointed to Article 3(a) with Explanation I for Ladakh, and, for the conversion of the remainder, made the submission that ultimately removed the question from the case: the Union territory status of Jammu and Kashmir was temporary, elections would be held, and statehood would be restored.

Two features of the argument deserve record because they shaped the judgment’s silences. The petitioners’ challenge to the proclamations was, as the lead opinion observed, not the centre of their case: the petitions attacked the actions of August 2019, and the proclamations of 2018 and 2019 entered the argument as the setting for those actions rather than as independent targets pressed to a conclusion. That forensic choice is why the judgment contains a standard for judging actions under a proclamation but no decision on the proclamations themselves. And the Solicitor General’s assurance on statehood was given in response to the Court’s questions during the hearing on the conversion issue, which is why the undecided question is recorded in the judgment in the form of a reliance: the Court acted on the statement because the statement was made, and the direction for restoration at the earliest is the form in which the reliance was memorialised.

Sovereignty: the first holding

The petitioners’ case on sovereignty began from the terms of the accession itself. The Instrument of Accession, they argued, was a limited grant: it ceded authority over three subjects and expressly preserved the ruler’s sovereignty over everything else, and Article 370 was drafted to honour that reservation, making the State’s Constituent Assembly, rather than any Indian institution, the final arbiter of the relationship. On this reading, supported by the reasoning in Prem Nath Kaul, the State retained an element of internal sovereignty after 1947, and Article 370 was the constitutional recognition of a compact between two entities that had bargained.

The lead opinion rejected the premise. Reading the Instrument of Accession together with the proclamation of 25 November 1949, by which the State adopted the Constitution of India, and with the State Constitution’s own declaration that the State is an integral part of the Union, the Chief Justice held that the State of Jammu and Kashmir did not retain any element of sovereignty after the execution of the Instrument of Accession and the issuance of that proclamation. The State, the lead opinion held, does not possess internal sovereignty distinguishable from the powers and privileges enjoyed by other states of the Union. Article 370 was, on this reading, a feature of asymmetric federalism and not an acknowledgment of sovereignty: an arrangement under which one unit of the federation held terms different in degree from its neighbours, within a constitutional order that remained single and indivisible. The opinion pointed to the family of provisions around Article 371, which confer special arrangements on a series of states for reasons of history, geography and tribal governance, as the constitutional neighbourhood to which Article 370 belonged, and treated asymmetry of that kind as an ordinary, if unevenly distributed, technique of Indian federal design. The treatment of the distribution of legislative power in the Seventh Schedule lists is the background against which that claim of ordinariness is made in this series. The lead opinion also drew support from State Bank of India v. Santosh Gupta, where the Court had held that the residents of the State were citizens of India first and that the State’s Constitution was subordinate to the Constitution of India.

The lead opinion reached that conclusion by way of a general examination of what sovereignty can mean inside a federation, drawing on the classical accounts of Dicey and Austin and on the Constituent Assembly record, and its care on the point is explained by the stakes: if the State had retained sovereignty in any legally operative sense, the arrangements of 2019 would have required its consent in a form no proclamation could manufacture, and the case would have ended at the first question. The opinion instead treated sovereignty in the Indian Union as vested in the Constitution itself, with the Union and the states alike exercising distributed and limited powers under it, so that a state’s special arrangements, however generous, are grants within the constitutional order and not reservations against it. The illustrations from the Article 371 family carry weight in that argument because several of those provisions are more intrusive, in favour of the state concerned, than anything Article 370 conferred in its later decades: the Governor’s special responsibility for law and order in Nagaland under Article 371A, the protections for land and resources in Nagaland and Mizoram, and the regional development boards under Article 371 are all asymmetries, and none has ever been described as a reservation of sovereignty by the state concerned. Asymmetry, on the lead opinion’s account, is simply how the Indian federation accommodates unlike histories, and the special position of Jammu and Kashmir differed from its neighbours in the family in degree and in origin, not in juridical kind.

Justice Khanna’s short opinion aligned itself with the conclusion on this point in its own terms: the abrogation of Article 370 does not negate the federal structure, because the citizens living in Jammu and Kashmir do and will enjoy the same status and rights as citizens residing in other parts of the country. The federalism the judgment protects is, on that formulation, a federalism of equal citizenship rather than of uniform institutional design, and the distinction resurfaces in the criticism the judgment attracted, since opponents of the outcome regard the federal structure as residing precisely in the institutions a state’s people elect. The federal structure of the Constitution is part of its basic structure, and the series article on the basic structure doctrine carries the general doctrine of which this exchange is an instance.

Justice Kaul’s concurrence parted from the lead opinion on the historical question while arriving at the same legal destination, and the difference is the single most misreported feature of the separate opinions. Justice Kaul held, in light of the Court’s prior finding in Prem Nath Kaul, that the State retained an element of internal sovereignty despite the signing of the Instrument of Accession, and that Article 370 recognised this internal sovereignty by recognising the Constituent Assembly of the State. Where the lead opinion saw no residue of sovereignty at any point after accession, Justice Kaul saw a residue that the constitutional text itself acknowledged. The practical difference between the two routes then collapses, because Justice Kaul held that clause (3) of Article 370 contained the mechanism to bring the temporary arrangement to an end and, in turn, to de-recognise the internal sovereignty of the State and apply the Constitution of India in full. Sovereignty retained, on his reading, was sovereignty held on terms that included its own terminability by the President under clause (3). A report which says simply that the Court held the State retained no sovereignty is accurate as a statement of the lead opinion and of the result, but it erases a disagreement inside the unanimous Bench about the constitutional past, a disagreement the opinions preserve in terms.

Temporary or permanent: the second holding

The text of the provision carried the first mark of temporariness on its face: it sat in Part XXI, among temporary and transitional provisions, under a marginal heading calling it temporary. The lead opinion held that the historical context for the inclusion of Article 370 and its placement in Part XXI together established that it was a temporary provision. The history mattered in two registers. The circumstances of 1949, a war ended by ceasefire the year before, a question pending before the United Nations, and an expectation that the State’s own Constituent Assembly would complete the settlement, made the provision an interim bridge between accession and full constitutional integration. And the body to which the proviso to clause (3) assigned the recommending function was itself a temporary institution, convened for the single task of framing the State’s Constitution and dissolved when the task was done. A provision whose termination mechanism names a body designed to disappear is, the opinion reasoned, a provision designed to operate for a transitional period rather than for all time.

The petitioners’ counter-reading was textual and structural rather than historical. They argued that whatever the intention of 1949, the dissolution of the Constituent Assembly without a recommendation changed the legal position: the termination route closed, and a provision that can no longer be terminated by its own mechanism has, in effect, acquired permanence. The argument drew strength from the anomaly that a power exercisable only on a named body’s recommendation was being exercised in that body’s permanent absence. The Court did not accept that the mechanism’s dependence on the Assembly converted the provision’s character. Temporariness, on the Court’s approach, described the provision’s purpose and design, assessed from its text, its placement and its history, and a provision does not become permanent because the transitional period it governed turned out to be long. Justice Kaul concurred on this point by a combination of factors, the historical context, the text and the subsequent practice, and Justice Khanna recorded that Article 370 was enacted as a transitional provision and did not have permanent character.

The petitioners had an answer to the placement argument, and the judgment’s handling of it shows the care with which the holding was assembled. Part XXI, they pointed out, is a miscellany: it houses provisions that are temporary, provisions that are transitional, and the special provisions of the Article 371 family, several of which are permanent features of the federation and show no sign of expiring. Placement in that part therefore cannot by itself classify a provision, since the part’s own title lists three different characters disjunctively. The lead opinion did not rest on placement alone. It combined the placement with the article’s own marginal heading, which singles Article 370 out as temporary in a part where its neighbours are described as special, with the drafting history recounted earlier in this article, and with the structure of clause (3), which builds the provision’s ending into the provision itself. A permanent special provision, of the Article 371 kind, contains no clause contemplating its own cessation on a presidential declaration; Article 370 does, and the presence of a self-termination mechanism, the opinion reasoned, is the mark of an arrangement its authors expected to end. The petitioners’ permanence argument also had to confront an awkward feature of its own logic: if the dissolution of the Constituent Assembly made Article 370 permanent, the provision’s permanence would date not from any decision of that Assembly but from its failure to decide, and the Assembly had in fact used its sitting to frame a State Constitution within the article’s framework rather than to invoke the article’s end. Silence, the Court in substance held, is a frail foundation for permanence in a provision whose heading says the opposite.

The subsequent practice weighed in this holding deserves a sentence of its own, because it is where the third element of the temporariness reasoning operates. The President had exercised the power under clause (1) continuously from 1950 onwards, applying the Constitution to the State provision by provision across seven decades, and the Court treated that unbroken practice as evidence of what the provision was for: a device of staged integration, under which the distance between the State and the rest of the Union narrowed order by order, until the declaration under clause (3) completed a process the provision had been conducting throughout its life. The petitioners read the same practice as demonstrating the need for consent at each stage, since nearly every order of consequence had been issued with the concurrence of an elected State Government until the practice changed character in 2019. Both readings of the practice were before the Court. The judgment adopted the first, and recorded the second without accepting it.

The survival of the power under clause (3): the third holding

The question on which the whole case most nearly turned was narrower than the sovereignty and temporariness questions, because those questions frame the provision while this one operated it. If the President’s power under clause (3) had lapsed when the Constituent Assembly dissolved in January 1957, then Constitutional Order 273 was issued in exercise of a power that no longer existed, and the abrogation failed at its root. If the power survived, the remaining questions were about the route by which it was exercised in 2019.

The lead opinion held that the power under clause (3) did not cease to exist upon the dissolution of the Constituent Assembly of Jammu and Kashmir. Its reasoning proceeded from the distribution of roles inside the clause. The clause confers the power on the President. The proviso assigns to the Constituent Assembly a transitional function: to make a recommendation. When the Assembly was dissolved, what ceased to exist was the transitional power recognised in the proviso, the Assembly’s capacity to make its recommendation. The President’s power under the main clause was untouched by the disappearance of the body named in the proviso, because the proviso regulated one route to the power’s exercise and did not constitute the power itself. The opinion further held that the recommendation of the Constituent Assembly was not binding on the President: the text says recommendation, and a recommendation, in its ordinary constitutional usage, advises rather than commands. The President could therefore issue a notification under clause (3) unilaterally, and the absence of a recommendation from a body dissolved in 1957 was not a defect in Constitutional Order 273.

This reasoning engages, and recasts, the earlier authority of Sampat Prakash v. State of Jammu and Kashmir. In that decision the Court had upheld the continued exercise of the President’s powers under Article 370 long after the Constituent Assembly had dissolved, rejecting the argument that the provision had spent itself with the Assembly’s disappearance, and had treated the continuance of the power as flowing from the provision’s own scheme. The 2023 Bench built on that foundation and declined to disturb it: the line of authority running from Sampat Prakash through Mohd. Maqbool Damnoo had consistently treated Article 370 as operative and its mechanisms as available, and the dissolution of the Assembly had not, in any of those decisions, been held to disable the President. The petitioners’ invitation to treat the proviso as a condition precedent whose failure destroyed the power was, in the judgment’s assessment, an invitation to overrule the practical construction the provision had carried for over fifty years, and the Court declined it.

Justice Kaul concurred on the survival of the power in terms that follow from his own sovereignty reasoning: since Article 370 was a temporary arrangement, it could not be said that the mechanism under clause (3) came to an end after the State Constituent Assembly was dissolved, and the President could exercise the power at any time after the dissolution, in line with the aim of full integration of the State. Justice Khanna concurred with both opinions on the point. The consequence for the architecture of the judgment is that the fourth question, the want of the Constituent Assembly’s recommendation, was answered at the level of power rather than of procedure: there was nothing for the 2019 route to cure, because the recommendation the proviso contemplated was neither binding nor a surviving condition of the power once the recommending body had ceased to exist. That answer is also what gives the judgment’s fourth holding, on the amendment of Article 367, its peculiar character, since it made the irregularity the Court found in Constitutional Order 272 legally inconsequential, a feature taken up in the next section.

The integration process on which the fifth holding rests

The lead opinion’s description of Constitutional Order 273 as the culmination of a gradual process was not a metaphor; it pointed to a documented sequence of exercises of the clause (1) power across seven decades, and the sequence is the factual spine of the holding. The Order of 1950 applied a comparatively short list of constitutional provisions to the State, essentially those needed to give effect to the accession subjects and to the Union’s institutional minimum. The Order of 1954, negotiated against the Delhi Agreement of 1952 and issued with the concurrence of the State Government, changed the scale entirely: it applied the great body of the Constitution to the State with exceptions and modifications, among them the insertion of Article 35A protecting the permanent residents regime, a modified application of the fundamental rights chapter, and the extension of the jurisdictions of the Supreme Court, the Election Commission and the Comptroller and Auditor General in forms that later orders completed. The State’s own constitutional practice moved in parallel: in 1965 the State amended its Constitution to replace the elected Sadr-i-Riyasat with a Governor appointed by the President and to retitle its Prime Minister as Chief Minister, aligning its offices with the all-India pattern. Direct elections to the Lok Sabha from the State followed in 1967, replacing the earlier arrangement under which the State’s representatives were chosen on the State Legislature’s recommendation.

Across the decades that followed, successive orders under clause (1) applied further provisions and amended earlier applications, and the Union List subjects on which Parliament could legislate for the State widened from the accession subjects to nearly the whole field, each extension formally resting on the concurrence of the government of the day in the State. The political character of that concurrence changed over time, and the petitioners made the change a theme: concurrence given by elected governments bargaining with the Union in the 1950s and 1960s was one thing, they argued, and concurrence given during periods of central rule, or extracted in the conditions of the 1975 accord between Indira Gandhi and Sheikh Abdullah, under which the Sheikh returned to office on the express footing that the State’s position under Article 370 continued, was another. The judgment did not audit the quality of the concurrences across the decades. Its use of the sequence was doctrinal rather than historical: the continuous exercise of the power under clause (1), by Presidents of every political complexion and with governments of every complexion in Srinagar, demonstrated that the provision was functioning as a mechanism of progressive application, and a power exercised in that direction for seventy years supported reading the final declaration under clause (3) as the process completing itself rather than as a rupture. The counter-reading, that a process of integration by consent cannot be completed by an act without consent, was the petitioners’ deepest objection to the fifth holding, and it is recorded here because the judgment answers it only by the equivalence reasoning set out in the section on Constitutional Order 272: the final step had the same effect as a clause (3) declaration, for which consent was never required at all.

The earlier decisions the Bench had to work with

Four earlier decisions framed the field, and the judgment’s treatment of each is part of its content, because the 2023 Bench was not writing on a blank page.

Prem Nath Kaul v. State of Jammu and Kashmir, decided in 1959, arose from a detention under the State’s preventive detention law and required the Court to describe the authority of the State’s Constituent Assembly. The Court there held that the Constituent Assembly of the State was the body competent to determine the State’s constitutional relationship, and passages in the judgment spoke of the State’s position in terms the petitioners in 2023 pressed as a recognition of retained internal sovereignty. The lead opinion read the decision more narrowly, as a case about the Assembly’s authority within the constitutional scheme during the period of transition, and declined to extract from it a doctrine of surviving sovereignty. Justice Kaul read it as the petitioners did on the historical point, and the resulting divergence inside the Bench, described in the sovereignty section of this article, is a disagreement about the correct reading of a 1959 judgment rather than about any fact.

Sampat Prakash v. State of Jammu and Kashmir, decided a decade later, also arose from preventive detention, and the challenge there was that the President’s power under Article 370 had lapsed with the Constituent Assembly, so that orders extending central legislation to the State were void. The Court rejected the argument and upheld the continuance of the power, treating Article 370 as a living provision whose mechanisms remained available to the President notwithstanding the Assembly’s dissolution. The 2023 judgment’s third holding is, in substance, Sampat Prakash restated at the level of clause (3), and the Bench’s refusal to reopen it carried the weight the Court usually gives to a construction under which governments and courts have acted for half a century.

Mohd. Maqbool Damnoo v. State of Jammu and Kashmir, decided in 1972, upheld the adaptation by which references to the Sadr-i-Riyasat in the constitutional scheme came to be read as references to the Governor of the State, an adaptation effected through the machinery of clause (1) and Article 367. The case mattered in 2023 because it marked the boundary the Court was later to draw with care: adaptations of nomenclature and reference through the interpretation clause had been accepted, because they made the text workable without altering the distribution of power inside Article 370. The 2019 adaptation was different in kind, since it substituted one constitutional actor for another inside the very procedure for ending the article, and the Damnoo line of acceptance therefore did not save it.

State Bank of India v. Santosh Gupta, decided in 2017, concerned the applicability of a central debt recovery statute to the State and required the Court to describe the relationship between the two Constitutions. The Court held that the Constitution of Jammu and Kashmir was subordinate to the Constitution of India and that the residents of the State were citizens of India, rejecting the submission that the State’s people held a separate sovereign allegiance. The lead opinion in 2023 drew on that reasoning in its sovereignty analysis. Taken together, the four decisions show the pattern the Bench confronted: the mechanisms of Article 370 had been construed generously and its continuance upheld whenever challenged, while no decision before 2023 had been required to rule on the termination of the article itself, which is why the clause (3) question, though governed by the spirit of Sampat Prakash, still had to be reasoned out in full.

The one invalidity: Paragraph 2 of Constitutional Order 272

Constitutional Order 272 did two things of different legal characters, and the judgment’s treatment of it cannot be understood unless they are kept apart. Its principal operation was to supersede the 1954 Order and apply all the provisions of the Constitution of India to the State of Jammu and Kashmir. Its Paragraph 2 additionally modified Article 367, the interpretation clause, as it applied to the State, by inserting a clause (4) into that article. Among the constructional directions the new clause contained, one changed the proviso to Article 370(3): the expression “Constituent Assembly of the State” in that proviso was to be read as “Legislative Assembly of the State”. The purpose of the substitution was plain. If the body whose recommendation clause (3) required was to be the Legislative Assembly, then Parliament, exercising the Assembly’s powers during President’s Rule, could supply the recommendation, and the declaration under clause (3) could proceed with the procedural box ticked. The petitioners attacked Paragraph 2 as a colourable device: an amendment of Article 370 dressed as an interpretation of it.

The lead opinion accepted that attack in terms that leave no room for softening. Article 370, the Court held, cannot be amended by exercise of power under clause (1)(d). If Article 370 is to cease to operate, or to be amended or modified in its application to the State, recourse must be taken to the procedure contemplated by clause (3). Paragraph 2 of Constitutional Order 272, by which Article 370 was amended through Article 367, is ultra vires clause (1)(d), because it modifies Article 370 in effect without following the procedure prescribed for modifying Article 370. An interpretation clause cannot be used to bypass the procedure laid down for amendment. The opinion reasoned from the function of an interpretation clause: such a clause directs how words are to be read; it does not reallocate the powers the substantive provision distributes. To read “Constituent Assembly” as “Legislative Assembly” was not to construe Article 370 but to change one of its institutional actors, and thereby to change the provision’s own termination procedure. Permitting that method, the Court observed, would expose constitutional provisions generally to amendment by the circuitous route of redefinition, without the procedure their texts prescribe. The general discipline that a constitutional provision with its own amendment procedure must be amended by that procedure, and not by a device operating elsewhere in the text, is the subject of the series article on the constitution amendment process in India, where this holding takes its place in the doctrine.

The earlier authority pointed the same way, and the Court said so. In Puranlal Lakhanpal v. President of India the Court had held, in the early 1960s, that the power under clause (1)(d) to apply provisions of the Constitution with modifications was a power of application, not a power to change the basic structure of the provision being applied, and the long-accepted position was that Article 370 could not be reached by the general amending power in Article 368 but only through its own clauses. The 2023 holding is the completion of that line: the provision’s own clause (3) was the only door, and the 367 route was an attempt to walk through a wall.

Both concurring opinions agreed on this holding, which means the invalidity commanded all five judges, and each stated it in a distinctive idiom. Justice Kaul held that the power under clause (1)(d) read with Article 367 cannot be used to do indirectly what cannot be done directly, that the power to make modifications under clause (1)(d) cannot be used to amend Article 370, and that Article 367, being an interpretation clause, cannot be used to alter the character of a provision; Paragraph 2 of Constitutional Order 272, insofar as it amended Article 367(4), was accordingly ultra vires Article 370. In argument and in his opinion Justice Kaul put the point in the language of prescribed procedure: where a procedure is prescribed, it has to be followed, and amendment through the back door is not permissible. Justice Khanna, in the short opinion, recorded the same conclusion in his own summary: Paragraph 2 of Constitutional Order 272, by which Article 370 was amended by taking recourse to Article 367, is ultra vires and bad in law. Any account of the judgment which reports that the Constitutional Orders were upheld in their entirety is therefore inaccurate on a point all five judges decided the same way.

Why the invalid paragraph changed nothing

The petitioners resisted severance directly, and their argument should be recorded because the Court’s answer to it is the hinge of this part of the case. Constitutional Order 272, they submitted, was a single integrated exercise: the amendment of Article 367 was not an appendage to the Order but its operative heart, the step that unlocked everything else, and an order whose central device is void cannot survive in fragments. The submission drew on the ordinary doctrine that an instrument falls as a whole where its valid and invalid parts are inextricably mixed, so that what remains after excision is something the authority would not have enacted on its own. The Court’s answer proceeded by separating functions rather than sentences. The Order performed two operations that rested on two distinct grants of power: the application of the Constitution, which clause (1)(d) authorised in terms, and the redefinition of the recommending body, which no clause authorised. The first operation did not depend for its validity on the second, because the power to apply the whole Constitution existed whether or not the proviso’s vocabulary was tampered with, and the second operation, being ineffective, simply failed to change the law. What remained after the excision was therefore not a mutilated instrument but the Order’s principal and independently authorised act. Severance of that kind, where the invalid part is an addition that never took legal effect and the valid part is complete without it, is the easiest case for the doctrine, and the Court treated it accordingly. The critics’ response, set out in the criticism section of this article, is that easiness in doctrine is not the point: the device was adopted because the government of the day believed it necessary, and a holding that it was both void and unnecessary converts a contested political judgment into a drafting error after the fact. The judgment does not answer that response in terms, because the response is about the politics of the sequence rather than about the law of severance, and the distance between the two is characteristic of the case as a whole.

The natural sequel to a finding of ultra vires is the fall of the instrument that contains it, and the most distinctive feature of this judgment is that the sequel did not follow. The Court severed the invalid operation of Constitutional Order 272 from its valid operation and upheld the Order’s substantive achievement, for reasons that have to be followed step by step, because together they constitute the judgment’s answer to the charge that it found an impropriety and attached no consequence to it.

The first step is the breadth of clause (1)(d) itself. The President’s power to apply other provisions of the Constitution to the State with exceptions and modifications extends to applying all of them. Paragraph 2 of Constitutional Order 272, insofar as it applied all the provisions of the Constitution of India to the State, was a valid exercise of that power, and the Court held it so. The second step answers the objection that the concurrence for that application was defective, because in August 2019 the concurrence of the State Government was in substance the concurrence of the Governor acting under a Proclamation, which the petitioners equated with the Union concurring with itself. The Court held that the President did not have to secure the concurrence of the Government of the State, or of the Union Government acting on behalf of the State Government under the second proviso to clause (1)(d), while applying all the provisions of the Constitution to the State, because such an exercise of power has the same effect as an exercise of power under clause (3), for which the concurrence or collaboration of the State Government was never required. The reasoning is an equivalence argument: a total application of the Constitution exhausts the subject matter on which concurrence could operate, and therefore cannot be made to depend on a consent designed for piecemeal application. The third step concerns good faith in the manner of application. The petitioners had argued that applying the entire Constitution at one stroke, rather than provision by provision across time, was a mala fide use of a power whose design was gradual. The Court held that the exercise of power by the President under clause (1)(d) to issue Constitutional Order 272 was not mala fide, and that such an exercise is not mala fide merely because all the provisions were applied together without following a piecemeal approach.

Justice Khanna’s formulation of the same conclusion adds a qualification worth preserving exactly: the paragraph was ultra vires and bad in law, albeit sustainable in view of the corresponding power under clause (1)(d). Justice Kaul’s route was equivalent in result: the President had the power to apply all provisions of the Constitution to the State under clause (1)(d), a power similar in effect to the power under clause (3), and therefore the remainder of Paragraph 2 of Constitutional Order 272 was valid. The combined effect of the four holdings to this point is the judgment’s signature structure. The substitution of the Legislative Assembly for the Constituent Assembly was beyond power. The substitution was also unnecessary, because the President needed no recommendation from any assembly to act under clause (3) after 1957. And the application of the whole Constitution, which the substitution was designed to facilitate, was independently valid under the clause that had been used to apply it. The impropriety was real, identified, and held ultra vires by five judges, and it carried no remedy because every act it was meant to enable stood on another, valid, footing.

Constitutional Order 273 and the culmination of integration: the fifth holding

Constitutional Order 273, issued on 6 August 2019, was the declaration under clause (3) that all the clauses of Article 370 would cease to be operative except clause (1), which would operate in a restated form applying all provisions of the Constitution, as amended from time to time, to the State. Three challenges were pressed against it. The first was the want of the Constituent Assembly’s recommendation, answered by the third holding: the power survived the Assembly and the recommendation was not binding. The second was that the recommendation actually obtained, from Parliament sitting in the shoes of the dissolved Legislative Assembly through the substituted reading of the proviso, was worthless; on the Court’s construction of clause (3), that recommendation was surplusage, since none was required, and the resolutions of 5 and 6 August 2019 neither validated nor invalidated anything. The third was an attack on the exercise itself as a mala fide and unreviewable-in-name-only act, and it is on this challenge that the Court articulated the standard governing clause (3).

The lead opinion held that the President had the power to issue a notification declaring that Article 370 ceases to operate without the recommendation of the Constituent Assembly. It held further that the continuous exercise of power under clause (1) by the President, across the decades, indicated that a gradual process of constitutional integration was ongoing, and that the declaration issued under clause (3) was the culmination of that process and, as such, a valid exercise of power. Constitutional Order 273 was accordingly valid. The standard of review embedded in this holding is narrow by design: the exercise of the clause (3) power is a political and constitutional act of the President on the aid and advice of the Union Council of Ministers, and the judgment subjects it to challenge on the ground of mala fides, a ground the petitioners had to make out and did not. Absent mala fides, the Court would not sit in appeal over the judgment that the process of integration was complete. Justice Kaul concurred that Constitutional Order 273, which declares that Article 370 shall cease to operate except as provided, was issued under clause (3) and is valid, and Justice Khanna recorded the point in a sentence: Article 370 has been made inoperative in terms of clause (3) of Article 370, and Constitutional Order 273 is valid.

The resolutions Parliament adopted on 5 and 6 August 2019, recommending the declaration in the name of the State Legislature whose powers it was exercising, occupy a curious position after this holding. Politically, they were the centrepiece of the government’s presentation in 2019: the recommendation the proviso appeared to require had been obtained, from the only body then capable of giving it, through the substituted reading of the proviso that Constitutional Order 272 supplied. Legally, after the third and fourth holdings, the resolutions do no work at all. The recommendation was not binding and was not required, and the substitution that made Parliament’s recommendation the Assembly’s recommendation was ultra vires. The declaration rests instead on the President’s unassisted power under clause (3), a power the Court found to have been intact since 1950. A government that had simply issued Constitutional Order 273 without the 367 device and without the resolutions would, on the judgment’s reasoning, have achieved the same legal result by a shorter and regular route. That counterfactual is the sharpest measure of the judgment’s structure: the steps the government took most trouble over were the steps the Court held either irregular or unnecessary, and the step that mattered was the one the text had always allowed.

One consequence of the declaration received its own sentence in the lead opinion and belongs in any accurate account. Following the application of the Constitution of India in its entirety to the State, the Constitution of Jammu and Kashmir was rendered inoperative, and the Court declared it to have become redundant. The State’s separate constitutional text, framed in 1956 and the subject of its own body of doctrine, ceased to operate as a constitution because the field it occupied had been occupied in full by the Constitution of India, which the Court described as a complete code for constitutional governance. No separate holding struck down the State Constitution; it fell as a consequence, and FAQ accounts of the judgment that list it among the instruments invalidated misstate the mechanism.

What became of the Constitution of Jammu and Kashmir

The sentence declaring the State Constitution redundant is easy to pass over, because the Court reached it as a corollary, but it disposes of an entire constitutional text and its consequences deserve a section. The Constitution of Jammu and Kashmir had been framed by the State’s Constituent Assembly under the authority the settlement of 1949 contemplated, and for seven decades it operated alongside the Constitution of India in a divided field: the State had its own constitutional provisions for its legislature, executive and judiciary, its own definition of the permanent residents who held special rights under its Section 6 and the laws made under it, and its own amending procedure, which required a two-thirds majority of its legislature for most changes and barred any amendment affecting the State’s relationship with the Union. Where the two Constitutions overlapped, the constitutional orders under Article 370 specified which provisions of the Indian text applied and with what modifications, and the courts, including the Supreme Court in Santosh Gupta, treated the Indian text as paramount in case of conflict.

The declaration under clause (3) removed the instrument that had performed that allocation. Once all provisions of the Constitution of India applied to the State without exception or modification, there was no field left in which a separate state constitution could operate: the legislature, executive and courts of the successor units derive their authority from the Indian Constitution and from the Reorganisation Act made under it, and the permanent residents regime lost the constitutional shelter of Article 35A and of the State Constitution together. The lead opinion expressed the result by saying that the Constitution of India is a complete code for constitutional governance and that the State Constitution had become inoperative and redundant. Two precisions matter for anyone using the holding. The Court did not strike the State Constitution down for inconsistency, as it might have done provision by provision; it treated the text as displaced wholesale by the complete application of the superior code, which is why there is no schedule of invalidated sections and why attempts to litigate surviving fragments of the State Constitution face the threshold answer that the instrument has ceased to operate. And the consequence was not produced by the Reorganisation Act, which reorganised territory and institutions, but by the Constitutional Orders, which is the doctrinal reason the point appears in this article rather than in the article on the statute.

President’s Rule: the framework, and the questions not adjudicated

The fifth and sixth questions the Court framed concerned the Governor’s proclamation under Section 92 of the State Constitution in June 2018, the dissolution of the Legislative Assembly in November 2018, and the Proclamation under Article 356 of December 2018 with its extension. The Court’s disposition of these questions is the second silence of the judgment, and it is a qualified one. The lead opinion recorded that the petitioners did not challenge the issuance of the proclamations under Section 92 and Article 356 until the special status of the State was abrogated, and held that the challenge to the proclamations did not merit adjudication, because the principal challenge was to the actions taken after the proclamations were issued. The validity of the proclamations themselves was therefore not decided. What the Court decided instead was the standard by which actions taken under a proclamation are to be judged, and it decided that in full.

Did the petitioners challenge the proclamations of President’s Rule in time?

The lead opinion recorded that the proclamations under Section 92 of the State Constitution and under Article 356 were not challenged until after the special status had been abrogated in August 2019, and held for that reason that the challenge to the proclamations did not merit adjudication in these proceedings.

The doctrine into which this framework fits had been built in two great cases. In S. R. Bommai v. Union of India, a bench of nine judges held in 1994 that a proclamation under Article 356 is subject to judicial review, that the President’s satisfaction must rest on material relevant to the breakdown of constitutional machinery, and that federalism and secularism form part of the Constitution’s basic structure; the opinions of Justices Sawant and Jeevan Reddy, on which the 2023 lead opinion drew, supplied the analytical account of the power’s limits. In Rameshwar Prasad v. Union of India, the Court held the dissolution of the Bihar Legislative Assembly unconstitutional while declining, on the facts, to restore it, establishing that action under a proclamation can be condemned in principle yet stand in effect. The 2023 judgment works inside that tradition but refines the point of entry for a challenger: earlier doctrine had concentrated on the proclamation itself, whereas the 2023 standard, framed for a case in which the proclamations were not in issue, concentrates on the individual exercises of power that follow a proclamation, and allocates the burden of showing mala fides or extraneous purpose to the challenger in the first instance. That allocation is the operative rule the case contributes to Article 356 law, and it will govern challenges to action taken during President’s Rule in any state, on facts far removed from Kashmir.

The standard the Court laid down has four elements. The exercise of power by the President after a proclamation under Article 356 is issued is subject to judicial review. The exercise must have a reasonable nexus with the object of the proclamation. The person challenging the exercise must prima facie establish that it is a mala fide or extraneous exercise of power; once a prima facie case is made out, the onus shifts to the Union to justify the exercise. And the power of Parliament under Article 356(1)(b) to exercise the powers of the legislature of the State cannot be restricted to law-making power, excluding the non-legislative powers of the State Legislature, because such an interpretation would read a limitation into the provision contrary to its text. This framework draws on the Court’s earlier doctrine in S. R. Bommai, where Justices Sawant and Jeevan Reddy had analysed the limits of Article 356, and on Rameshwar Prasad, and Justice Khanna recorded his agreement with the lead opinion’s examination of Article 356 and with its reading of those authorities. Its significance for the 2019 sequence is direct. The petitioners’ argument that Parliament, acting as the State Legislature, could not perform acts of a non-legislative character, such as recommending the abrogation of the State’s constitutional position, failed on the fourth element: the power Parliament assumes under a proclamation is the whole power of the State Legislature, legislative and non-legislative alike. Their argument that the actions were mala fide failed on the third: no prima facie case of mala fide or extraneous purpose was established to the Court’s satisfaction, so the onus never shifted.

Justice Kaul’s concurrence on this part of the case reasoned from the sequence of events: President’s Rule could be imposed after the dissolution of the Assembly because the failure of constitutional machinery had preceded both Governor’s rule and the dissolution; once the proclamation had been approved by both Houses of Parliament, reflecting the will of the people, the President had the power under Article 356 to make irreversible changes, and Articles 356 and 357 did not bar the exercise of the non-legislative powers of the State Legislature. He joined the lead opinion in insisting, at the same time, that a proclamation of emergency is bound by judicial and constitutional scrutiny, so that the exercise of emergency powers is not unfettered and without limit. The petitioners’ strongest structural objection, that the 2019 route permitted changes of a kind no temporary administration should be able to make, because they could not be undone by any restored state government, was squarely presented to the Court and squarely answered: irreversibility of consequence is not, without more, a ground of invalidity, and the discipline the Constitution imposes on action during President’s Rule is the discipline of nexus and good faith, applied through the standard the judgment records.

Six holdings, one invalidity, two silences: the artifact

The table below states the judgment’s disposition of the questions before it in the form the article has argued for throughout: each holding on its own line, the reasoning compressed to its operative step, and the effect stated without enlargement. It is the reference against which summaries of the case can be checked.

Question before the Court	Holding	Reasoning in one line	Effect
Did the State retain sovereignty after accession?	No element of sovereignty was retained (lead opinion; Justice Kaul dissenting from the reasoning, not the result)	The Instrument of Accession and the proclamation of 25 November 1949 completed the State’s integration as a unit of the Union	Article 370 is read as asymmetric federalism, not as a compact between sovereigns
Was Article 370 temporary or permanent?	Temporary and transitional (all three opinions)	Its placement in Part XXI, its marginal heading and the circumstances of 1949 show a provision designed for a transitional period	The termination mechanism in clause (3) is read as a working part of the design
Did the clause (3) power survive the Constituent Assembly?	The power survived; the recommendation was not binding (all three opinions)	Only the proviso’s transitional function lapsed with the Assembly; the President’s power under the main clause was unaffected	The absence of a recommendation after 1957 is no defect in a clause (3) declaration
Could Article 367 be amended to rewrite the clause (3) proviso?	No; Paragraph 2 of Constitutional Order 272, insofar as it amended Article 370 through Article 367, is ultra vires (all five judges)	An interpretation clause cannot be used to amend a substantive provision or to bypass the procedure the provision prescribes for its own amendment	The substitution of the Legislative Assembly for the Constituent Assembly is bad in law
Was Constitutional Order 272 otherwise valid?	Yes; the application of all provisions of the Constitution was a valid exercise of clause (1)(d) and was not mala fide	Total application has the same effect as a clause (3) declaration, for which concurrence was never required	The Order stands, and the Constitution of Jammu and Kashmir became inoperative and redundant
Was Constitutional Order 273 valid?	Yes (all three opinions)	The declaration was the culmination of a continuous, gradual process of integration under clause (1), and no mala fides was shown	Article 370 operates only in the restated clause (1) form; the special status stands ended
Could Ladakh be carved out as a Union territory?	Yes (lead opinion, with the concurrences)	Article 3(a) read with Explanation I permits forming a Union territory by separation of territory from a state	The Union territory of Ladakh stands

The reorganisation holding: Ladakh upheld, conversion undecided

The seventh and eighth questions concerned the Reorganisation Act, and the Court’s answers divide along the seam between the two Union territories the Act created. On procedure, the Court held that the views of the legislature of a state under the first proviso to Article 3 are recommendatory in character: a Bill affecting the area, boundaries or name of a state must be referred to that legislature for its views, but the views expressed do not bind Parliament. Applying that settled position, which runs back to Babulal Parate v. State of Bombay, the Court held that Parliament’s exercise of power under the first proviso during the Proclamation, when Parliament itself discharged the functions of the State Legislature, was valid and not mala fide. Justice Kaul added that it was in any event not possible to ascertain the views of a State Legislature that stood dissolved, and that the functions of that legislature were performed by Parliament, from which it followed that the corresponding provisions of the Reorganisation Act were valid. The challenge founded on the second proviso to Article 3, which in its application to Jammu and Kashmir had required the consent of the State Legislature to changes in the State’s area, boundaries or name, fell with the application of the Constitution in full and with the recommendatory character of the reference procedure under the Proclamation.

The second proviso point deserves its own statement, because it was the strongest procedural card the petitioners held on reorganisation. As Article 3 applied to Jammu and Kashmir under the constitutional orders, a further proviso required that no Bill increasing or diminishing the area of the State, or altering its name or boundary, be introduced in Parliament except with the consent of the Legislature of the State. That consent requirement was a creature of the Article 370 scheme: it did not exist for any other state, and it embodied, at the level of ordinary territorial legislation, the same principle of state concurrence that structured clause (1) of Article 370. In August 2019 the State Legislature stood dissolved and its powers were exercised by Parliament, and the consent the proviso contemplated was therefore supplied, if at all, by Parliament itself. The petitioners submitted that a consent requirement designed as a protection for the State could not be satisfied by the very authority against which it protected. The Court’s disposition of the reorganisation challenge did not sustain that submission: with the Constitution applied in full by Constitutional Orders 272 and 273, the special application of Article 3 that carried the second proviso fell away with the rest of the Article 370 scheme, and the reference procedure that remained was the first proviso’s, whose views are recommendatory and whose performance by Parliament during the Proclamation the Court upheld. The episode closes the circle of the judgment’s federal reasoning: every consent the special scheme required of the State was held, for a distinct textual reason in each case, either unnecessary, recommendatory in character, or competently supplied under the Proclamation.

On the substantive question, the Court upheld one part of the reorganisation and declined to decide the other. The validity of the decision to carve out the Union territory of Ladakh was upheld, in view of Article 3(a) read with Explanation I, which permits the formation of a Union territory by separation of a territory from any state. The separation of territory from a state and its constitution as a Union territory is, on the text, an enumerated use of the Article 3 power, and the Court treated it as such. The conversion of the remaining territory of the State, the successor unit also named Jammu and Kashmir, from a state into a Union territory presented the harder and unprecedented question, and the Court did not answer it. The Solicitor General had stated, in the course of the hearing, that the statehood of Jammu and Kashmir would be restored, the carving out of Ladakh apart, and that the Union territory status of the remainder was a temporary arrangement pending that restoration. The lead opinion recorded the statement and held that in view of it, it was not necessary to determine whether the reorganisation of the State into two Union territories is permissible under Article 3. The question whether Parliament may convert a state into Union territories therefore stands expressly undecided, and the series guide to state reorganisation in India records the same reservation in its treatment of the reorganisation power, because any future exercise of that power will meet this silence first.

The two directions of the judgment attach to this part of the case. The Court directed the Election Commission of India to take steps to conduct elections to the Legislative Assembly of Jammu and Kashmir, constituted under Section 14 of the Reorganisation Act, by 30 September 2024. And it directed that restoration of statehood take place at the earliest and as soon as possible. The first direction carried a date and was complied with on schedule: elections were held in September and October 2024 and an elected government took office in the Union territory in October 2024. The second direction carried no date, and its content is the assurance the Union gave, converted by the judgment into an obligation of restoration without a timetable fixed by the Court. The contrast between the two directions, one dated and one not, is a deliberate feature of the disposition and is regularly flattened in summaries that report both as orders of the same kind.

The differing texture of the two directions reflects a deliberate judicial choice about the limits of the Court’s own role, and it connects the reorganisation disposition to the undecided conversion question. A direction to hold elections by a date certain is administrable: it is addressed to a constitutional authority with a statutory duty, the Election Commission, and compliance can be verified on a calendar. A direction to restore statehood operates on a decision that the Constitution commits to Parliament under Article 3, and a court that has declined to define the limits of that article’s power is poorly placed to fix the date by which the power must be exercised in a particular direction. The formula the Court chose, at the earliest and as soon as possible, converts the Solicitor General’s assurance into a recorded obligation while leaving its timing to the authority that gave it. Critics of the judgment treat that deference as the point at which the assurance did its real work: the conversion question was not merely postponed but disarmed, since a future challenge to a prolonged continuance of Union territory status would have to overcome a direction the Court framed without a deadline and a record that the arrangement was accepted as temporary. Defenders answer that the alternative, a judicially imposed date for a parliamentary act of reorganisation, would have been the genuinely lawless course, and that the assurance, recorded in a judgment and repeated as a direction, binds the Union politically and constitutionally in ways a dismissed petition would not. The passage of the election deadline without controversy, and the holding of the elections it required, is the factual epilogue the judgment has so far received; the statehood direction has received none, and the asymmetry between a direction performed and a direction pending is the live residue of the case.

The concurring opinions in full

Justice Kaul’s opinion is a concurrence in result built on a different account of the constitutional past, and it adds the judgment’s only recommendation. On sovereignty, as set out earlier in this article, he read Prem Nath Kaul as having found that the State retained an element of internal sovereignty after accession, recognised by Article 370 through its recognition of the State’s Constituent Assembly. On the mechanism of ending, his reasoning turned that recognition to the opposite conclusion from the petitioners’: clause (3) contained the mechanism to bring the temporary arrangement to an end and, in turn, to de-recognise the internal sovereignty of the State and apply the Constitution of India in full, and since the arrangement was temporary, the mechanism could not have lapsed with the Assembly’s dissolution. On the amendment of Article 367, he held with the lead opinion that the paragraph was ultra vires, in the idiom of prescribed procedure. On the proclamations and the powers exercised under them, he held that irreversible changes could be made under a duly approved proclamation, while insisting that emergency power remains subject to judicial and constitutional scrutiny. On the reorganisation, he held the challenge did not require debate in view of the Union’s assurance on statehood.

The recommendation appears in an epilogue to Justice Kaul’s opinion, written from his own connection with the region, in which he recorded that the social fabric of the State had been damaged across decades and that a generation had grown up amid distrust. He recommended the setting up of an impartial Truth and Reconciliation Commission to investigate and report on the violations of human rights both by state and by non-state actors, at least since the 1980s, and to recommend measures for reconciliation. He specified that the exercise should be carried out in a time-bound manner and before memory escapes, and that, considering the sensitivities involved, it was for the government to decide the manner in which such a commission should be constituted. The status of this passage is essential to state exactly: it is a recommendation in a concurring opinion, not a direction of the Court. It binds no authority, it fixes no obligation, and it forms no part of the operative order. Reports which describe the Court as having ordered an inquiry into human rights violations in the State misstate the judgment; the Court ordered nothing of the kind, and one judge recommended it.

Justice Khanna’s opinion is the briefest of the three, and its method is agreement with particularisation: he recorded that he found it difficult to agree with the reasoning of one of his colleagues and not the other, and that he therefore concurred with both, adding particulars. His particulars carry two points already noticed in this article, the formulation that Paragraph 2 of Constitutional Order 272 was ultra vires and bad in law, albeit sustainable in view of the corresponding power under clause (1)(d), and the statement that abrogation does not negate the federal structure because the State’s citizens enjoy the same status and rights as citizens elsewhere in the country. His third particular is the one with the longest forward reach. Union territories, he observed, are normally geographically small territories, or are created for aberrant reasons or causes. The conversion of a state into a Union territory has grave consequences: among others, it denies the citizens of the state an elected state government of their own and it impinges on federalism. Conversion or creation of a Union territory out of a state, he held, has to be justified by giving very strong and cogent grounds, and must be in strict compliance with Article 3. Justice Khanna concurred, even so, in not deciding the conversion question, given the Union’s assurance. His observations therefore operate as a recorded caution rather than as law laid down: they are the terms on which a future bench, facing a conversion without an assurance of restoration, has been told by one member of this Bench the question ought to be approached.

Mala fides: the thread that runs through the whole judgment

A reader who follows the judgment’s use of one idea across its parts will understand its method better than a reader who follows any single holding. That idea is mala fides, bad faith, and it appears at four distinct points in the disposition, each time performing the same structural function: it is the sole ground on which the Court was willing to supervise an act the Constitution commits to a political authority, and at each point the petitioners failed to satisfy it.

It appears first in relation to Constitutional Order 272’s application of the whole Constitution, where the Court held the exercise not mala fide and refused to treat total application as bad faith merely because a piecemeal method had been the practice. It appears second in relation to Constitutional Order 273, where the declaration under clause (3) was treated as reviewable for mala fides and for no broader error of judgment about whether integration was complete. It appears third in the Article 356 framework, where the prima facie showing of a mala fide or extraneous exercise is the trigger that shifts the burden to the Union. And it appears fourth in the Article 3 reference, where Parliament’s exercise of the power under the first proviso during the Proclamation was held valid and not mala fide. The pattern is consistent and deliberate. On each question where the Constitution’s text gave the Court a rule to apply, the identity of the recommending body, the reach of an interpretation clause, the survival of a power, the Court applied the rule and decided. On each question where the attack was on the wisdom, timing or motive of a political act, the Court confined itself to the mala fides inquiry and, finding the threshold unmet, stopped. The judgment’s admirers describe this as the correct division of labour between court and polity. Its critics describe the same pattern as the case’s evasion, because a mala fides standard that requires proof of corrupt purpose will almost never be met by challengers to a determined government, however structural their objection. Both descriptions are of the same architecture, and the architecture, rather than any one of the six holdings, is what a future court will inherit from the case when it is asked to supervise emergency power or a contested reorganisation again.

The criticisms, stated at full strength, and the answers

Three criticisms of the judgment have been made persistently by named critics, and each deserves to be stated in its strongest form before the answers are set against it.

The first is that the temporariness holding reads the safeguard out of the provision. The proviso to clause (3) made the recommendation of the State’s Constituent Assembly necessary before the President could end the article, and that requirement was, on the critics’ reading, the price of the accession settlement: the State’s own constituent body, and no one else, would say when the special arrangement could close. To hold that the recommendation was never binding, and that the power survived the Assembly’s dissolution intact, is to hold that a safeguard drafted in mandatory terms, “shall be necessary”, imposed no constraint at any point in the provision’s life. Fali S. Nariman, the senior advocate and constitutional jurist, described the Court’s approach in published comment as the wrong one, characterising the reasoning on the central questions as an interpretive muddle. Gautam Bhatia, the constitutional scholar, argued in his published analysis of the judgment that the Court reached temporariness by assembling text, history and practice into a conclusion none of them compelled, and that the reasoning on the proviso dissolved a textual condition into an intention attributed to the framers. On this criticism, the judgment does not interpret the proviso; it retires it.

The second criticism is that the judgment permitted the Union to act as the State for the purpose of consenting to a change in the State’s own status. Every federal safeguard in the sequence depended on a state organ that did not exist in August 2019: the concurrence recorded for Constitutional Order 272 was the Governor’s, the recommendation for Constitutional Order 273 came from Parliament exercising the dissolved Assembly’s powers, and the views under Article 3 were Parliament’s own. Critics, among them the Supreme Court Observer’s journal analysis of the verdict, put the point as one of constitutional morality: a federation in which the Union may suspend a state’s elected organs and then supply, from its own side of the table, every consent the Constitution requires of the state, has converted federalism’s procedures into formalities that operate only when they cannot obstruct. The judgment’s answer, that the petitioners had not made out a prima facie case of mala fides and that irreversibility is not itself a ground of challenge, is described by the critics as an answer to a different question, since the objection was not that the Union acted from corrupt motives but that the structure of the situation removed every check the text had built.

The third criticism is that the judgment identified a constitutional impropriety and attached no consequence to it. Paragraph 2 of Constitutional Order 272 was held ultra vires by all five judges, in language about circuitous amendment that the critics quote with approval. Yet the amendment the paragraph effected had no work left to do once the Court held the recommendation unnecessary, and the outcome the paragraph was designed to produce, the declaration under clause (3), was upheld on independent grounds. The critics’ formulation is that the means were declared invalid and the end was upheld in the same breath, and that a holding of ultra vires which changes nothing teaches governments that procedural invention carries no cost, because a court that approves the destination will excuse the road. The related complaint, made by Alok Prasanna Kumar of the Vidhi Centre for Legal Policy among others, is that the Court resolved the largest federalism question in the case, the conversion of a state into Union territories, by accepting an assurance from the government’s law officer rather than by deciding the law, leaving the power’s limits undefined for the next occasion on which it may be used without any assurance being given.

The answers, as the judgment and its defenders give them, are as follows. To the first: the word in the proviso is “recommendation”, and the framers of the Constitution knew the difference between a recommendation and a consent, having used concurrence elsewhere in the same article where consent was meant; a recommendation the President was bound to obtain from a body that no longer existed would have made the termination power unusable, and would thereby have made permanent, by accident of drafting, a provision every contemporary account described as transitional. To the second: the discipline on action during President’s Rule is the discipline the Court articulated, reasonable nexus with the proclamation’s object and freedom from mala fides, and that discipline was applied, not waived; Parliament’s assumption of the State Legislature’s powers during a proclamation is the Constitution’s own arrangement in Articles 356 and 357, not an improvisation of 2019, and the alternative rule, that no act of consequence may be taken while a state is under President’s Rule, would make every proclamation a suspension of governance in the affected state for its duration. To the third: a court does not set aside an outcome that was lawfully reachable by a route plainly available, merely because a redundant and irregular step was also taken; the declaration under clause (3) required no recommendation, so the irregular device that supplied one could not vitiate it, and the holding of ultra vires stands as doctrine governing every future attempt to amend by interpretation, which is where its consequence lies. To the complaint about the assurance: the Union’s statement that statehood would be restored was made on behalf of the government in open court, was recorded in the judgment, and was converted into a direction for restoration at the earliest; a court that is told a measure is temporary, by the authority that took it, does not act improperly by holding the authority to its word instead of pronouncing on a wider power in the abstract.

This article adjudicates none of these exchanges. What can be said without taking a side is that the three criticisms and the three answers do not meet on common ground: the criticisms measure the judgment against the federal safeguards as a scheme of protection for states, and the answers measure it against the text of the provisions and the standards of review the Court has built for them. That divergence, rather than any single holding, is why the judgment continues to be argued about by lawyers who agree completely on what it held.

The durable doctrine, and what remains open

A judgment’s reputation rests on its result, and its doctrinal significance usually rests on a narrower holding. The result in this case ends the special status of Jammu and Kashmir, and it is the result that will be remembered. The narrower holding with the longest life is the fourth: an interpretation clause cannot be used to amend a substantive provision, and a power to apply a constitution with modifications is not a power to rewrite the provision that confers it. That holding is not about Kashmir. It governs every future occasion on which an executive, armed with a power to adapt or apply a constitutional text, is tempted to reach a result the text’s own procedure reserves to another route. The Court’s language on the point was general, the Bench was unanimous on it, and it decides a question of method that arises wherever interpretation clauses and adaptation powers appear, which is to say across the whole Constitution. Readers who work with the judgment as an advocate or a researcher will find that the paragraph counts for least in the reporting of the case and the most in its citation by later benches.

What does the judgment mean for interpretation clauses elsewhere in the Constitution?

The holding establishes that an interpretation clause may direct how a provision is read but may not change what the provision does or who holds the powers it distributes, and that any change of that character must travel by the amendment procedure the provision prescribes for itself.

The holding also sits comfortably in the ordinary law of interpretation, which is one reason it is likely to endure. Courts have long treated definition and interpretation clauses as servants of the substantive text: they resolve ambiguity in the words a statute uses, and they yield where the substantive provision’s own scheme shows a contrary intention. The vice the Bench identified in Paragraph 2 of Constitutional Order 272 was not that it used the interpretation clause energetically but that it used the clause to change the identity of an actor in a substantive procedure, converting a question of power into a question of meaning. Every constitution and every code contains adaptation and interpretation devices of this family, and governments reach for them precisely because they appear technical. The judgment’s answer, that a change in who must recommend, consent or decide is never a mere matter of interpretation, draws a line that administrators and drafters will have to respect in instruments having nothing to do with Jammu and Kashmir, from adaptation orders under Article 372 to the application clauses of reorganisation statutes themselves.

The practical consequences for the law of Jammu and Kashmir can be stated briefly, because the judgment leaves very little of the old scheme standing. Article 370 operates only in the restated form of its clause (1): all provisions of the Constitution of India, as amended from time to time, apply to the Union territory of Jammu and Kashmir and to Ladakh as they apply elsewhere in the Union. Article 35A, which had been inserted by the 1954 Order and had empowered the State Legislature to define permanent residents and their special rights, ceased to operate when the Order that contained it was superseded and the parent provision was declared inoperative; the judgment did not adjudicate Article 35A separately, and accounts attributing to the Court a distinct holding on it go beyond the text. The Constitution of Jammu and Kashmir is redundant and inoperative, as the lead opinion declared. Nothing in the judgment revives any element of the pre-2019 arrangement, and restoration of a special status in any form would require a constitutional amendment enacted for the purpose, a route the judgment neither opens nor closes because it was not asked to. This article is legal information and analysis of a decided case; it is not advice on any individual matter, and a person whose rights are affected by the changes the case upheld should take advice on their own position, since forums and limitation periods in this area are unforgiving. For readers preparing teaching or briefing material from the case, the holdings lend themselves to a structured deck, and working through them slide by slide in the ReportMedic presentation viewer is a serviceable way to test the six holdings, the one invalidity and the two silences against the table set out in this article.

One textual point completes the practical picture and is persistently misstated. Article 370 was not deleted from the Constitution. Constitutional Order 273 declared that all its clauses would cease to be operative except clause (1), and clause (1) survives in the restated form the Order gave it, under which all provisions of the Constitution, as amended from time to time, apply to the State the declaration concerned. The article therefore still stands in the printed text of the Constitution, in Part XXI, bearing its original marginal heading, but its operative content is the opposite of the scheme it once housed: where it formerly limited the application of the Constitution to Jammu and Kashmir, it declares that application complete. Clauses (2) and (3) and the provisos that generated the litigation have no operative force, and the mechanism of termination, having been used, cannot be used again in either direction. Readers comparing editions of the Constitution should therefore not expect to find the article absent; they should expect to find it present and spent, a condition without close parallel elsewhere in the text, and the source of recurring confusion in accounts that describe the article as having been removed from the Constitution altogether.

Two matters remain open in law, and both were left open deliberately. The conversion question stands undecided: whether Article 3 empowers Parliament to convert a state into one or more Union territories, as a durable act rather than as a temporary incident of a promised restoration, awaits a case in which the question cannot be avoided, and Justice Khanna’s observations about grave consequences and cogent grounds are the only judicial text a future bench has been given on it. The validity of the 2018 and 2019 proclamations stands unadjudicated on the record of these petitions, which matters less for Jammu and Kashmir than for the doctrine, since the standard the Court laid down for challenging actions taken under a proclamation will govern the next dispute in any state. Review petitions against the judgment were dismissed in May 2024, the Bench recording that there was no error apparent on the face of the record, so the judgment stands as the final word of the Court on every question it decided, and as no word at all on the two it did not.

The closing verdict is the one the artifact states. In Re: Article 370 of the Constitution is a unanimous judgment that upheld the ending of the special status, upheld the application of the entire Constitution to the State, upheld the carving out of Ladakh, directed elections by a date that was met and restoration of statehood at the earliest, and, inside that result, held one paragraph of one Constitutional Order beyond power in terms that will outlast the case. It decided the questions about Article 370 itself completely. It left the largest question about Article 3 untouched, deliberately and on the record. A reader who can hold both halves of that sentence at once understands the judgment better than most of the accounts written about it.

Frequently Asked Questions

Q: What did the Supreme Court hold about Article 370?

A Constitution Bench of five judges, unanimous in result, held that Article 370 was a temporary provision, that the President’s power under clause (3) to declare it inoperative survived the dissolution of the State’s Constituent Assembly, that Constitutional Order 273 validly ended the special status, and that the carving out of Ladakh as a Union territory was valid. The lead opinion of the Chief Justice, for himself and two colleagues, added that the State retained no element of sovereignty after accession and that one paragraph of Constitutional Order 272 was beyond power. Two judges wrote separate concurrences, and two questions, on conversion of a state and on the proclamations, were left undecided.

Q: Was Article 370 a temporary provision?

Yes. All three opinions agreed that Article 370 was temporary and transitional in character. The lead opinion rested the conclusion on the historical context of its inclusion and on its placement in Part XXI of the Constitution, the part devoted to temporary, transitional and special provisions, under a marginal heading that describes the article as a temporary provision for the State. Justice Kaul reached the same conclusion from a combination of the provision’s history, its text and the practice under it, and Justice Khanna recorded that the article was enacted as a transitional provision and did not have permanent character. The petitioners’ argument that the provision had become permanent when the State’s Constituent Assembly dissolved without recommending termination was rejected.

Q: Did Jammu and Kashmir retain sovereignty after accession?

The lead opinion held that it did not. The Chief Justice, for himself and Justices Gavai and Surya Kant, held that the State retained no element of sovereignty after the execution of the Instrument of Accession and the proclamation of 25 November 1949 adopting the Constitution of India, and that Article 370 was a feature of asymmetric federalism rather than sovereignty. Justice Kaul took a different view of the history: reading Prem Nath Kaul, he held that the State retained an element of internal sovereignty which Article 370 recognised, but that clause (3) supplied the mechanism to end that arrangement, so he concurred in the result. The difference between the opinions is one of reasoning; the operative conclusion was unanimous.

Q: What did the Court decide about Constitutional Order 272?

The Court split the Order into its two operations. Constitutional Order 272, issued on 5 August 2019, superseded the 1954 Order and applied all provisions of the Constitution of India to the State; that application was upheld as a valid exercise of the power under Article 370(1)(d), and the Court held the exercise was not mala fide merely because every provision was applied at once rather than piecemeal. Paragraph 2 of the Order also amended Article 367 so that the Legislative Assembly would be read in place of the Constituent Assembly in the proviso to Article 370(3). All five judges held that paragraph, insofar as it amended Article 370 through the interpretation clause, ultra vires and bad in law.

Q: Why was Constitutional Order 273 upheld?

Constitutional Order 273, issued on 6 August 2019 under Article 370(3), declared that all clauses of Article 370 would cease to be operative except clause (1) in a restated form. The Court upheld it because the power under clause (3) had survived the dissolution of the State’s Constituent Assembly in 1957, because the recommendation of that Assembly was not binding on the President and could not operate as a surviving condition of the power, and because the declaration was the culmination of a continuous and gradual process of constitutional integration conducted under clause (1) across seven decades. No mala fides in the exercise was shown, and absent mala fides the Court would not review the merits of the declaration.

Q: What did the Court expressly leave undecided about Article 370?

Two matters. First, whether Parliament may convert a state into Union territories under Article 3: the Court declined to determine the question because the Solicitor General had assured it that the statehood of Jammu and Kashmir would be restored, Ladakh apart, and it instead directed restoration at the earliest while upholding the Ladakh carve-out under Article 3 read with Explanation I. Second, the validity of the proclamations themselves, the Governor’s proclamation under Section 92 of the State Constitution and the Proclamation under Article 356 of December 2018 with its extension: the Court held that the challenge to the proclamations did not merit adjudication, since they had not been challenged until after the abrogation, and decided instead the standard governing actions taken under a proclamation.

Q: What directions were given on statehood and elections?

Two directions appear in the lead opinion. The Court directed that steps be taken by the Election Commission of India to conduct elections to the Legislative Assembly of Jammu and Kashmir, constituted under Section 14 of the Reorganisation Act, by 30 September 2024. Elections were held in September and October 2024 and an elected government took office in the Union territory in October 2024. The Court also directed that restoration of statehood take place at the earliest and as soon as possible. That second direction, unlike the first, fixed no date, and it converted into an obligation the assurance the Solicitor General had given during the hearing that statehood would be restored.

Q: What did the concurring opinions in the Article 370 case add?

Justice Kaul’s concurrence reasoned differently on sovereignty, holding that the State had retained an element of internal sovereignty which clause (3) was designed to de-recognise, and he added an epilogue recommending an impartial Truth and Reconciliation Commission on human rights violations in the State. Justice Khanna’s brief concurrence agreed with both opinions and added particulars, including the statement that abrogation does not negate the federal structure because the State’s citizens enjoy the same rights as citizens elsewhere, and a caution that converting a state into a Union territory has grave consequences and must be justified by very strong and cogent grounds in strict compliance with Article 3.

Q: What is the citation of the Article 370 case and which five judges decided it?

The case is In Re: Article 370 of the Constitution, decided on 11 December 2023, reported at [2023] 16 SCR 1 and cited as 2023 INSC 1058; the lead writ petition was Writ Petition (Civil) No. 1099 of 2019. The Constitution Bench comprised the Chief Justice, Dr D. Y. Chandrachud, and Justices Sanjay Kishan Kaul, Sanjiv Khanna, B. R. Gavai and Surya Kant. Three opinions were delivered: the lead opinion by the Chief Justice for himself and Justices Gavai and Surya Kant, a separate concurrence by Justice Kaul, and a short separate concurrence by Justice Khanna agreeing with both. The result on every operative question was unanimous.

Q: Why was the recommendation of the Constituent Assembly not binding on the President?

Because the proviso to Article 370(3) speaks of a recommendation, and a recommendation in constitutional usage advises rather than commands. The lead opinion reasoned that the Constituent Assembly’s role under the proviso was a transitional function attached to a body that was itself temporary, convened to frame the State’s Constitution and dissolved when that task ended in January 1957. When the Assembly ceased to exist, the function it alone could perform ceased with it, but the President’s power under the main clause was not constituted by the proviso and did not lapse. Earlier authority, including Sampat Prakash, had treated the President’s powers under Article 370 as continuing notwithstanding the dissolution, and the Bench declined to unsettle that construction.

Q: What part did the Sampat Prakash decision play in the Article 370 judgment?

Sampat Prakash v. State of Jammu and Kashmir, decided at the end of the 1960s, had upheld the continued exercise of the President’s powers under Article 370 long after the State’s Constituent Assembly had been dissolved, rejecting the contention that the provision had spent itself when the Assembly disappeared. In the 2023 case the petitioners invited the Court to confine or depart from that line, since their permanence argument depended on the dissolution having legal consequences Sampat Prakash had denied. The Constitution Bench instead treated Sampat Prakash, with Mohd. Maqbool Damnoo, as part of a consistent line of authority under which Article 370 remained operative and its mechanisms available, and built its third holding, the survival of the clause (3) power, on that foundation.

Q: Why did the Court reject the argument that applying the whole Constitution at once was mala fide?

The petitioners argued that clause (1)(d) was designed for gradual, provision-by-provision application, almost always with the concurrence of an elected State Government, and that applying the entire Constitution in a single order, during President’s Rule, betrayed the design and showed bad faith. The Court held that the exercise of power under clause (1)(d) to issue Constitutional Order 272 was not mala fide, and that such an exercise is not mala fide merely because all the provisions were applied together without following a piecemeal approach. A total application, the lead opinion reasoned, has the same effect as a declaration under clause (3), for which the concurrence of the State Government was never a requirement, so the manner of application could not be condemned by reference to a consent the total exercise did not need.

Q: What standard did the Court set for challenging decisions taken during President’s Rule?

The lead opinion laid down a four-part position drawn from S. R. Bommai and Rameshwar Prasad. The exercise of power by the President after a proclamation under Article 356 is subject to judicial review. The exercise must bear a reasonable nexus to the object of the proclamation. A challenger must first make out a prima facie case that the exercise was mala fide or extraneous; only then does the onus shift to the Union to justify it. And Parliament’s power under Article 356(1)(b) extends to the whole of the State Legislature’s power, legislative and non-legislative alike, so an act cannot be attacked merely because the function Parliament performed was not a law-making one. On the facts, no prima facie case of mala fides was made out.

Q: On what ground did the Court uphold the carving out of Ladakh as a Union territory?

The lead opinion upheld the decision to carve out the Union territory of Ladakh in view of Article 3(a) read with Explanation I to Article 3. Clause (a) of Article 3 empowers Parliament to form a new state by separation of territory from any state, and Explanation I provides that the word “State” in clauses (a) to (e) includes a Union territory, from which the Court derived the power to form a Union territory by separating territory from a state. The holding is confined to that operation, the separation of territory and its constitution as a Union territory. It was expressly not a decision on the distinct question whether a state may be converted, as a whole, into Union territories, which the Court left open.

Q: Was the Truth and Reconciliation Commission a direction of the Court or only a recommendation?

Only a recommendation, and of one judge. Justice Kaul, in an epilogue to his concurring opinion, recommended the setting up of an impartial Truth and Reconciliation Commission to investigate and report on violations of human rights by state and non-state actors in Jammu and Kashmir at least since the 1980s, and to recommend measures for reconciliation. He proposed that the exercise be time-bound and completed before memory escapes, and he left the manner of constituting the commission to the government, citing the sensitivities involved. The lead opinion contains no such recommendation, the operative order creates no commission, and reports describing the Court as having ordered an inquiry into human rights violations misstate the judgment.

Q: What caution did Justice Khanna record about converting a State into a Union territory?

Justice Khanna observed that Union territories are normally geographically small territories, or are created for aberrant reasons or causes. The conversion of a state into a Union territory, he held, has grave consequences: among others, it denies the citizens of the state an elected state government and it impinges on federalism. Conversion or creation of a Union territory from a state therefore has to be justified by giving very strong and cogent grounds, and it must be in strict compliance with Article 3 of the Constitution. He recorded those observations while concurring in the decision not to determine the conversion question, given the Union’s assurance that statehood would be restored, so the passage operates as a caution for any future case rather than as a holding.

Q: Why did the Court describe the Constitution of Jammu and Kashmir as redundant?

The lead opinion held that, following the application of the Constitution of India in its entirety to the State, the Constitution of Jammu and Kashmir is inoperative and was declared to have become redundant. The reasoning is one of occupation of the field: the Constitution of India is a complete code for constitutional governance, and once it applied to the State without exceptions or modifications, no operative space remained for a parallel state constitution framed under the earlier dispensation. The State Constitution was not struck down by a separate holding, and it was not amended; it ceased to operate as a consequence of the declaration under Article 370(3) and of Constitutional Orders 272 and 273, whose validity the Court upheld.

Q: What did the Court hold about the views of a state legislature under the first proviso to Article 3?

That they are recommendatory. The first proviso to Article 3 requires that a Bill affecting the area, boundaries or name of a state be referred by the President to the legislature of that state for expressing its views within a period the President specifies. Following the line of authority from Babulal Parate v. State of Bombay, the lead opinion held that the views of the State Legislature under the first proviso are recommendatory in character and do not bind Parliament. On that footing, Parliament’s exercise of the power during the Proclamation, when Parliament itself discharged the functions of the dissolved State Legislature, was valid and not mala fide. Justice Kaul added that the views of a dissolved legislature could not in any event have been ascertained.

Q: What happened to Article 35A as a result of the Article 370 judgment?

Article 35A was not a part of the Constitution as enacted; it was inserted into the constitutional scheme as it applied to Jammu and Kashmir by the Constitution (Application to Jammu and Kashmir) Order, 1954, and it empowered the State Legislature to define permanent residents and to confer on them special rights in employment, property and settlement. Constitutional Order 272 superseded the 1954 Order, and Constitutional Order 273 then declared Article 370 inoperative except in its restated clause (1) form. Article 35A accordingly ceased to operate as a consequence of those Orders. The judgment contains no separate holding adjudicating Article 35A, and its fate follows from the validity the Court accorded to the two Orders, not from any independent reasoning about permanent residents.

Q: Can the special status of Jammu and Kashmir be restored after the judgment?

Not through Article 370 itself. The judgment upheld the declaration under clause (3) that rendered the article inoperative except for its restated clause (1), under which all provisions of the Constitution of India apply to Jammu and Kashmir and Ladakh as amended from time to time. The mechanism by which the special status was created and ended has therefore been exhausted on the Court’s construction, and nothing in the judgment preserves a route by which a President could revive the pre-2019 arrangement by order. A special constitutional position for the region could be created again only by a constitutional amendment under Article 368 enacted for that purpose, a question of political choice on which the judgment expresses no view. The restoration the Court directed was of statehood, which is a different matter.