The Emergency has never really ended

The story Indian constitutional law tells about itself is that the Emergency of 1975 to 1977 was a closed parenthesis. The proclamation came; the Court briefly disgraced itself in ADM Jabalpur (Shivakant Shukla)¹; the people threw out Mrs Gandhi at the polls; the Janata Parliament passed the 44th Amendment; Articles 20 and 21 were placed beyond the reach of any future Presidential Order; the Cabinet's written advice was required to commence an emergency; the door was bolted from inside. After 1978, the textbooks say, the Emergency could not happen again in that form. I want to argue that this account is wrong in a way that is not merely inconvenient. The proclamation ended. The constitutional order it produced did not. It survived inside the preventive-detention statute book, inside the executive's standing power to switch off communication, and inside Part IVA of the Constitution itself. The Emergency ended as proclamation in 1977. As habit and doctrine, it stayed.

The settled view, taken at its strongest

Let me say first what the settled view actually claims, because polemic that ignores its target is shouting and I want to do something else. The settled view is not naive. It rests on three propositions and each of them is true.

The first is textual. The 44th Amendment changed Article 352 in a way that makes the 1975 mechanism legally impossible. Mrs Gandhi proclaimed the Emergency on the President's satisfaction alone, on the advice of the Prime Minister alone, with the Cabinet first informed and then asked to ratify. Today the President cannot proclaim under Article 352 unless the Union Cabinet recommends it in writing.² The proclamation lapses in a month unless both Houses approve, and must be re-approved every six months. One-tenth of the Lok Sabha can force a special session in fourteen days to revoke it by simple majority. The phrase 'internal disturbance' was excised from Article 352(1) and replaced with 'armed rebellion'. None of this is a small change. The framers' Article 352, with its single executive trigger, was an instrument designed for one kind of person; the post-1978 Article 352 is designed for a different one.

The second proposition is rights-protective. The defining horror of Shivakant Shukla was that the High Courts were told they could not entertain a habeas corpus petition during an emergency, because Article 359 had suspended the enforcement of Articles 14, 21 and 22 by Presidential Order. That trapdoor has been welded shut. The 44th Amendment's proviso to Article 359(1) excludes Articles 20 and 21 from any future suspension. A Presidential Order cannot reach them. The right to life and the right not to be punished retroactively are now, in the constitutional text, beyond emergency suspension. The Supreme Court completed the work in K.S. Puttaswamy v UOI when D.Y. Chandrachud J., writing for himself and three other judges of the nine-judge bench, formally overruled Shivakant Shukla.⁷ The judgment that gave the executive a free hand over the body of the citizen during an emergency is no longer good law.

The third proposition is institutional. The Court that decided Shivakant Shukla in 1976 was an institution that had been intimidated, packed by supersession, and rendered ideologically uncertain. The Court of 2026 is a very different body — not always brave, not uniformly principled, but unrecognisable from the bench of A.N. Ray, M.H. Beg, P.N. Bhagwati, and Y.V. Chandrachud J. that surrendered habeas corpus in Shivakant Shukla four to one. Indira Nehru Gandhi v Raj Narain⁶ was decided during the Emergency itself and struck down a constitutional amendment shielding the Prime Minister's election. Minerva Mills followed two years later, undoing the 42nd Amendment's worst clauses; Bommai in 1994 brought Article 356 under meaningful judicial review. The doctrinal corrections came quickly, and the institutional confidence to make them came soon enough after.

Take all three of these together and the conclusion writes itself. The Emergency, as a constitutional event, was a one-off. We diagnosed it; we corrected it; the corrections were textual, doctrinal, and institutional; the system has never since produced anything that resembles it. To say the Emergency 'continues' is to flatten real institutional differences and to insult the Janata Parliament that did the corrective work. This is the settled view. It is honestly held by serious people. I disagree with it, but it deserves to be stated in its strongest form before being argued against.

Where the settled view fails: preventive detention as the through-line

The settled view fails first at the place where the Emergency hurt people most. Preventive detention is the mechanism that filled MISA cells in 1975 and 1976. MISA was repealed in 1978. The settled view treats this as the closing of the question. It is not. MISA was the second major preventive-detention statute, not the first; it was the Preventive Detention Act of 1950, succeeded by MISA, succeeded by the National Security Act of 1980, succeeded by TADA in 1985 and 1987, succeeded by POTA in 2002, succeeded by the amended UAPA from 2008 onward. None of these statutes was a copy of MISA. Each one was different in some particular and continuous in the essential. The essential is the power to detain a person on the executive's satisfaction, without trial, on the basis of materials the detenu cannot fully see, for periods that begin in months and accumulate into years.

The constitutional cover for this entire series is Article 22(3) and the carve-outs that follow it.⁴ Article 22(3) tells us that the procedural safeguards in clauses (1) and (2) — the right to be informed of the grounds of arrest, the right to consult a lawyer, the right to be produced before a magistrate within twenty-four hours — do not apply to a person detained under any law providing for preventive detention. The framers wrote this carve-out into the Constitution itself in 1949. They wrote it because they expected preventive detention to remain an instrument of the State after independence; they wrote it because the Constituent Assembly could not bring itself to ban what the colonial regime had used so frequently against the people now writing the Constitution. Article 22 is the part of the Bill of Rights that is least like a Bill of Rights. It is, structurally, a permission slip to the State.

The 44th Amendment did not touch Article 22(3). It did not touch the carve-out. It did not touch the principle that the State may, on satisfaction recorded internally, deprive a citizen of liberty without trial for an extended period. What it did touch was the duration of detention without an Advisory Board's view, lengthening the constitutional minimum and allowing Parliament to legislate longer ones — which Parliament did, almost immediately, in the National Security Act, and again in TADA, and again in POTA, and continues to do under UAPA.⁵ The UAPA in its current form allows detention for 180 days before a chargesheet must be filed; sets bail standards under Section 43D(5) that the Supreme Court has held require the accused to disprove the prosecution's prima facie case to obtain bail; permits the Union to designate individuals as terrorists by executive notification without a hearing; and is now the principal vehicle for political detention in the Republic.

The settled view says: but UAPA is not MISA. There are courts; there are chargesheets; there are bail applications. This is true in form and false in the matter that counts. The matter that counts is whether the executive can take a citizen's liberty for years without a trial that produces a verdict on guilt. Under UAPA, on the existing bail jurisprudence, the answer is yes. Under MISA the answer was also yes. The constitutional architecture that permits both is Article 22, which the 44th Amendment did not touch and which the Supreme Court has not, in seventy-six years, cut back to anything resembling the safeguards available to a person detained for an actual offence.

The Emergency made unusual use of MISA. The post-Emergency order has made ordinary use of MISA's successors. The volume is lower; the architecture is identical; the constitutional cover is the same. To call this 'the Emergency continuing' is not to deny that the proclamation has ended. It is to insist that the structural feature the Emergency exposed — preventive detention as a freestanding executive power, constitutionally insulated from ordinary criminal-process safeguards — was not corrected. It was incorporated.

Where the settled view fails: Article 359, repurposed

The second place the settled view fails is more subtle and to me more revealing. Article 359 of the Constitution permits the President, during an emergency, to suspend the enforcement of fundamental rights other than Articles 20 and 21. Read narrowly, the article is a device for one specific situation that has not been used since 1977. Read for what it actually authorised, it is a device for converting fundamental rights from claims against the State into permissions held by the State and revocable at the State's convenience. The 44th Amendment's contribution was to ringfence two of those rights — the criminal-procedure right under Article 20 and the right to life under Article 21 — from any future Presidential Order. The other rights, including Article 19 in its entirety, can still be reached. The mechanism that produced the worst of the Emergency was not formally retired. It was narrowed.

What the post-Emergency state has done is something the framers of the 44th Amendment did not anticipate. It has learned to do without Article 359. It has discovered that ordinary statutes, ordinary executive orders, and ordinary administrative discretion can produce most of what Article 359 once produced, without needing to declare an emergency at all. Section 144 of the old Code of Criminal Procedure, now Section 163 of the BNSS, gives a District Magistrate the power to prohibit assemblies, processions, and movements indefinitely on the executive's satisfaction. The Telegraph Act of 1885 and the Telecom Suspension Rules of 2017 give the Union and the States the power to suspend internet and telephone services on the executive's satisfaction. The Foreigners Act, the Public Safety Act of Jammu and Kashmir, and the National Security Act allow administrative detention without the Article 352 machinery being engaged at all. The Information Technology Act gives the Union the power to block specified online content, with the rules requiring confidentiality of the takedown order itself.

Each of these powers, taken alone, is defensible. Each has been read by the Court at one point or another to require some form of proportionality, recorded reasons, or periodic review. Anuradha Bhasin v UOI⁸ is the high-water mark of this jurisprudence. The Court held that a complete and indefinite suspension of internet services in Jammu and Kashmir was impermissible, that proportionality applied, that orders had to be published and re-justified, that judicial review was available. The judgment is widely celebrated. What it did not do is order the lifting of the shutdown. It did not free a single detenu. It returned the matter to the executive with instructions to write better orders. The shutdown continued for months after Bhasin was decided.

This is the technique. The constitutional cover is no longer Article 359; it is the Telecom Rules, the BNSS, the Public Safety Act, and the UAPA. The procedural review is not the absence of review but its conversion into a form the executive can satisfy by drafting better paperwork. The result, on the ground, is what the citizen of 1976 would have recognised as Emergency conditions — sustained communication blackout, prolonged administrative detention, restrictions on movement and assembly imposed by executive fiat — produced through ordinary legal channels and reviewed, when reviewed at all, on a standard that asks only whether the executive considered the relevant material. The framework of the 44th Amendment was built for the wrong threat. It was built for a Prime Minister who would proclaim an emergency. The successor state has learned to govern as if an emergency were always in operation, while never proclaiming one.

Where the settled view fails: the 42nd Amendment did not go

The third failure is textual and obvious enough that I find it strange that it is not made more often. The 44th Amendment is celebrated as the constitutional repudiation of the Emergency. The 42nd Amendment is treated as the body of work the 44th repudiated. This is roughly half right. The 44th rolled back the 42nd's most aggressive incursions — the attempt to bar judicial review of Emergency proclamations, the lengthening of Lok Sabha terms to six years, the supremacy of Directive Principles over Fundamental Rights for a class of legislation. It did not roll back several other things the 42nd Amendment did, and those surviving footprints are not minor.³

Part IVA of the Constitution — the Fundamental Duties — was inserted by the 42nd Amendment in 1976. It is still there. It was a Soviet-style addition lifted, the Swaran Singh Committee acknowledged, from Article 59 of the 1977 Soviet Constitution. The duty 'to abide by the Constitution', 'to cherish and follow the noble ideals which inspired our national struggle for freedom', and 'to safeguard public property and to abjure violence' are not, in themselves, objectionable. What is objectionable is the constitutional logic of inserting them. Bills of Rights catalogue what the State owes the citizen. Catalogues of duties owed by the citizen to the State are a different kind of document; they are the language of a State that conceives of the citizen as instrumentally accountable to it. The framers of 1949 did not include such a catalogue. The framers of 1976 did. The Janata Parliament left it in.

The Preamble itself was amended by the 42nd. The words 'socialist' and 'secular' were inserted into the description of the Indian State. There has been litigation about whether these insertions changed anything; the Supreme Court has said they did not. The point is narrower. The Preamble, which the framers chose with care, was edited during the Emergency by a Parliament whose democratic legitimacy was at its lowest. The 44th Amendment did not undo the edit. It accepted it. The Constitution as it stands today is a document drafted in 1949, amended in major part in 1976 by an Emergency Parliament, and corrected only partially in 1978. Its preambular self-description is the Emergency Parliament's. We have lived with this so long that we no longer notice. We should notice.

Beyond the Preamble and Part IVA, the Seventh Schedule changes — the transfer of education, forests, weights and measures, and the administration of justice from the State List to the Concurrent List — were 42nd Amendment work and remain in the Constitution today. They have shaped Centre-State legislative competence ever since. None of them were Emergency-essential; all of them were enacted as part of the Emergency project of recentralising power. The 44th did not return them to the State List.

The alternative: name the continuity

If the settled view is wrong, what is the alternative? The alternative is to give up the parenthesis. The Emergency was not a closed episode; it was a moment at which structural features of the Indian constitutional order — the carve-outs, the executive discretion, the deference of the Court to administrative satisfaction, the Constitution's tolerance for liberty-restricting laws so long as they bore the right recitals — were used at maximum amplitude. The amplitude has come down. The features have not.

This reframing has consequences for how we teach the Constitution and for how we litigate it. It means treating Article 22 not as a procedural article that the framers reasonably excluded from the ordinary criminal-process guarantees, but as a structural authorisation of preventive detention that the framers should not have included and that the Court should have read down. It means treating Anuradha Bhasin not as a victory but as the last clear opportunity the Court had to read Article 19 in the digital age in a way that made indefinite executive shutdown impossible — an opportunity the Court took partially, leaving the substantive test in the executive's hands. It means treating the 42nd Amendment's surviving footprints as legitimate objects of constitutional repair, not as settled background.

Most of all it means treating the elevation of preventive-detention statutes — from MISA's emergency-era prominence to UAPA's normal-times prominence — as the central story of the post-Emergency constitutional order, rather than as a footnote to it. The argument is not that the Indian state today is governed under an emergency. The argument is that the operative substance of the 1975 Emergency — extended administrative detention without trial, executive control over communication and assembly, a constitutional architecture that places the burden of justification on the citizen rather than on the State — was substantially preserved through the 44th Amendment and substantially extended in the decades since. We have not lifted the Emergency. We have learned to govern through its instruments without needing to declare it.

The counter-argument, addressed honestly

The strongest reply to all this is also the most honest, and it deserves a section to itself. It runs as follows. The Emergency was not just a set of statutes. It was the proclamation, the suspension of habeas corpus, the censorship of the press by prior restraint, the forced sterilisations, the demolitions, the arrest of the entire opposition leadership in a single night. None of these things has happened since. The 44th Amendment, the post-1978 Court, the press, the political opposition, the civil-society infrastructure that did not exist in 1975 — these together make a recurrence of that scale impossible. To equate UAPA detention with MISA detention is to flatten precisely the differences that the 44th Amendment was designed to create. The detenu under UAPA can move a court. The detenu under MISA could not. That is not a small thing.

I take this seriously. I take it seriously enough that I will not pretend the difference is rhetorical. It is real and it matters. A person detained under UAPA today can file a habeas corpus petition; the High Court will entertain it; the courts will examine the order under Section 482 of the BNSS; bail under Section 43D(5) is hard but not categorically barred. The press, however constrained by ownership and informal pressure, is not under prior censorship. The opposition, however depleted, sits in the Lok Sabha and is not in jail. The judges, however selected, are not the judges of ADM Jabalpur. The reply is correct that the worst features of 1975 to 1977 — the comprehensive simultaneity of suppression, the proclamation hanging over everything — have not recurred.

The reply is also too narrow. It treats the Emergency as defined by its peak intensity rather than by its instruments. The settled view, in its strong form, says: the peak cannot be reached again, therefore the Emergency is over. My reply is that the peak need not be reached for the architecture to be doing the same work at lower amplitude. The 1975 Emergency took eighteen months to detain roughly one hundred thousand people under MISA. The post-Emergency order, drawing on the National Security Act, the Public Safety Act, TADA, POTA, and UAPA across forty-six years, has detained a comparable cumulative population without ever needing the Article 352 machinery. The 1975 Emergency censored the press by Order. The post-Emergency order has used the IT Rules, the Telegraph Act, defamation law, and selective tax enforcement to produce a press whose self-censorship is as effective as direct censorship was. These differences in form are real. The reply is that they are also the achievement the 44th Amendment claims credit for. They are not. The instruments that produce them are not 1978 reforms; they are 1980, 1987, 2002, 2008, 2017, and 2019 enactments, layered over a constitutional architecture the 44th Amendment did not change.

What changes if this is right

Three things change if this account is correct. The first is the way we read the Shah Commission report.⁹ The Commission was the Janata government's attempt to produce an authoritative account of Emergency abuses. Its three interim reports documented detentions, deaths in custody, suppression of the press, the conduct of the Maintenance of Internal Security Act administration. The reports were tabled in 1978. After the Congress returned in 1980, the reports were withdrawn from public circulation. The withdrawal was administrative; no statute repealed them; they were simply pulled from official sale and from libraries. We do not have an authoritative account of what the Indian State did to its citizens between 1975 and 1977. The closest we have is a set of reports that the State itself decided we should not consult. That this happened in 1980 and has not been undone in forty-six years is itself evidence of the continuity I am pointing to. A constitutional order that treats its own diagnostic record as politically inconvenient has not closed the book on the Emergency. It has merely closed the cover.

The second change is in how we read Maneka Gandhi v UOI¹⁰. The standard account treats Maneka as the moment Indian fundamental-rights jurisprudence came of age — the moment Article 21 was read to require fair, just, and reasonable procedure. The standard account is right about the doctrinal content. It is wrong about the meaning. The judges who decided Maneka in January 1978 were, in significant part, the judges who decided Shivakant Shukla in April 1976. P.N. Bhagwati J., who in Shivakant Shukla declared that judicial conscience must find its resting place in the letter of the constitutional text, in Maneka read into Article 21 a substantive due-process review the text does not announce. The doctrinal U-turn was complete. The institutional reflection was not. The Court that delivered Maneka never sat to explain why it had delivered ADM Jabalpur. The repudiation came through doctrine, not through accountability. Puttaswamy in 2017, with its explicit overruling of Shivakant Shukla, came thirty-nine years late and from a bench none of whose members had been on the 1976 Court. Doctrinal repair is real but it is not the same thing as institutional repair, and the difference matters when the next emergency-equivalent arrives.

The third change is in how we read the use of these instruments in Jammu and Kashmir between August 2019 and the present.¹¹ The detentions under the Public Safety Act, the communications blackout, the administrative restrictions on movement and assembly, the holding of mainstream political leaders without charge for periods running to months and in some cases more than a year — none of these required Article 352. None of them produced a constitutional crisis comparable to 1975. They were processed through ordinary administrative law, reviewed by the Supreme Court in Anuradha Bhasin and connected matters, and substantially upheld in their procedural form. The settled view's response is: this is not the Emergency, this is national security in a contested region under specialised laws. My response is: the citizen of Srinagar in late 2019 would not have been able, from her own experience, to say that her constitutional position differed materially from the citizen of Allahabad in late 1975. The architecture that produced her position did not need a proclamation. That is the achievement of the post-Emergency constitutional order, and it is not the achievement the 44th Amendment is celebrated for.

The Emergency, as proclamation, is over. The Emergency, as the underlying constitutional logic that allows the State to take liberty without trial, communication without notice, and assembly without permission — that Emergency is the operating system. The 44th Amendment patched the user interface. The kernel is unchanged. To say this is not to dismiss the patch. The patch matters. It is to refuse the consoling story that the patch is the whole repair, and to insist that the architecture beneath it is the next constitutional question this Republic has not yet been willing to ask.

Get the week's biggest decisions, distilled.

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