The judgment in Krishna Kumar Singh v State of Bihar, reported at (2017) 3 SCC 1 and delivered on 2 January 2017, is the point at which the Supreme Court stopped merely deploring the re-promulgation of ordinances and instead rearranged the legal consequences of governing by ordinance. A bench of seven judges held four things. Re-promulgation of an ordinance is a fraud on the Constitution and a subversion of the democratic legislative process. The satisfaction of the President or the Governor that immediate action is necessary is open to judicial review on settled and limited grounds. The laying of an ordinance before the legislature is a mandatory constitutional obligation, and a failure to lay is a serious constitutional infraction. And, in the holding that changed settled law, an ordinance that has ceased to operate does not by default leave behind rights, privileges, obligations and liabilities that survive it: survival is a question of construction, answered case by case through a test of public interest and constitutional necessity. The first three holdings restated and hardened doctrine that had been available since D.C. Wadhwa in 1987. The fourth displaced two Constitution Bench decisions, one from 1962 and one from 1985, which had treated an ordinance on the analogy of a temporary Act whose effects endure after its expiry.

This article owns the judgment and its doctrine. The ordinance power itself, its textual conditions in Articles 123 and 213, its session and duration limits and its disapproval mechanism, belongs to the companion treatment of the ordinance-making power in India, and the questions that article answers are not repeated here. What follows is an account of how the doctrine was built, what the seven judges decided holding by holding, where the dissent parted company, and what the decision means for the person whose appointment, payment or entitlement rests on an ordinance that has ceased to operate. That last question is the one the judgment most changed, and it is the one least often taught.

The dispute that forced the questions

The litigation began with schools. In 1989 the Governor of Bihar promulgated the Bihar Non-Government Sanskrit Schools (Taking Over of Management and Control) Ordinance, by which the State took over the management and control of 429 private Sanskrit schools, and the teachers and staff of those schools were to be treated on the footing of State employment. The Ordinance was not converted into an Act. When its life ran out it was promulgated again in the same terms, and then again, through seven successive re-promulgations recorded in the judgment, and none of the ordinances in the series was ever laid before the Bihar Legislative Assembly. The teachers and other employees of the taken-over schools petitioned the Patna High Court for salary and the emoluments of government service, resting their claim on the ordinances under which the takeover had been effected.

The Patna High Court dismissed the writ petitions. Relying on D.C. Wadhwa, it held the serial re-promulgation unconstitutional, and a claim founded on instruments produced by that practice could not be sustained. The teachers carried the matter to the Supreme Court. In the Supreme Court the case acquired a second and larger dimension. The appellants’ position drew support from T. Venkata Reddy v State of Andhra Pradesh, (1985) 3 SCC 198, in which a Constitution Bench had held that the validity of an ordinance could not be questioned on the ground that the executive’s motives were improper, and, following State of Orissa v Bhupendra Kumar Bose, 1962 Supp (2) SCR 380, that rights and liabilities created under an ordinance endure after the ordinance ceases to operate, on the analogy of a temporary statute. Those propositions sat uneasily with the later jurisprudence of the Court on judicial review of constitutional functionaries, in particular S.R. Bommai v Union of India, and the conflict was the reason the matter was placed before a bench of seven judges for an authoritative pronouncement. A bench larger than the five judges who decided Bhupendra Kumar Bose and T. Venkata Reddy was required because those decisions could be displaced only by a bench of greater strength.

The questions before the seven judges therefore ran in two directions at once. One set concerned the discipline of the power: whether re-promulgation is permissible, whether the satisfaction behind an ordinance can be examined, and whether laying is obligatory. The other concerned the residue of the power: what remains when an ordinance dies. The first set had been substantially answered in Wadhwa thirty years earlier, though without durable effect on practice. The second had been answered in 1962 and 1985 in terms that made re-promulgation worth practising, because a government that could not keep an ordinance alive could at least rely on its consequences surviving. The significance of Krishna Kumar Singh is that it answered both sets together, and closed the loop between them.

D.C. Wadhwa and the anatomy of ordinance raj

The doctrinal starting point is D.C. Wadhwa v State of Bihar, (1987) 1 SCC 378, a Constitution Bench decision delivered on 20 December 1986, with Bhagwati C.J. writing for the Court. Dr D.C. Wadhwa, a professor of economics, had researched the ordinance practice of Bihar and published his findings in a book whose title the Court adopted as the description of the practice: Repromulgation of Ordinances: Fraud on the Constitution of India. The research, reproduced in the judgment, recorded that between 1967 and 1981 the Governor of Bihar promulgated 256 ordinances, and that all of them were kept alive for periods ranging from one year to fourteen years by re-promulgation from time to time. Of the 256, sixty-nine were re-promulgated several times and kept alive with the prior permission of the President. The judgment reproduced the life-groups of the ordinances in a table whose cumulative picture was of a parallel statute book: measures on sugarcane supply, land, education and local government that had never been enacted by the legislature at all, yet governed Bihar for years, in one case for thirteen years and in another for a period within the fourteenth year.

The mechanism was administrative routine rather than concealment. An ordinance would be promulgated when the legislature was not in session. The session would come and go without the ordinance being converted into a Bill. When the ordinance approached its constitutional limit, it would be promulgated again in identical terms. The Court in Wadhwa held that this practice was unconstitutional in terms that Krishna Kumar Singh later quoted and reaffirmed. The power to promulgate an ordinance is conferred to meet an emergent situation and is an exceptional power, not a parallel source of legislation. A systematic practice of re-promulgation, carried on without placing the measure before the legislature, is a fraud on the Constitution and a subversion of the democratic legislative process, because it transfers the law-making function from the body the Constitution charges with it to the executive, which is responsible to that body and acts in its name only within the limits the ordinance clauses allow.

Wadhwa was not, however, a judgment with teeth of its own. The Court struck down the particular ordinance before it, the Bihar Intermediate Education Council Ordinance of 1985, as unconstitutional, but for the general practice it expressed the hope and trust that the constitutional authorities would not repeat it. No consequence was attached to a failure to lay, no general ruling was given on the fate of rights created under re-promulgated ordinances, and the enduring rights doctrine of Bhupendra Kumar Bose stood untouched, so a government that lost an ordinance still kept its effects. The three decades between Wadhwa and Krishna Kumar Singh are the demonstration of what a condemnation without consequences achieves. Bihar’s practice had been the extreme case, but the technique of serial re-promulgation remained available across India, and the Sanskrit schools ordinances that produced Krishna Kumar Singh were themselves promulgated, and re-promulgated seven times, after Wadhwa had declared the practice a fraud on the Constitution. The 2017 bench knew it was writing the second judgment on a subject on which the first had been ignored, and the design of the second judgment, attaching legal consequences to each element of the practice, is best understood as a response to that failure.

What role did Dr Wadhwa’s research play in building the doctrine?

Dr Wadhwa supplied the facts on which the doctrine rests. His research counted every ordinance promulgated in Bihar between 1967 and 1981, the repetitions of each, and the total period each spent in force: 256 ordinances kept alive for up to fourteen years. The Supreme Court reproduced his tables and drew the constitutional conclusion from them.

The doctrine of re-promulgation as constitutional fraud rests on that empirical foundation. A single re-promulgation might be an expedient; 256 ordinances kept alive for up to fourteen years was a system, and a system of executive legislation was what the Constitution forbids. Krishna Kumar Singh inherited the conclusion because the record had been built in that form.

There is an institutional lesson in the route the research travelled. The facts were gathered outside government, published as a book, carried into court as the annexures to a writ petition under Article 32, and adopted by a Constitution Bench as the factual basis of a constitutional rule. Neither the Bihar legislature nor any audit institution had assembled the count, although every figure in it came from the State’s own gazettes. The episode is a standing illustration of why the documentary record of ordinance practice matters: each individual re-promulgation looks like a minor administrative act, and only the aggregated record discloses the constitutional violation. It also explains a feature of the later judgment. In Krishna Kumar Singh the Court did not need fresh fact-finding about the practice in general, because Wadhwa’s record and reasoning were available to it; what it needed, and what it supplied, was the set of consequences Wadhwa had omitted.

The old law on what an ordinance leaves behind

The holding that Krishna Kumar Singh changed can only be measured against the position it displaced, and that position had two pillars. The first was State of Orissa v Bhupendra Kumar Bose, decided in 1962 by a Constitution Bench of five judges. The case arose from an ordinance of the Governor of Orissa which had invalidated an election to the Cuttack Municipality and ordered a fresh election, and which had ceased to operate by the time the dispute matured. The question was whether the acts done under the ordinance, including the invalidation of the election, fell with it. The Court held that they did not. An ordinance, the Court reasoned, is a law made under a power conferred by the Constitution and has the same force as an Act; the correct analogy was with a temporary Act, and under the law governing temporary enactments, proceedings and effects completed during the life of the enactment are not undone by its expiry unless a contrary intention appears. The rights and liabilities acquired under the ordinance were therefore treated as surviving its cessation, and the fresh election ordered under it stood.

The second pillar was T. Venkata Reddy v State of Andhra Pradesh, (1985) 3 SCC 198, also a Constitution Bench of five judges. The Andhra Pradesh Abolition of Posts and Prohibition of Appointments Ordinance had abolished certain posts of part-time village officers and barred fresh appointments. The challenge contended that the ordinance was a colourable exercise, actuated by the motive of removing the incumbents. The Court rejected the challenge at its root. An ordinance is a legislative act of the President or the Governor, and just as the motives of a legislature in enacting a law cannot be questioned, the satisfaction behind an ordinance could not be examined on the ground of mala fides in the manner contended; and following Bhupendra Kumar Bose, the Court treated the consequences of the ordinance as capable of enduring beyond its life. The two decisions together gave the executive a comfortable jurisprudence: the satisfaction was effectively beyond scrutiny, and the effects of an ordinance were effectively permanent. A government minded to legislate by ordinance faced a judicial doctrine under which even failure carried little cost, since what had been done under the ordinance remained done.

Two features of that old position deserve emphasis, because the 2017 majority fastened on both. The first is that the enduring rights theory rested on an analogy, not on the text of Articles 123 and 213. The Constitution states what happens to an ordinance: it ceases to operate at the expiration of six weeks from the reassembly of the legislature, or earlier upon disapproval, and it may be withdrawn at any time. The Constitution also shows, in provisions such as Articles 249(3), 250(2), 357(2), 358 and 359(1A), that when the framers intended the effects of a time-limited law to be saved, they wrote a saving clause saying so. Articles 123 and 213 contain no saving clause for rights, privileges, obligations or liabilities arising under an ordinance that has ceased to operate. The temporary Act analogy filled that silence with a presumption drawn from a different legal category, and it was the analogy, rather than any constitutional text, that the seven judges eventually rejected.

The second feature is the relationship between the enduring rights doctrine and re-promulgation. So long as consequences survived automatically, re-promulgation was a low-risk instrument even when it failed: the worst outcome of the practice was the position the executive had already achieved under the ordinance. The doctrine therefore blunted Wadhwa. A declaration that re-promulgation is a fraud on the Constitution carries limited deterrent force if the fruits of the fraud are secure. The majority in Krishna Kumar Singh understood the two questions as one system, and its treatment of the residue question is the mechanism by which its restatement of the re-promulgation rule acquired practical force. That is also why the enduring rights holding, which looks technical beside the rhetoric of fraud, is the part of the judgment that decides cases.

The satisfaction question before 2017

The reviewability of the satisfaction behind an ordinance had its own history, and the seven judges had to bring three strands of it together. The first strand is the early willingness of the Court to look behind a promulgation. In R.C. Cooper v Union of India, (1970) 1 SCC 248, the Bank Nationalisation case, the majority accepted that an ordinance could be challenged on the ground that the satisfaction was mala fide, even though the challenge failed on its facts. The second strand is constitutional amendment. The Constitution (Thirty-eighth Amendment) Act, 1975 inserted clause (4) into Articles 123 and 213, declaring the satisfaction of the President or the Governor to be final and conclusive and not questionable in any court on any ground. The Constitution (Forty-fourth Amendment) Act, 1978 omitted that clause. The third strand is the retreat and return of review in the case law. In A.K. Roy v Union of India, (1982) 1 SCC 271, the Court declined to hold that the satisfaction was justiciable in the ordinary way, and T. Venkata Reddy treated the legislative character of the ordinance as excluding an enquiry into motive. Against that, S.R. Bommai v Union of India, (1994) 3 SCC 1 established for Article 356 the modern position that the satisfaction of the President is reviewable where it rests on no material, on irrelevant material, or is mala fide, while the court does not sit in appeal over the sufficiency of the material.

How did the Forty-fourth Amendment alter the review of the satisfaction?

The Thirty-eighth Amendment had inserted clause (4) into Articles 123 and 213 to make the satisfaction final and unquestionable in court. The Forty-fourth Amendment omitted clause (4) from both Articles. That omission removed the textual bar on review and left the satisfaction examinable under the general law of judicial review.

The majority in Krishna Kumar Singh treated that omission as decisive of the direction of travel: the constitutional text had contained a bar on review, the bar was deliberately removed, and no doctrine of immunity could be rebuilt by interpretation on ground the amending power had cleared. Review of the satisfaction therefore stands where the general law of judicial review of constitutional satisfaction places it, governed by the approach settled in Bommai, and not by any ordinance-specific immunity.

The majority’s reasoning on the point also disposed of the strongest textual argument for immunity, which was the legislative character of the act. An ordinance is indeed legislative in character, and the court does not review a statute for the wisdom or motives of the legislature that enacted it. But the promulgation of an ordinance is not only a legislative act; it is a legislative act that the Constitution permits only upon a condition, the existence of circumstances rendering immediate action necessary, attested by the satisfaction of the President or the Governor. Judicial review in this field examines the condition, not the legislation: whether there was a satisfaction at all, and whether it was formed on material that the Constitution allows to count. That distinction, between reviewing the law and reviewing the satisfaction that alone authorised the law to be made by the executive, is the analytical centre of the second holding, and it is what allowed the bench to affirm reviewability without asserting any power to strike down an ordinance for bad policy.

The bench, the opinions and the shape of the decision

The bench of seven judges comprised T.S. Thakur, Chief Justice of India, and Madan B. Lokur, S.A. Bobde, A.K. Goel, U.U. Lalit, D.Y. Chandrachud and L. Nageswara Rao, JJ. Three opinions were delivered. Chandrachud J. wrote the principal judgment, stating the conclusions of the majority on each of the questions. Thakur C.J. wrote a separate concurring opinion, agreeing that the successive re-promulgations before the Court were a fraud on the Constitution and that the appeals must fail, but expressly leaving open, in his own reasoning, the interpretation of Articles 123(2) and 213(2) on the obligation to place an ordinance before the legislature. Lokur J. dissented in part. He agreed that Bhupendra Kumar Bose and T. Venkata Reddy required to be overruled on the enduring rights question, though for reasons of his own, and he held that re-promulgation is not of itself a fraud on the Constitution, since exigencies may require it, though it must never be a mechanical exercise and the Governor must be satisfied of the necessity for immediate action on each occasion. His principal dissent was on laying: in his view the requirement in Article 213(2) is not mandatory, and a failure to lay does not deprive the ordinance of the force of law or render it of no consequence.

The structure matters for anyone citing the case. The propositions that re-promulgation is impermissible, that the satisfaction is reviewable, that survival of consequences is a matter of construction governed by public interest and constitutional necessity, and that Bhupendra Kumar Bose and T. Venkata Reddy do not state the correct law, are holdings of the Court. The proposition that laying is mandatory is the holding of the majority speaking through Chandrachud J., reached over the dissent of Lokur J. and with the Chief Justice reserving his position in a separate opinion; it remains the law declared by the bench, and it is so treated in later practice. The disposition of the appeals followed from the re-promulgation holding: the ordinances in the Sanskrit schools series were unconstitutional products of a practice the Constitution forbids, the Patna High Court had been right to dismiss the writ petitions, and the teachers’ claim to be treated as State employees on the strength of those ordinances could not be sustained. The Court thus denied relief to the very claimants whose case had exposed the practice, because their entitlement depended on the validity of instruments that the practice had vitiated. That hard outcome is part of the judgment’s logic about risk, and it is examined after the holdings.

The first holding: re-promulgation is impermissible

The majority reaffirmed Wadhwa and put its conclusion in the same words, that re-promulgation of ordinances is a fraud on the Constitution and a subversion of democratic legislative processes. The reasoning proceeds from the character of the power. The power conferred by Articles 123 and 213 is legislative in character, but it is conditional: it may be exercised only when the legislature is not in session, and only upon the satisfaction that circumstances exist which render immediate action necessary. The President and the Governor are not constituted by these Articles into a parallel source of law-making or an independent legislative authority. Promulgation is an exception carved out of the legislative supremacy of Parliament and the State Legislatures, justified by necessity and bounded by the conditions that create it. A practice that converts the exception into the ordinary route of legislation contradicts the premise on which the power exists.

Re-promulgation fails that test in a specific way. The constitutional design gives an ordinance a short and defined life, at the end of which the legislature must have the opportunity to approve it, disapprove it, or replace it with an Act. Re-promulgation is the device by which that opportunity is withheld while the law continues. Each successive promulgation is formally a fresh exercise of power, but where the series is maintained without laying the measure before the legislature, the series is in substance a single continuing act of executive legislation. The majority treated the Sanskrit schools series in exactly that way: seven successive re-promulgations of the first ordinance of 1989, none of them laid before the Assembly, constituted a fraud on the Constitution, and the ordinances produced by the practice were unconstitutional. Thakur C.J., in his concurring opinion, made the same point by describing the promulgations and re-promulgations as a single transaction and a single series on the same subject, so that the vice of invalidity attached to the whole series and not only to its later members.

Two qualifications keep the holding precise. The first is that the vice lies in the practice, not in the bare fact that an ordinance follows an earlier one on the same subject. Lokur J.’s dissent insisted that re-promulgation is not of itself a fraud, because an exigency may persist or recur, and the majority’s rule is best read with that pressure in mind: a single fresh promulgation upon a genuine and fresh satisfaction, in circumstances the executive can explain, stands on different ground from a mechanical series maintained to avoid the legislature. What the Constitution forbids is re-promulgation as a substitute for the legislative process, where the executive, in the majority’s words, persistently avoids placing the ordinance before the legislature. The second qualification concerns remedy. A finding that a series of re-promulgations is unconstitutional does not answer, by itself, the question of what happens to acts done under the ordinances in the series. That question belongs to the fourth holding, and the separation of the two questions is deliberate: invalidity of the practice is one matter, the survival of its consequences another, and the Court answered them under different tests.

The second holding: the satisfaction is reviewable

The majority held that the satisfaction of the President under Article 123 and of the Governor under Article 213 is not immune from judicial review, and stated the grounds and the limits of review together. The court will not determine the sufficiency or adequacy of the material on which the satisfaction rests; the executive’s assessment of whether circumstances require immediate action is not re-weighed. What the court may scrutinise is whether the satisfaction was formed at all and whether it was formed lawfully. Judicial review, the majority held, would enquire into whether there was no satisfaction at all, whether the satisfaction constitutes a fraud on power, and whether it was actuated by an oblique motive; the test is whether the satisfaction is based on some relevant material, so a satisfaction resting on extraneous or irrelevant considerations, or on no material, is open to challenge. The deletion of clause (4) from both Articles by the Forty-fourth Amendment was treated as removing the textual foundation of immunity, and the approach of S.R. Bommai to the review of a constitutional satisfaction supplied the governing model.

The holding displaced the contrary current in T. Venkata Reddy, under which the legislative character of an ordinance had been taken to exclude an enquiry into the satisfaction on grounds of mala fides. The majority’s distinction, between the law made and the condition for making it, answers that current directly. A statute enacted by a legislature is not reviewed for motive because the Constitution commits law-making to the legislature unconditionally. An ordinance is law-making committed to the executive conditionally, and the condition is justiciable because it is the only thing that separates an ordinance from legislation by the wrong institution. The practitioner consequence is concrete. A challenge to an ordinance may plead that no satisfaction existed, that the satisfaction was a fraud on the power, or that it rested on extraneous considerations; the pleadings must engage the existence and the material of the satisfaction, not its wisdom; and the government’s answer is to demonstrate the satisfaction and its relevant material, not to assert immunity. Challenges framed as invitations to assess whether the emergency was grave enough will fail at the threshold the majority itself set, since adequacy is expressly outside the enquiry.

There is a further, structural consequence. Because the satisfaction is reviewable, the record of satisfaction matters. The advice tendered to the President or the Governor, the material placed before the constitutional head, and the sequence of sessions and prorogations surrounding a promulgation all become potentially examinable material in litigation. An executive that promulgates as routine must expect the routine to be reconstructed in court. The holding therefore operates in advance of any particular challenge: it raises the internal standard of care for the promulgation file, in the same way that the laying holding raises the standard of care for the parliamentary file, and that indirect discipline is a substantial part of what the judgment was designed to achieve.

The third holding: laying is mandatory

The majority held that the requirement of laying an ordinance before Parliament or the State Legislature is a mandatory constitutional obligation cast upon the government, and that failure to comply with it is a serious constitutional infraction and an abuse of the constitutional process. The reasoning is functional. Laying is the mechanism by which the legislature exercises control over a legislative act of the executive, and the control has three distinct objects. Upon the ordinance being laid, the legislature must be able to determine the need for, the validity of and the expediency of the promulgation; to decide whether the ordinance ought to be approved or disapproved; and to decide whether an Act incorporating the provisions of the ordinance should be enacted, with or without amendments. Those are the same decisions the legislature makes in the ordinary course when it enacts law through the process described in how a Bill becomes law in India, and the ordinance clauses compress them into the laying procedure because the ordinance has already taken effect. If laying were directory, each of those decisions could be defeated by the simple expedient of never placing the instrument on the table, and the disapproval resolution and the six-week limit would operate in a vacuum. The mandatory character of laying is what makes the rest of the scheme real.

The Sanskrit schools series was the limiting case of the failure the holding addresses. Not one of the ordinances in the series was laid before the Bihar Legislative Assembly at any point in the years across which they were successively re-promulgated. The executive thereby obtained the substance of legislation while denying the Assembly each of the three determinations the Constitution reserves to it. The majority’s language of constitutional infraction attaches to that denial, and it gives the laying requirement an independent legal significance it had not previously carried: in Wadhwa the wrong was re-promulgation; in Krishna Kumar Singh the failure to lay is itself a constitutional wrong, whether or not the ordinance is later re-promulgated. A government that promulgates an ordinance, allows it to operate, and omits to lay it has committed an infraction even if the ordinance is never used again.

Lokur J.’s dissent on this point is the strongest counter-statement and is addressed in its own terms in the treatment of the dissent. Its core is textual and structural: the fate of an ordinance, in his view, is governed entirely by Article 213(2), under which it ceases to operate at the stated time or upon disapproval, and the only control the executive retains after promulgation is withdrawal; the legislature’s control is exercised through that cessation and disapproval machinery whether or not the instrument is formally laid. The majority’s answer, in substance, is that cessation and disapproval cannot operate on an instrument the legislature has not been given, and that a control which depends on the controlled party delivering the instrument is no control at all. The disagreement is genuine and narrow, and it is about where the Constitution locates the legislature’s grip on executive legislation: in the clock, as the dissent would have it, or in the laying, as the majority held.

The fourth holding: no enduring rights by default

The fourth holding is the one that changed the law. The majority held that the theory of enduring rights laid down in Bhupendra Kumar Bose and followed in T. Venkata Reddy does not state the correct position, and that those decisions are no longer good law, being also inconsistent with the decision in S.R. Bommai. The question whether rights, privileges, obligations and liabilities survive an ordinance that has ceased to operate must be determined as a matter of construction. The appropriate test is the test of public interest and constitutional necessity, which includes the issue whether the consequences that have taken place under the ordinance have assumed an irreversible character. In a suitable case it is open to the court to mould the relief. An ordinance is law while it lives; on the question of what it leaves behind, there is no presumption in favour of survival.

The reasoning dismantles the temporary Act analogy at three points. The first is categorical: there is a basic difference between an ordinance and a temporary enactment. A temporary Act is enacted by the legislature itself for a stated duration, and the law of temporary enactments, under which completed effects survive expiry, is a rule about the construction of a legislature’s own product. An ordinance is the executive’s product, issued under a conditional power, and importing the temporary Act presumption into it gives the executive’s instrument a durability the Constitution did not confer. The second point is textual. The framers used distinct expressions across the Constitution, including “repeal”, “void”, “cease to have effect” and “cease to operate”, and each carries a distinct connotation. “Cease to operate” in Articles 123 and 213 does not mean that the ordinance is rendered void from the beginning: the Articles separately prescribe the situations in which an ordinance is void, namely where it makes a provision the legislature would not be competent to enact. Cessation is therefore neither annihilation nor continuation; it is the end of operation, with the fate of accrued consequences left to be determined. The third point is the missing saving clause. Other constitutional provisions that create time-limited law-making, including Articles 249(3), 250(2) and 357(2), expressly save things done under the law that expires. Articles 123 and 213 contain no such saving, and while the majority held that absence is not conclusive by itself, it confirms that survival cannot be a default; it must be established as a matter of construction in each case.

What does “cease to operate” mean in Articles 123 and 213?

The expression marks the end of an ordinance’s operation without erasing its past. An ordinance that ceases to operate is not void from the beginning; the Articles reserve voidness for an ordinance beyond the legislature’s competence. Nor is cessation a repeal with a saving of accrued rights, since Articles 123 and 213 contain no saving provision.

The consequence, after Krishna Kumar Singh, is that acts done while the ordinance operated are neither automatically preserved nor automatically undone. Their fate is a question of construction for the court, answered by the test of public interest and constitutional necessity, with irreversibility at its centre.

The practical content of the test can be stated through the two directions in which it cuts. In one direction, consequences that have assumed an irreversible character, or whose unwinding would be impracticable and damaging to public interest, may be preserved by the court even though the ordinance that produced them has ceased to operate, and the power to mould relief allows preservation to be partial and tailored. In the other direction, an entitlement that rests entirely on the ordinance, and whose continuation would perpetuate the very governance the ordinance power does not authorise, will not survive merely because it was created while the ordinance lived. The Sanskrit school teachers stood in the second direction: their claimed status depended on the continued efficacy of ordinances held to be unconstitutional, and no consideration of public interest or irreversibility carried their claim across the line. Between the directions lies a field of cases the judgment deliberately leaves to decision case by case, and the uncertainty of that field is the price of the rule, weighed in the counter-reading taken up after the dissent.

The position before and after, in one view

The four holdings, set against the doctrine they replaced, show that three of them hardened existing law and one of them reversed it. The table states the position on each question before the judgment of 2 January 2017 and after it, with the authority for each entry.

Question	Before Krishna Kumar Singh	After Krishna Kumar Singh
Is re-promulgation permissible?	Condemned as a fraud on the Constitution in D.C. Wadhwa (1987), but without attached consequences, and the practice continued.	Impermissible: a fraud on the Constitution and a subversion of democratic legislative processes; a series maintained without laying is unconstitutional (Krishna Kumar Singh, majority).
Is the satisfaction reviewable?	Treated as effectively beyond enquiry in T. Venkata Reddy (1985), on the legislative character of an ordinance.	Reviewable on limited grounds: no satisfaction at all, fraud on power, oblique motive, or material that is extraneous or irrelevant; adequacy is not examined (Krishna Kumar Singh, applying the Bommai approach).
Is laying mandatory?	Required by the text of Articles 123(2) and 213(2), but its legal character was undecided and failure carried no stated consequence.	Mandatory: a constitutional obligation of the government, whose breach is a serious constitutional infraction and an abuse of the constitutional process (Krishna Kumar Singh, majority; Lokur J. dissenting).
Do rights survive expiry?	Yes, by default: Bhupendra Kumar Bose (1962) and T. Venkata Reddy (1985) treated an ordinance on the analogy of a temporary Act whose effects endure.	No default: survival is a matter of construction under the test of public interest and constitutional necessity, including irreversibility; the earlier decisions are overruled (Krishna Kumar Singh).

The last row is the row that matters in practice. The first three rows regulate the conduct of governments, and a government determined to test them can still calculate the political price of a challenge. The fourth row regulates the expectations of everyone who deals with an ordinance, and it operates without any litigation against the government at all: it operates at the moment a citizen, an employee, a contractor or a department decides how much weight to place on an entitlement that an ordinance has created.

The dissent, in its own terms

What did Lokur J. hold in dissent, and where did he agree?

Lokur J. agreed that an ordinance cannot create an enduring or irreversible right, and that Bhupendra Kumar Bose and T. Venkata Reddy required to be overruled, though for his own reasons. He dissented on laying, which he held not to be mandatory, and on re-promulgation, which he held not to be a fraud in itself.

On re-promulgation his discipline was a genuine satisfaction on each occasion, never a mechanical exercise, since exigencies may require an ordinance to be promulgated again. On laying, the fate of an ordinance was in his view governed entirely by Article 213(2) and by the Legislative Assembly, and failure to lay neither strips the ordinance of the force of law nor renders it of no consequence.

The dissent deserves to be taken seriously rather than treated as a footnote to the majority, for two reasons. The first is that its textual discipline is real. Article 213(2) prescribes consequences with some precision: the ordinance ceases to operate at the expiration of six weeks from reassembly, or earlier upon disapproval, and it may be withdrawn. It does not prescribe a consequence for failure to lay. A court that attaches the consequence of constitutional infraction to an omission whose consequence the text does not state is making constitutional law, and Lokur J.’s opinion is a reminder that the majority was doing so, on the strength of the function laying performs rather than any sanction the Article names. The second reason is that the dissent’s account of re-promulgation identifies the genuinely hard case. Emergencies do not always expire on the legislature’s timetable, and a rule that treats every second promulgation as fraudulent would punish the executive for the legislature’s calendar. The majority’s holding, fairly read, does not adopt that rule: its target is the series maintained to avoid the legislature, and its emphasis on persistent avoidance of laying is what separates the condemned practice from a justified repetition. But the line between the two is drawn by emphasis rather than by a stated test, and litigants since 2017 have had to argue where a given series falls. On the enduring rights question, by contrast, there was no dissent at all in substance: seven judges agreed that the old automatic-survival doctrine could not stand, and that agreement across the bench’s divisions is the surest measure of where the judgment’s centre of gravity lies.

The case for the fourth holding, and the case against it

The stance of this series on a contested question is to give each side its strongest statement before taking a position, and the enduring rights holding needs both. The case for the holding begins with the structure of the ordinance power. An ordinance is legitimate, if at all, because it is temporary and supervised: temporary because the legislature is expected to replace or reject it, supervised because it must be laid. A doctrine under which its consequences are permanent breaks both legs of that justification. Permanence of effect combined with brevity of life is the precise formula for legislation without a legislature: promulgate, act quickly, create facts on the ground, and let the ordinance die in the knowledge that its work survives it. The presumption of enduring effect was, on this view, not a neutral rule of construction but the incentive structure of ordinance raj. Wadhwa’s record showed governments exploiting duration; the enduring rights doctrine showed them how to exploit expiry as well. Removing the presumption removes the prize. The holding also has a textual honesty to it: the Constitution’s failure to save rights under a lapsed ordinance, in a document that saves rights expressly in neighbouring contexts, is better respected by a construction test than by an analogy imported from the law of temporary statutes, where the enacting body and the expiring body are the same legislature.

The case against the holding is a case about reliance, and it is not answered by pointing to the executive’s misconduct, because the person who bears the loss is usually not the executive. An ordinance in force is law. A citizen who takes an appointment under it, accepts a payment under it, pays a levy under it, enrols in a scheme under it or alters a legal position in reliance on it has done exactly what the legal system asks, which is to obey the law in force at the time. When the ordinance lapses and the entitlement is then subjected to a case by case test of public interest and constitutional necessity, the citizen’s position depends on a judicial assessment conducted after the event, under criteria, public interest and irreversibility, that are matters of judgment rather than rule. The old law, whatever its structural vice, at least told everyone where they stood: effects endured. The new law tells them that effects endure if a court, later, considers that they should. For the Sanskrit school teachers the point is sharpest. Their wrong, if any, was to have been employed in schools the State took over; the constitutional fraud was the government’s; the loss fell on them. A doctrine that allocates the cost of executive misconduct to the citizen who relied on the executive’s law is open to the charge of punishing the wrong party, and scholars of constitutional remedies have pressed exactly that charge against the remedilessness of the outcome even while accepting the doctrine.

The Court’s answer to the reliance objection is built into the test it chose, and it should be recorded fairly. The test does not direct that consequences fall with the ordinance. It directs the court to ask whether the consequences have assumed an irreversible character and whether unwinding them would be impracticable, and it arms the court with the power to mould relief, which is the power to preserve consequences in whole or in part where justice and public interest require it. The reliance interest is therefore not excluded from the enquiry; it is the enquiry’s central subject. What the holding withdraws is only the automaticity. The verdict of this article is that the exchange was justified, for a reason the reliance argument does not reach. A presumption of survival is not merely a protection for the relying citizen; it is also, and at the same time, a guarantee to the promulgating executive that its ordinance will achieve permanent effects at temporary cost, and a legal system cannot hold both ends of that proposition without subsidising the practice it condemns. The case by case test protects reliance where reliance is genuine and irreversible, and denies protection where the claimed consequence is simply the continuation of executive legislation by other means. The cost is predictability, and the cost is real; but predictability in favour of survival was the subsidy, and ending the subsidy was the point of the judgment. What courts owe the relying citizen, within the test, is a disciplined and generous reading of irreversibility, so that appointments acted upon, payments received and positions altered in good faith are unwound only where a constitutional necessity genuinely outweighs them.

The judgment as a reallocation of risk

Read as a whole, Krishna Kumar Singh is less a set of prohibitions than a repricing of ordinance government. Wadhwa had addressed the executive in the language of constitutional morality, and the executive had absorbed the language and continued the practice. The 2017 judgment addresses the executive in the language of consequences, and it distributes those consequences across every participant in an ordinance’s life. For the government, re-promulgation is unconstitutional and a failure to lay is an infraction, so the parliamentary file must be perfect. For the constitutional head and the advising ministry, the satisfaction is reviewable on the record, so the promulgation file must be real. For the beneficiary of an ordinance, and this is the fourth holding’s work, the entitlement is provisional in a way an entitlement under an Act is not, because its survival past the ordinance’s death must be affirmatively established rather than presumed. Risk that previously rested on the legislature, which had to find the time and the numbers to disapprove or replace an ordinance before its effects hardened, is transferred to the executive that promulgates and to those who act on the promulgation. Series of ordinances maintained to avoid the legislature are the extreme case, but the same logic reaches the single ordinance carelessly laid or optimistically relied upon.

Subsequent practice in adjacent fields shows the same contest continuing under the judgment’s shadow. The Tribunals Reforms sequence, in which provisions on tribunal appointments that the Court had struck down in the Madras Bar Association litigation returned by ordinance and then by re-enactment, presented the cognate question whether the executive may reproduce by legislation the substance of what judicial review has removed, and the Court’s impatience in that line of cases draws on the same premise as Krishna Kumar Singh, that the forms of legislative power may not be used to defeat the substance of constitutional control. The land acquisition amendments of 2014 and 2015 supply the complementary example from within ordinance practice itself: the amendments to the land acquisition statute were carried by an ordinance promulgated at the end of December 2014 and then by successive ordinances in April 2015 and again in May 2015, before the government of the day allowed the measure to lapse in August 2015 rather than continue the series. That episode, occurring while Krishna Kumar Singh was pending, is the pattern Wadhwa condemned, run at the Union level and abandoned under political and legal pressure; under the doctrine settled in 2017, the consequences claimed under such a series would face the further difficulty that none of them would carry a presumption of survival.

What happens to an appointment made under an ordinance that later lapses?

An appointment made under an ordinance is made under law in force, and the later cessation of the ordinance does not invalidate it, because cessation is not voidness from the beginning. Its continuation after the lapse is another question: no presumption of endurance assists the appointee, and survival must be established under the construction test.

The court asks whether the consequence has assumed an irreversible character and whether public interest and constitutional necessity support its survival, and it may mould relief accordingly. An appointment fully acted upon, with service rendered and third-party arrangements built on it, presents the strongest case for survival; an appointment that merely continues an arrangement the ordinance created, where the ordinance itself was part of an unconstitutional series, presents the weakest.

The documentary dimension of the doctrine should not be missed, because the whole line of cases is built on records. Wadhwa’s tables came from gazettes. The satisfaction enquiry examines the promulgation file. The laying enquiry examines the records of the legislature. Practitioners who litigate ordinance questions, and researchers and journalists who reconstruct ordinance practice from public records, work with material whose preservation decides what can be proved; the discipline described in the VaultBook account of how journalists protect sources and their working papers belongs to the same family of practices as the scholar’s archive that made Wadhwa possible. A doctrine that polices executive legislation through records rewards those who keep them.

The text the seven judges construed

Doctrine of this weight is best kept tied to the clauses that produced it. Articles 123 and 213 are structured in parallel, the first for the President and Parliament, the second for the Governor and the State Legislature, and the provisions on which Krishna Kumar Singh turns are four. The opening clause confers the power conditionally: if at any time, except when both Houses are in session, the President is satisfied that circumstances exist which render it necessary for him to take immediate action, he may promulgate such ordinances as the circumstances appear to him to require, with Article 213 adding, for Governors, a proviso about instructions from the President in specified cases. The force and effect clause gives a promulgated ordinance the same force and effect as an Act. Clause (2) then carries the whole scheme of control. An ordinance so promulgated shall be laid before both Houses, and shall cease to operate at the expiration of six weeks from the reassembly of the legislature, or, if before the expiration of that period resolutions disapproving it are passed, upon the passing of the second of those resolutions; and it may be withdrawn at any time. Clause (3) supplies the only ground on which an ordinance is declared void: if and so far as an ordinance makes any provision which the legislature would not be competent to enact, it shall be void.

Each of the four holdings is a reading of one of those elements. The re-promulgation holding reads the opening clause: a power exercisable only upon satisfaction of immediate necessity cannot be exercised in a series whose purpose is to dispense with the legislature, because the necessity asserted on the seventh promulgation is not the necessity the clause contemplates. The satisfaction holding reads the same clause from the other side: satisfaction is the condition of the power, and conditions in a constitutional grant are examinable. The laying holding reads the first limb of clause (2), and treats the imperative “shall be laid” as the hinge of legislative control rather than a direction about parliamentary housekeeping. The enduring rights holding reads the second limb of clause (2) together with clause (3): the framers provided that the ordinance shall cease to operate, and separately provided when it shall be void, and the space between those two expressions is where the fate of accrued consequences must be determined by construction. Students who map the judgment onto the text in this way will find that nothing in it floats free of the clauses, which is part of its authority; the majority’s method was textual first, structural second, and consequential third.

One more textual observation explains the choice of battleground. The Constitution (Thirty-eighth Amendment) and the Forty-fourth Amendment had fought over clause (4), the review bar, a generation earlier, and the Forty-fourth Amendment’s deletion of that clause meant that by 2017 the text no longer contained anything that even purported to immunise the satisfaction. The defenders of the old position in T. Venkata Reddy were therefore driven to rest immunity on the legislative character of the ordinance rather than on any words in the Article, and character arguments of that kind are vulnerable to the distinction the majority drew between the act of legislating and the condition that authorises the executive to legislate. The text, as amended by 1978, was on the majority’s side before the argument began, and the judgment is in part the delayed working out of what the Forty-fourth Amendment had decided.

Legislative supremacy as the judgment’s premise

Beneath the four holdings sits a single premise which the majority stated expressly: the ordinance making power does not constitute the President or the Governor into a parallel source of law-making, and promulgation is subject to the principle of legislative supremacy. The premise has a democratic component and a mechanical component. The democratic component is that the President and the Governor act on the aid and advice of a Council of Ministers collectively responsible to the legislature, so executive legislation is legislation by a body that the legislature can remove, issued in a gap in the legislature’s own sittings, and answerable to the legislature when it reassembles. The mechanical component is the set of devices by which that answerability is enforced: the laying requirement, the disapproval resolution, the six-week cessation, and the expectation of replacement by an Act where the measure is to continue. Krishna Kumar Singh is best read as a judgment that takes the mechanical component seriously as law, rather than as convention or as drafting surplus. Before 2017 it was possible to describe the laying requirement as a formality and the six-week rule as a self-executing limit that needed no assistance. After 2017 the devices are holdings: laying is mandatory and its omission an infraction, cessation carries no implied saving, and the satisfaction that starts the whole machine is examinable.

The premise also explains the judgment’s treatment of the constitutional fiction in the force and effect clause. The majority observed that the fiction attributing to an ordinance the same force and effect as a law enacted by the legislature comes into being if the ordinance has been validly promulgated and complies with the requirements of Articles 123 and 213. Force and effect are therefore the reward of compliance, not a cloak for non-compliance. That sentence does a great deal of quiet work in later litigation, because it locates the consequences of a defective promulgation inside the promulgation itself: an ordinance produced by a fraudulent series does not borrow the dignity of an Act for the purpose of saving what was done under it. Read together with the fourth holding, the position is coherent. Validity governs whether the ordinance ever had the force of law in the full sense; construction governs what survives its cessation; and neither enquiry is answered by the bare fact that the instrument was called an ordinance and printed in the gazette.

The thirty years between the two judgments

The interval between Wadhwa and Krishna Kumar Singh is itself evidence in the case for the second judgment’s design. Wadhwa was decided in December 1986. The Sanskrit schools ordinance at the heart of Krishna Kumar Singh was first promulgated in 1989, and its seven re-promulgations ran into the following years, in the very State whose practice Wadhwa had examined, under a declaration by a Constitution Bench that the practice was a fraud on the Constitution. No feature of the second case is more damning of the first judgment’s remedial modesty. The Bihar government of the day cannot have misunderstood Wadhwa; the judgment was famous, its language was unqualified, and its author had been the Chief Justice. What the episode shows is that a constitutional condemnation addressed to the conscience of the executive competes, inside government, with the convenience of the condemned practice, and convenience won for three decades because nothing in the legal position made the practice expensive. The teachers’ litigation then supplied what Wadhwa had lacked: a concrete claim, by identified persons, whose success or failure would force the Court to say what followed from the condemnation.

The interval also changed the legal materials available to the second bench. Bommai, decided in 1994, had normalised the review of a constitutional satisfaction and had done so in the setting of Article 356, where the satisfaction displaces an elected government; the argument that the lesser satisfaction behind an ordinance was uniquely immune became difficult to maintain with a straight face after the greater one had been held reviewable. The jurisprudence of the 1990s and 2000s on colourable legislation, on fraud on power, and on the review of executive action for irrelevant considerations had thickened around the ordinance cases without disturbing them, because Bhupendra Kumar Bose and T. Venkata Reddy stood in the way as Constitution Bench authority. A five-judge bench could distinguish them at the margins; it could not displace them. The constitution of a seven-judge bench in Krishna Kumar Singh was therefore not a ceremonial escalation but the necessary instrument: the questions referred could be answered authoritatively only by a bench stronger than five, and the referring process, in which the conflict between T. Venkata Reddy and the later law was identified, is part of the judgment’s pedigree that students should note.

There is a third feature of the interval. Ordinance practice at the Union level during those decades was heavier than the constitutional design contemplated, though it never approached the Bihar pattern of Wadhwa’s tables: measures of major policy were repeatedly introduced by ordinance, and some were re-promulgated, most visibly where the government’s majority in the second chamber was uncertain. Commentators on the judgment, writing in the constitutional law journals and in the press, treated Krishna Kumar Singh as addressed to that practice as much as to Bihar’s. The holdings are framed generally, for Articles 123 and 213 alike, and their discipline falls on the Union executive in terms. Whether the discipline has been effective is the subject of the practice discussion that follows the risk analysis; the point here is that the 2017 bench wrote for a national practice, on facts supplied by one State, using a record assembled thirty years earlier in the same State. The continuity of the Bihar thread through all three moments, research, first judgment and second judgment, gives the doctrine an unusually complete factual biography.

Fraud on power: the doctrine behind the rhetoric

The phrase fraud on the Constitution, which Wadhwa coined for this field and Krishna Kumar Singh adopted, belongs to a wider family of doctrines by which Indian constitutional law polices the misuse of a power for purposes outside its grant. Its relatives include the doctrine of colourable legislation, under which a legislature may not do indirectly what it cannot do directly; the principle that a power conferred for one purpose and exercised for another is a fraud on the power; and the administrative law rule that a discretion exercised on extraneous considerations, or in bad faith, is no exercise of the discretion at all. The family resemblance matters because it tells the reader what kind of wrong re-promulgation is. The wrong is not that the executive legislated badly, nor that it legislated on a subject reserved to the legislature, since the ordinance power covers the legislature’s whole field. The wrong is one of purpose and of fidelity to the conditions of the grant: a power given to bridge the legislature’s absence is used to engineer the legislature’s irrelevance, and the form of the power is observed precisely in order to defeat its function. That is why the Court’s language moves between fraud, subversion and colourable exercise; each term in the family describes the same structure, the use of a lawful instrument for an object the instrument was not given to achieve.

Understanding the wrong in that way also clarifies the limits of the first holding. A fraud on power is characteristically a conclusion drawn from a pattern and a purpose, not from a single act inspected in isolation. The first promulgation of the Sanskrit schools ordinance was, on its face, an ordinary takeover measure of a kind State governments enact by Act every year. The fraud emerged from the series and from the omission that accompanied it: seven repetitions, no laying, no Bill. Courts asked to apply Krishna Kumar Singh to a shorter series, or to a series punctuated by laying and by a failed Bill, must therefore reason about purpose from evidence, the session calendar, the fate of replacement Bills, the intervals between promulgations, and the government’s own explanations. The judgment gives them the conclusion and the paradigm case; the evidential discipline of the fraud on power family gives them the method. Governments, correspondingly, defend a re-promulgation not by asserting a right to re-promulgate, which the judgment forecloses, but by evidencing the exigency and the legislative engagement that take the series outside the paradigm. That is a healthier forensic structure than the one it replaced, under which the executive asserted the legislative character of the ordinance and the enquiry ended.

How the construction test is argued, and how it is likely to be applied

The fourth holding replaces a presumption with a method, and the method deserves a slower statement than it usually receives. The enquiry begins with the ordinance itself. Because survival is a matter of construction, the first question is what the ordinance, read as a whole, indicates about the fate of things done under it: some ordinances by their nature execute themselves in a single completed act, such as the holding of an election or the making of an acquisition, while others create continuing states of affairs, such as an appointment to a tenure post or the imposition of a recurring levy, whose continuation after cessation requires a legal basis that the dead ordinance can no longer supply. The second question is the character of the consequence. The majority’s test names public interest and constitutional necessity, and includes within it the enquiry whether the consequences have assumed an irreversible character. Consequences arrange themselves along a spectrum. At one end stand consequences that cannot be undone without absurdity or grave public harm: an election conducted and a legislature constituted on its result, a concluded trial, a payment received and spent in the ordinary course, a third party’s rights acquired for value on the faith of the arrangement. At the other end stand consequences whose whole content is the continued enjoyment of a status the ordinance created and nothing else: the teachers’ claim in Krishna Kumar Singh itself is the type case, because the status claimed, employment by the State, had no existence apart from the ordinances and conferred no rights on third parties that unwinding would injure. Between the ends lie the cases that will occupy the courts, and the judgment’s instruction is that they be decided by the stated criteria and by the power to mould relief, not by loyalty to the discarded presumption in either direction.

Two comparisons sharpen the method. The first is with repeal. When a legislature repeals an Act, section 6 of the General Clauses Act, 1897 supplies a default saving of rights acquired and liabilities incurred, unless a different intention appears; the repeal of an enactment is therefore wrapped in a statutory presumption of survival. No corresponding provision governs the cessation of an ordinance, and the majority’s reasoning treats that difference as deliberate constitutional design rather than an oversight to be corrected by analogy. A litigant who argues from section 6 to a lapsed ordinance is arguing for the very analogy the judgment rejected, and the argument should be expected to fail; the governing instrument is the test of public interest and constitutional necessity, applied to the ordinance’s own terms. The second comparison is with the fate of a temporary Act, the analogy at the root of Bhupendra Kumar Bose. Under the law of temporary statutes, the expiry of the Act does not undo proceedings completed under it, a rule the English and Indian courts built for enactments of the legislature itself. Krishna Kumar Singh does not disturb that rule for temporary Acts; it holds only that an ordinance is not a temporary Act for this purpose, because the identity of the maker differs and the conditions of the making differ. The precision of that distinction is a strength of the judgment, since it confines the change to the field where the constitutional objection lies, executive law-making, and leaves the legislature’s own temporary enactments under their settled rule.

Bhupendra Kumar Bose itself is the best illustration of how outcomes may survive a change of doctrine. The facts there, an election invalidated under an ordinance and a fresh election held, present consequences of the most irreversible kind known to the law: a municipal election cannot be unheld, and the affairs of the municipality had been conducted by the body the fresh election produced. A court applying the construction test to those facts could reach the same result the 1962 bench reached, preservation of what was done, while reasoning in the opposite direction, from irreversibility established in the case to survival, instead of from a presumption of survival to the same end. T. Venkata Reddy presents the sterner test of the new doctrine, since the abolition of posts and the removal of incumbents effected by that ordinance were treated as enduring under the old law, and under the new law the incumbents’ position would have to be established through the construction enquiry rather than conceded by it. The difference the judgment makes is therefore not that every consequence once saved is lost; it is that the party asserting survival must prove the features that justify it. In litigation, burdens decide cases in which the evidence is thin, and the reallocation of the burden is the operative change.

Litigating an ordinance after 2017

For the practitioner, Krishna Kumar Singh reorganises ordinance litigation into four enquiries that correspond to the four holdings, and a well-drawn challenge will plead them distinctly because each has its own evidence and its own remedy. The re-promulgation enquiry asks for the series: every promulgation on the subject, its dates against the session calendar, whether any instrument in the series was laid, and whether a replacement Bill was introduced and what became of it. The satisfaction enquiry asks for the condition: whether a satisfaction was formed, and whether the material before the President or the Governor contained considerations extraneous to the necessity of immediate action, remembering that the adequacy of the material is not in issue and that a challenge framed on adequacy invites dismissal. The laying enquiry is the shortest: it asks whether the instrument was laid, and if it was not, the infraction is established by the record of the legislature itself. The survival enquiry arises at the remedial stage and asks the construction questions set out in the preceding section, with the party asserting survival carrying the burden of establishing irreversibility or constitutional necessity.

Interim strategy is affected by the fourth holding in a way practitioners should note. Under the old doctrine, a beneficiary of an ordinance could treat interim protection as the preservation of a presumptively permanent position, and courts could reason likewise. Under the construction test, the beneficiary’s position at the interim stage is explicitly provisional, and the court seised of the main challenge will approach final relief through public interest and irreversibility, in which the interim history, what was done on the faith of interim orders, itself becomes material. Conversely, a challenger who attacks an ordinance late, after arrangements have hardened, must expect the hardening to be argued against complete relief even if the challenge succeeds on the law, because moulding relief is part of the doctrine and not an afterthought. The judgment thus rewards early challenge and honest interim disclosure on both sides, which is consistent with its general design of making ordinance government more expensive to conduct carelessly.

A word belongs here to those who advise governments. The compliance lessons of the judgment are administrative and unglamorous. Lay every ordinance, promptly, and keep the record of laying. Do not re-promulgate without a fresh and documented satisfaction that addresses why the legislature has not replaced the instrument and why immediate action remains necessary. Where continuation is genuinely needed, introduce the replacement Bill and be able to show that it was introduced; the legislative history of the Bill is the government’s best answer to a fraud on power challenge. And where acts of consequence are done under an ordinance, a tax collected, appointments made, a scheme opened, do them in the knowledge that their survival past cessation will be tested, and consider whether the object can wait for an Act. This is general information about the doctrine and not advice on any particular matter; its content is simply the judgment read from the government’s side of the table, and a government that internalises it will seldom need to meet Krishna Kumar Singh in court.

Practice after the judgment

Whether re-promulgation has stopped is a question the brief for this article directs to be answered with attribution rather than assertion, and the attributed answer is that it has not stopped, though its legal environment has changed. Parliamentary records show, for example, that the Muslim Women (Protection of Rights on Marriage) Ordinance, first promulgated in September 2018, was promulgated again in January 2019 and yet again in February 2019 before the measure was enacted in July 2019, a series of three ordinances on the same subject within a single inter-session arc, defended by the government of the day on the ground of continuing necessity and criticised by commentators in the constitutional press as sitting uneasily with Krishna Kumar Singh. Legislative researchers who track ordinance practice, including the legislative research organisations that publish session-by-session counts of ordinances at the Union and in the States, have recorded in their published analyses that resort to ordinances remained substantial in the years after 2017 and that re-promulgation continued to appear in State practice in particular. The honest summary is that the judgment changed the law of re-promulgation completely and its frequency only partly, which is the usual fate of doctrine in this area, and which is itself a reason the fourth holding matters: where the practice persists, the survival rule determines what the practice can achieve.

The deeper change is forensic rather than statistical. Before 2017, a government re-promulgating an ordinance faced, at worst, a declaration in the Wadhwa register and the preservation of everything done under the series. After 2017, the same course of conduct exposes the series to invalidation, the satisfaction to examination, the omission to lay to a finding of constitutional infraction, and the beneficiaries of the series to the loss of their claimed entitlements under the construction test. Litigation risk of that order enters the advice given inside government before the file reaches the Governor, and its deterrent effect, though unmeasurable in published counts, is the mechanism by which the judgment was designed to work. Opposition parties and parliamentary commentators have also taken up the judgment’s vocabulary: debates on ordinances in Parliament and in the press after 2017 are conducted in its terms, fraud on the Constitution, mandatory laying, enduring rights, in a way that the corresponding debates of the Wadhwa decades were not, because Wadhwa supplied rhetoric without a remedial structure and Krishna Kumar Singh supplies both.

The map of authorities, for the student

The line of authority is long, and examinations test its joints, so the sequence is worth stating as a sequence. Bhupendra Kumar Bose in 1962 establishes the temporary Act analogy and the survival of effects. R.C. Cooper in 1970 accepts, in principle, that a promulgation may be challenged for mala fide satisfaction. The Thirty-eighth Amendment in 1975 inserts the bar on questioning satisfaction; the Forty-fourth Amendment in 1978 removes it. A.K. Roy in 1982 declines to make the satisfaction ordinarily justiciable. T. Venkata Reddy in 1985 consolidates the old law: no enquiry into motive, effects endure on the Bose analogy. D.C. Wadhwa, decided in December 1986 and reported in 1987, condemns re-promulgation as a fraud on the Constitution but attaches no consequences and leaves Bose and Venkata Reddy standing. S.R. Bommai in 1994 settles the modern approach to reviewing a constitutional satisfaction in the Article 356 field. Krishna Kumar Singh in 2017, by a bench of seven, reaffirms Wadhwa with consequences, holds the satisfaction reviewable on the Bommai pattern, holds laying mandatory, and overrules the enduring rights doctrine of Bose and Venkata Reddy.

Three errors recur in summaries of this material and are worth naming because the brief for this article flags them. The first is the statement that an ordinance operates as a temporary Act whose effects survive its expiry; that was the law until 2017 and is the opposite of the law the seven judges declared, and any answer that states it without the Krishna Kumar Singh qualification is wrong in the present tense. The second is the statement that the satisfaction of the President or the Governor is not justiciable; that statement confuses the limits of review, which are real and which exclude any enquiry into adequacy, with an immunity that the Forty-fourth Amendment removed and the 2017 bench refused to restore. The third is the omission of the fourth holding altogether, in summaries that report the fraud language and the laying rule and stop there. The omission is understandable, since the fourth holding is the least rhetorical, but it inverts the judgment’s practical order of importance: the holding that governs the position of every person who has acted under a lapsed ordinance, in every field from municipal elections to public employment, is the survival holding, and it is the one the bench of seven was assembled to reconsider.

What the judgment did not decide

A judgment’s authority includes its silences, and four of the silences in Krishna Kumar Singh should be marked so that the doctrine is not overread. The first concerns the single justified re-promulgation. The majority condemned the series maintained to avoid the legislature and described the paradigm in the facts before it; it did not lay down a numerical rule, a maximum count, or a test stated in terms of the number of repetitions, and Lokur J.’s insistence that exigency may require repetition stands in the reports as a warning against converting the holding into a rule that one repetition is always fatal. Cases between the paradigm and the single justified repetition will be decided on the fraud on power method, by evidence of purpose, and are not decided in advance by anything in the 2017 judgment. The second silence concerns the consequence of a failure to lay, beyond the characterisation of the failure as a serious constitutional infraction. Lokur J. would have held that failure carries no consequence for the ordinance’s force; the majority held the obligation mandatory and the failure an infraction, but the precise remedial consequence for an ordinance that was validly promulgated, operated, and was never laid, in a case where no re-promulgation followed, is a question the facts of Krishna Kumar Singh, in which non-laying accompanied a fraudulent series, did not require the majority to isolate, and it should be argued as an open question rather than cited as decided either way.

The third silence concerns the outer boundary of the satisfaction enquiry. The majority specified the grounds of review and excluded adequacy, but the application of the extraneous considerations ground to a promulgation whose file contains both relevant and irrelevant material, or to a satisfaction attacked for the political character of the government’s motive in choosing the ordinance route over a Bill, will require further decision. The distinction between an oblique motive, which the majority makes examinable, and a political motive, which is the ordinary currency of government and cannot by itself vitiate a legislative act, is easy to state and difficult to administer, and it is the point on which future satisfaction challenges will turn. The fourth silence concerns the content of constitutional necessity in the survival test. Public interest and irreversibility are given; constitutional necessity is named in the test and illustrated by nothing in the facts, since the teachers’ case was decided at the threshold of the fraudulent series. Whether constitutional necessity adds a distinct enquiry, for instance where unwinding a consequence would itself breach a constitutional obligation owed to third parties, or whether it is a compendious label for the public interest enquiry, awaits a case that presents it. A commentary that filled these silences with confident answers would misrepresent the judgment; the silences are the agenda of the next decade of ordinance litigation, and the judgment is stronger, not weaker, for leaving them visible.

Conclusion: law while it lives

Krishna Kumar Singh endures in the reports because it completed a piece of constitutional work that had been left half done for thirty years. Wadhwa had identified the wrong and named it; the 2017 bench supplied the legal consequences that turn a name into a rule. Re-promulgation is a fraud on the Constitution, and the series that practises it is unconstitutional. The satisfaction is reviewable, on grounds that examine its existence and its material and stop short of its adequacy. Laying is mandatory, and its omission is an infraction and not a formality. And an ordinance, when it ceases to operate, leaves behind only what a court, construing the instrument under the test of public interest and constitutional necessity, with irreversibility at the centre, decides should survive it. An ordinance is law while it lives and leaves nothing behind by default. The executive that legislates by ordinance, and the citizen who builds on an ordinance, both act after 2017 in the knowledge that the instrument’s effects are on loan, and that knowledge, distributed to exactly the actors whose conduct needed to change, is the judgment’s real remedy.

The road of the appeal

The procedural history of Krishna Kumar Singh explains why the judgment ranges as widely as it does. The teachers and staff of the taken-over Sanskrit schools petitioned the Patna High Court under Article 226 for the salary and status of State employees. The State’s answer exposed the difficulty at the centre of the case: the petitioners’ status depended on a series of ordinances, and the series, on the authority of Wadhwa, was constitutionally indefensible. The High Court dismissed the petitions, holding the successive re-promulgations unconstitutional. The petitioners appealed to the Supreme Court, and in the Supreme Court the argument opened outward. For the appellants, T. Venkata Reddy offered two supports: that the satisfaction behind an ordinance was not open to the kind of challenge the High Court’s reasoning implied, and that, in any event, rights created under an ordinance endured after its cessation, so that the status conferred on the teachers survived the ordinances that conferred it. For the State, and for the doctrine of Wadhwa, both supports had to fail if the High Court was to be affirmed on principle rather than on the narrow ground of the series alone. The bench seised of the appeals recognised that the second support rested on two Constitution Bench decisions, Bhupendra Kumar Bose and T. Venkata Reddy, whose reconsideration required a bench of seven, and that the first support sat in tension with the Forty-fourth Amendment’s deletion of the review bar and with Bommai. The reference that followed framed the questions the seven judges answered, and the appeals were disposed of in accordance with the answers: dismissed, because the ordinances were unconstitutional and because no enduring right carried the petitioners’ claim past that conclusion.

The history carries a lesson about how constitutional doctrine in this field is made. The governing precedents were not reconsidered in a reference in the abstract; they were reconsidered because identified employees, in an unglamorous service dispute about salary in Sanskrit schools, needed them reconsidered in order to win. The grandeur of the eventual judgment, seven judges, four holdings, two Constitution Benches overruled, sits on a foundation of pay slips that were never issued. That is a common pattern in Indian constitutional law, and it has a consequence for how the judgment should be read: its propositions were framed to decide a dispute about reliance on ordinances, and their authority is at its strongest exactly there, in the reliance setting, and correspondingly more argumentative as they are carried further from it. The Government of Bihar, whose conduct was condemned across three decades and two judgments, was in formal terms the successful respondent; the State defended the dismissal of the claims by relying on the unconstitutionality of its own predecessor governments’ ordinances. Nobody in the case occupied a morally simple position, and the Court’s willingness to let the loss fall where the doctrine placed it, rather than to rescue the petitioners at the doctrine’s expense, is the clearest sign that the bench intended a rule and not an episode.

The design the framers defended

The ordinance power was contested at its birth, and the Constituent Assembly’s debate supplies the premises from which both Wadhwa and Krishna Kumar Singh reason. Members who opposed the draft provisions corresponding to Articles 123 and 213 warned, in terms, that an executive power to legislate during recess would be used to bypass the legislature, and amendments were moved to restrict the power and to shorten the life of ordinances. Dr B.R. Ambedkar, defending the draft, justified the power as an answer to emergency: situations could arise in which the legislature was not in session and immediate action was required, and the power was hedged by the session requirement, by the short cessation period, and by the legislature’s power of disapproval. The defence is significant because it concedes the objectors’ premise while denying their conclusion. The power is not defended in the Assembly as an alternative legislative channel; it is defended as an emergency bridge, tolerable because it is temporary and because the legislature resumes control on reassembling. Every later doctrine in this line is a gloss on that exchange. Wadhwa’s insistence that the power is exceptional and not a parallel source of legislation is Ambedkar’s defence converted into a rule of construction. Krishna Kumar Singh’s holdings go one step further along the same line: the safeguards on which the defence rested, laying, disapproval, cessation, are treated as legal obligations with legal consequences, because the framers’ bargain was that the power would be acceptable only if the safeguards were real.

The historical point also disciplines the counter-reading of the judgment. Critics of Krishna Kumar Singh sometimes present it as a judicial invention imposed on a neutral text. The Assembly record makes that presentation difficult to sustain. The text was carried on an assurance, that the power would remain exceptional and supervised, and the judgment enforces the assurance. Where the judgment is genuinely creative is narrower: in attaching the construction test to cessation, and in characterising non-laying as an infraction, the Court supplied content that the debates gesture towards without stating. Those are the points on which Lokur J.’s dissent fastens, and the debate between majority and dissent is therefore best understood as a debate about how far a court may go in completing a constitutional bargain whose terms are clear in purpose and spare in machinery. The majority completed it boldly; the dissent would have left the completion to the text’s own devices; and the division between those instincts, rather than any disagreement about the evils of ordinance raj, is the judgment’s deepest fault line.

Laying in practice

The laying holding rewards a closer look at what legislatures do with a laid ordinance, because the three determinations the majority enumerated correspond to three distinct parliamentary acts, each with its own politics. The first, examining the need for and expediency of the promulgation, is retrospective scrutiny of the executive’s emergency claim; it is the point at which a government may be required to explain, on the floor, why the matter could not wait, and the record of that explanation is the material on which a later satisfaction challenge may draw. The second, approval or disapproval, is exercised through the statutory resolution: a member may move a resolution disapproving the ordinance, and the debate on such a resolution, usually paired with the debate on the replacement Bill, is the legislature’s substantive vote on the policy. The third, enactment with or without amendments, converts the ordinance into an Act and is the normal terminus of a well-conducted ordinance: the great majority of ordinances that are laid in due course travel this route, and their legislative history is unremarkable precisely because the machinery worked.

Against that normal course, the modes of failure the judgment addresses come into focus. An ordinance may be laid and then abandoned, its replacement Bill never pressed, so that it dies by the clock while its subject matter is left in suspense; the judgment does not condemn that course, because the legislature had its opportunity, and the death of the measure is the system working. An ordinance may be laid in one House and neglected in the other, in a bicameral legislature; the text requires laying before both Houses, and the infraction analysis applies to the omission, though its political visibility is lower. And an ordinance may simply never be laid at all, the Sanskrit schools pattern, in which the legislature is denied all three determinations and the executive’s only remaining discipline is the cessation clock, which re-promulgation then defeats. The gradation matters for litigation and for parliamentary practice alike. After Krishna Kumar Singh, the laying record is the first document a challenger calls for and the first document a defending government must be able to produce, and presiding officers and legislative secretariats, whose tables record what was laid and when, have become custodians of a constitutional obligation that before 2017 was treated as a clerical step. The judgment’s least dramatic holding is therefore, in the daily working of legislatures, the one with the most continuous application.

Temporary law, and why an ordinance is not a species of it

The doctrinal error the fourth holding corrected is worth isolating, because it was an error of classification with a respectable pedigree. Legislatures frequently enact temporary statutes: taxing measures for a year, emergency legislation for a crisis, experimental schemes with sunset clauses. A body of construction law grew around such enactments, and its central rule, that rights acquired and proceedings completed under a temporary Act are not defeated by its expiry, served the sensible purpose of preventing legislative time limits from becoming traps. Bhupendra Kumar Bose placed the ordinance inside that classification, and the placement looked natural: an ordinance, like a temporary Act, is law with a short stated life. The seven judges’ correction begins by noticing what the classification suppressed. A temporary Act is temporary because the legislature, in the exercise of its plenary power, chose a duration; its effects endure, where they endure, as an inference about the intention of the body that made it and that could have made it permanent. An ordinance is temporary because the Constitution imposes a duration on a maker that is not the legislature at all, and the inference of intended permanence is not available: the executive cannot intend, on the legislature’s behalf, that its stopgap measure shall have the enduring consequences of an enactment. Classification by lifespan confused a limit imposed on the power with a choice made in exercising it, and the confusion was the whole doctrine.

Sunset legislation in the modern practice throws the distinction into relief. When Parliament enacts a measure with a sunset clause, it deliberates on the duration as part of the policy, committees examine it, and the expiry date is itself a legislative act that a later Parliament can extend. When the executive promulgates an ordinance, no deliberative body has chosen anything; the duration is a constitutional default, and the deliberation is scheduled for later, at the laying and replacement stage. Krishna Kumar Singh can be read as restoring that scheduling to legal significance. The survival of an ordinance’s consequences is made to depend on a court’s construction precisely because no legislature has performed the deliberative act from which survival could be inferred, and where a legislature does perform it, by replacing the ordinance with an Act that saves accrued rights, the saving is effective because it is legislative. Governments that want durability for what an ordinance does therefore have a lawful route to it, through the legislature, and the judgment’s severity falls only on the attempt to obtain durability without taking that route. Doctrinally, that is the cleanest statement of the case: the fourth holding is not hostile to enduring consequences; it is hostile to enduring consequences that no legislature ever authorised.

The doctrine in the courts below

In the High Courts, where most ordinance litigation is actually conducted, Krishna Kumar Singh operates as a checklist discipline on service matters, municipal matters and regulatory disputes in which an ordinance forms part of the legal history. The recurring pattern is a claim founded on an instrument of the ordinance family: an appointment regularised by an ordinance that was later replaced or allowed to lapse, a municipal body’s composition altered by an ordinance, a fee or levy imposed by an ordinance and demanded after its cessation. Before 2017 such claims were commonly argued, and sometimes decided, on the Bose presumption, with the survival of the claimed right treated as the natural consequence of its creation. After 2017 the argument must be run through the construction test, and judgments of the High Courts in ordinance-adjacent disputes since that year show the characteristic shape of the new enquiry: identification of the precise consequence claimed, examination of the ordinance’s terms for indications of intended duration, and assessment of irreversibility and public interest, with relief moulded where a consequence is partly preserved and partly denied. The change is visible less in outcomes than in the structure of reasoning, which is usually how a Supreme Court doctrine announces that it has taken hold.

The judgment also disciplines the State’s own conduct of such litigation. A government defending a claim founded on a lapsed ordinance can no longer rest on the ordinance’s former validity, and a government resisting such a claim can no longer assume the claim fell with the instrument; both sides must address the construction enquiry, and the record of laying, of any replacement Act, and of what was done in reliance becomes the evidential centre of the case. Litigants have learned to build that record deliberately. The promulgation file, the laying record, the gazette series and the legislative history of any successor Bill are standard requisitions in ordinance litigation, and their absence is itself argued as telling. In that forensic sense the judgment achieved something Wadhwa never attempted: it made the paper trail of ordinance government into admissible constitutional evidence, and gave every future challenger a discovery agenda written by the Supreme Court itself.

The satisfaction record: what counts as material

Since the second holding makes the satisfaction examinable for the relevance of its material, the content of a lawful satisfaction deserves statement. The material must bear on the two elements the opening clause of Articles 123 and 213 requires: that the legislature is not in session, which is a matter of fact established by the session calendar, and that circumstances exist which render immediate action necessary, which is a matter of judgment on facts. Material bearing on the second element includes the nature of the exigency, the harm apprehended if action waits for the session, the expected interval before the legislature can address the subject, and the reason the measure takes the form it does. Material that does not bear on those elements, and whose presence may vitiate the satisfaction if it operated on the decision, includes the electoral convenience of the government, the desire to avoid a debate the government expects to lose, and pressure from interests the measure benefits, where those considerations substitute for an exigency rather than accompany one. The line is not between pure and impure governments; it is between material that answers the constitutional question and material that answers a different question, and the court’s enquiry, as the majority framed it, is confined to whether the satisfaction was based on some relevant material and was not a fraud on power or actuated by an oblique motive.

Two practical corollaries follow. The first is that contemporaneity matters: a satisfaction reconstructed after challenge, from material assembled for the litigation, invites the finding that there was no satisfaction at all in the constitutional sense, because the enquiry is into the satisfaction that preceded the promulgation. Governments are therefore advised by their own law officers to perfect the file at the time, and the post-2017 promulgation file, with its noting of the exigency and of the session position, is a recognisably more careful document than its predecessors. The second corollary concerns the aid and advice structure. The satisfaction is formally that of the President or the Governor, and the material is in substance the advice and record of the Council of Ministers; review of the satisfaction is therefore, in reality, review of the ministerial decision to advise promulgation, conducted through the constitutional form. That is why the majority could describe the power as legislative in character and yet reviewable: the review fastens on the executive decision embedded in the legislative act, and leaves the act’s policy, once the condition is satisfied, to the legislature at the laying stage. The two controls, judicial review of the condition and legislative control of the content, divide the ordinance between the institutions the Constitution assigns to each, and the judgment’s architecture is complete only when both are seen operating together.

An agenda for research

For the researcher, Krishna Kumar Singh closes one set of questions and opens another, and the open questions are empirical as much as doctrinal. Wadhwa’s contribution was a count; the corresponding count for the period after 2017 has yet to be assembled with the same completeness for all States: how many ordinances each government promulgated, how many were re-promulgated, how many were laid and after what interval, and how many series were maintained across the replacement of governments. Legislative secretariats publish the raw materials in bulletins and gazettes, but the aggregation that made Wadhwa’s book a constitutional event has to be performed by scholarship, and the judgment supplies the categories the aggregation should use, laying, repetition, interval and survival litigation, rather than a bare total of ordinances. A second research field is the survival test in action: the decided cases in which courts have applied the construction enquiry to appointments, levies, elections and schemes since 2017, coded by the kind of consequence claimed and the treatment of irreversibility, would show whether the test is hardening into predictable sub-rules or remaining an open discretion, and whether the relying citizen is in practice protected as generously as the majority’s formulation promises. A third field is comparative within the federation: States differ widely in ordinance culture, and the States in which re-promulgation was endemic are natural subjects for before and after study against States in which the ordinance remained exceptional. The judgment gives researchers a doctrine whose success is measurable, and measuring it is the tribute the next Wadhwa will pay to this one.

Democratic accountability as the judgment’s measure

It is worth stepping back from the four holdings to the value they jointly serve, because the judgment is sometimes taught as a technical ruling on Articles 123 and 213 when it is in substance a ruling on democratic accountability. A legislature is not merely a factory for statutes. It is the institution in which the government’s programme is exposed to an opposition with the right to question it, in which amendments can be moved by members who do not control the executive, in which a measure’s defects are placed on a public record before the measure binds anyone, and in which the governing majority must periodically count itself. Every one of those functions is suspended when law is made by ordinance, and the Constitution tolerates the suspension only for the interval in which the legislature cannot sit and only under the discipline of the laying procedure, which is designed to restore each function in compressed form at the first opportunity. Re-promulgation is objectionable, on this account, not because it is untidy but because it converts a suspension into an abolition for the measure concerned: no questions, no amendments, no record, no count. The fraud language of Wadhwa and Krishna Kumar Singh is the law’s recognition that the injury is to the democratic process itself and not merely to a rule of legislative procedure, and the seriousness with which the seven judges treated a failure to lay, an act that in other settings would be a venial parliamentary omission, follows from the same recognition. The table of an Assembly is, for an ordinance, the whole of democracy there is.

That framing also explains the judgment’s otherwise puzzling distribution of severity. The Court was severe with governments, characterising their conduct in the vocabulary of fraud and infraction, and severe in doctrine, overruling two Constitution Benches; yet it was restrained in the remedies it prescribed, refusing to convert cessation into voidness, preserving the possibility of survival where irreversibility and public interest carry the case, and leaving the single justified re-promulgation outside the condemnation. The pattern is consistent with a judgment aimed at a process rather than at outcomes. Outcomes under ordinances are often benign, occasionally admirable; the process is what the Constitution guarantees, and the judgment protects the guarantee by pricing its breach rather than by confiscating everything the breach produced. Readers who approach the case asking whether the Court trusted the executive too little or the citizen too much are asking the wrong question. The Court trusted the legislative process, and arranged the incentives of executives and citizens alike so that the process would be used. In the series to which this article belongs, that arrangement has a familiar shape: the allocation of risk determines conduct, and a Constitution that wishes its legislatures to legislate must make the alternative expensive for those best placed to choose it. Krishna Kumar Singh is the ordinance chapter of that general lesson, and it is likely to be the last word on its subject for a long time, not because its questions are exhausted but because its answers are complete enough to govern the questions that remain.

A note on reading the judgment itself

Students and practitioners approaching the reports should read the three opinions in the order that reveals their disagreements most clearly. The majority judgment of Chandrachud J. should be read first and whole, because its conclusions are collected and numbered and because the later debate about the case is a debate about those conclusions. Thakur C.J.’s concurrence should be read next, and principally for its treatment of the series as a single transaction and for its reservation on the laying question, which is the point a careful advocate will note before citing the laying holding to a bench. Lokur J.’s opinion should be read last and without the prejudice the word dissent carries: on enduring rights it is an agreement with different reasons, on re-promulgation it is a refinement the majority’s language can accommodate, and on laying it is a complete alternative theory of Article 213(2), stated with a textual rigour that any future reconsideration of the point would have to answer. Citation practice should reflect that structure. The four holdings are cited to the majority; the fraud conclusion may also be cited to the concurrence; the laying holding should be cited with the dissent acknowledged; and the overruling of Bhupendra Kumar Bose and T. Venkata Reddy may be cited to the whole Court, since on that question the bench did not divide in substance. Reports that flatten the three opinions into a single voice deprive the reader of the judgment’s actual contours, and the case is a standard examination subject precisely because its contours are examinable.

Frequently Asked Questions

Q: What did the Krishna Kumar Singh judgment decide?

In Krishna Kumar Singh v State of Bihar, (2017) 3 SCC 1, a bench of seven judges decided four questions. Re-promulgation of ordinances is a fraud on the Constitution and a subversion of democratic legislative processes. The satisfaction of the President or the Governor behind an ordinance is open to judicial review on limited grounds. Laying an ordinance before the legislature is mandatory, and failure to lay is a serious constitutional infraction. And rights created by an ordinance do not automatically survive its cessation; survival is decided case by case under a test of public interest and constitutional necessity. The appeals were dismissed.

Q: Why is re-promulgation a fraud on the Constitution?

Because it uses a power given for emergencies to displace the legislature altogether. The ordinance power in Articles 123 and 213 is conditional: the legislature must not be in session, and immediate action must be necessary. An ordinance is then given a short life at the end of which the legislature must have the chance to approve, disapprove or replace it. Re-promulgation withholds that chance while keeping the law alive, series after series, so the executive legislates on a continuing basis without a vote, a debate or an amendment. D.C. Wadhwa named that practice a fraud on the Constitution in 1987, and Krishna Kumar Singh reaffirmed the description in 2017.

Q: Is laying an ordinance before the House mandatory?

Yes. The majority in Krishna Kumar Singh, speaking through Chandrachud J., held that laying an ordinance before Parliament or the State Legislature is a mandatory constitutional obligation of the government, because laying is the mechanism by which the legislature decides whether the promulgation was needed, whether the ordinance should be approved or disapproved, and whether an Act should replace it. A failure to lay is a serious constitutional infraction and an abuse of the constitutional process. Lokur J. dissented on this point, holding the requirement directory, and Thakur C.J. in his concurring opinion left the interpretation of the laying clauses open; the majority view is the law declared by the bench.

Q: Do rights created by an ordinance survive its lapse?

Not by default. Before 2017 the answer was yes, because State of Orissa v Bhupendra Kumar Bose and T. Venkata Reddy treated an ordinance on the analogy of a temporary Act whose effects endure after expiry. Krishna Kumar Singh overruled that doctrine. An ordinance that has ceased to operate leaves no presumption of survival behind it. Whether a particular right, privilege, obligation or liability survives is a question of construction, answered by the test of public interest and constitutional necessity, including whether the consequences of the ordinance have become irreversible or impossible to unwind. The person claiming survival must establish the case for it.

Q: Which earlier decisions on ordinances were overruled?

Two Constitution Bench decisions were overruled on the enduring rights question. State of Orissa v Bhupendra Kumar Bose, 1962 Supp (2) SCR 380, had held that rights and liabilities created under an ordinance survive its cessation, on the analogy of a temporary enactment. T. Venkata Reddy v State of Andhra Pradesh, (1985) 3 SCC 198, had followed that view and had also treated the satisfaction behind an ordinance as beyond enquiry on grounds of motive. Krishna Kumar Singh held that the theory of enduring rights in those decisions does not state the correct law, and that the decisions are no longer good law, being inconsistent with S.R. Bommai as well.

Q: What was decided in the Bihar re-promulgation case?

The Bihar re-promulgation case is D.C. Wadhwa v State of Bihar, (1987) 1 SCC 378. Dr Wadhwa’s research, reproduced in the judgment, showed that between 1967 and 1981 the Governor of Bihar promulgated 256 ordinances and kept them alive for periods ranging from one year to fourteen years by repeated re-promulgation, sixty-nine of them repeatedly with the prior permission of the President. A Constitution Bench held that the ordinance power is an emergency power and not a parallel source of legislation, that the practice of re-promulgation was a fraud on the Constitution and a subversion of the democratic process, and it struck down the particular ordinance before it, while expressing the hope that the practice would not recur.

Q: How is the enduring rights question decided now?

It is decided as a matter of construction in each case, under the test laid down in Krishna Kumar Singh. The court examines the ordinance’s own terms and asks whether public interest and constitutional necessity support the survival of the consequence in question. A central enquiry is whether the consequence has assumed an irreversible character, or whether unwinding it would be impracticable and harmful, and the court may mould its relief to preserve a consequence in part. There is no presumption either way: cessation of the ordinance is neither treated as voidness from the beginning nor as repeal with a saving of everything done under it.

Q: Has re-promulgation stopped?

No, though its legal consequences have changed. Parliamentary records show series of re-promulgated ordinances after January 2017, including the Muslim Women (Protection of Rights on Marriage) Ordinance, promulgated in September 2018 and again in January 2019 and February 2019 before enactment in July 2019. Legislative research organisations that publish session counts of Union and State ordinances have recorded continued resort to ordinances and instances of re-promulgation, particularly in the States. Commentators accordingly describe the judgment as having transformed the law while reducing the practice only partly: a re-promulgated series is unconstitutional, the satisfaction is examinable, and consequences carry no presumption of survival, so the practice persists under a heavier legal risk than before.

Q: How does re-promulgation differ from a fresh ordinance on a new subject?

A fresh ordinance on a new subject responds to a situation that has arisen while the legislature is not sitting, and it will be laid, debated and replaced or allowed to lapse in the ordinary course. Re-promulgation is the repetition of substantially the same ordinance as its predecessor approaches the end of its constitutional life, characteristically without the predecessor having been converted into an Act or even laid before the legislature. The difference is therefore one of function: the fresh ordinance uses the power for the interval the power exists to cover, while the re-promulgated series uses repetition to make the interval permanent and to govern without the legislature’s participation.

Q: What must a challenger prove to attack the satisfaction behind an ordinance?

A challenger must attack the existence or the lawfulness of the satisfaction, not its wisdom. After Krishna Kumar Singh, the grounds are that there was no satisfaction at all, that the satisfaction was a fraud on power, that it was actuated by an oblique motive, or that it rested on extraneous or irrelevant material or on no material. The pleading should engage the promulgation record: the material placed before the President or the Governor and the circumstances said to require immediate action. A challenge that the emergency was exaggerated, or that the policy was wrong, invites the answer that adequacy of material is outside the court’s enquiry.

Q: Will a court examine the adequacy of the material behind an ordinance?

No. The majority in Krishna Kumar Singh stated the limit expressly: in judicial review of the satisfaction under Articles 123 and 213, the court will not determine the sufficiency or adequacy of the material on which the satisfaction rests. The enquiry is confined to whether the satisfaction was based on some relevant material, whether there was a satisfaction at all, and whether it was vitiated by fraud on power or an oblique motive. This division follows the approach in S.R. Bommai to the review of a constitutional satisfaction: the court polices the legality of the decision to promulgate, and the legislature, at the laying stage, judges the merits of what was promulgated.

Q: What follows when an ordinance is not laid before the legislature?

The failure is itself a constitutional wrong. The majority in Krishna Kumar Singh held the laying obligation to be mandatory and described a failure to comply with it as a serious constitutional infraction and an abuse of the constitutional process, because non-laying denies the legislature its three determinations: the need for the promulgation, approval or disapproval, and replacement by an Act. Where non-laying accompanies a series of re-promulgations, as in the Sanskrit schools case, the ordinances produced are unconstitutional. The precise consequence for a single unlaid ordinance that was never re-promulgated was not isolated by the majority, and Lokur J. in dissent would have attached no consequence to the omission at all.

Q: What did the Patna High Court hold in the Sanskrit schools litigation?

The teachers and staff of the 429 taken-over Sanskrit schools petitioned the Patna High Court for salary and the emoluments of State employment, resting their claim on the series of ordinances by which Bihar had taken over the management of the schools from 1989 onwards. The High Court dismissed the writ petitions. Relying on D.C. Wadhwa, it held the successive re-promulgations of the takeover ordinance unconstitutional, since none of the ordinances had been placed before the State Legislature and the series had been maintained to govern without it. A claim founded on the instruments produced by that practice could not be sustained, and the Supreme Court affirmed the dismissal in Krishna Kumar Singh.

Q: Why was Krishna Kumar Singh heard by a bench of seven judges?

Because the questions referred required the reconsideration of two decisions of Constitution Benches of five judges. The appellants relied on T. Venkata Reddy, (1985) 3 SCC 198, and on the enduring rights doctrine of State of Orissa v Bhupendra Kumar Bose, each decided by five judges, for the propositions that the satisfaction behind an ordinance could not be questioned and that rights under an ordinance survive its cessation. A bench of coordinate or lesser strength cannot overrule such decisions. A bench of seven was therefore constituted to give an authoritative pronouncement, and it overruled the enduring rights doctrine and displaced the immunity that T. Venkata Reddy had attached to the satisfaction.

Q: How many times was the Sanskrit schools ordinance re-promulgated?

Seven times in succession. The Bihar Non-Government Sanskrit Schools (Taking Over of Management and Control) Ordinance was first promulgated in 1989, and the judgment records seven successive re-promulgations of that first ordinance. Thakur C.J., in his concurring opinion in Krishna Kumar Singh, described the conclusion drawn from the series: the repeated re-promulgations were a fraud on the Constitution, especially as none of the ordinances was ever tabled before the Bihar Legislative Assembly as Article 213(2) requires. The promulgations and re-promulgations were treated as a single transaction and a single series on the same subject, so the invalidity attached to the series as a whole and not only to its later members.

Q: What happened to the teachers’ claim for salary in Krishna Kumar Singh?

It failed. The Supreme Court dismissed the appeals and upheld the Patna High Court. The teachers’ claim to be treated as employees of the State rested on the takeover ordinances, and those ordinances were the product of a practice held to be a fraud on the Constitution: a first ordinance of 1989 re-promulgated seven times without any of the series being laid before the Assembly. With the instruments vitiated, and with the doctrine of enduring rights overruled in the same judgment, no surviving entitlement carried the claim. The case is a hard illustration of the risk allocation the judgment creates: the constitutional wrong was the government’s, and the loss fell on those who had relied on the government’s ordinances.

Q: What happens to payments made under an ordinance after it lapses?

A payment made and received under an ordinance while it was in force was made under law in force, and the lapse of the ordinance does not by itself make the payment recoverable or the receipt wrongful, because cessation is not voidness from the beginning. The harder question is the converse one, whether an unpaid entitlement under a lapsed ordinance can still be demanded, and after Krishna Kumar Singh that question is answered by construction: the court applies the test of public interest and constitutional necessity, asking whether the consequences have become irreversible and whether unwinding or enforcing them serves the public interest, and it may mould relief. There is no automatic saving in either direction.

Q: Who bears the risk if a benefit granted by ordinance is withdrawn after its lapse?

Primarily the beneficiary. Under the law before 2017, the risk sat with the legislature and the public, because effects created by an ordinance endured by presumption and the executive kept the benefit of its instrument even after the instrument died. Krishna Kumar Singh moved the risk: a benefit that rests on an ordinance alone carries no presumption of survival, and the person claiming it after the lapse must establish, under the test of public interest and constitutional necessity, why the consequence should be preserved. The promulgating executive bears the parallel risks of the other holdings, invalidation of the series and review of its satisfaction, which is the judgment’s design for discouraging ordinance government.

Q: Can a legislature validate acts done under a lapsed ordinance?

Yes, by legislating. The disability Krishna Kumar Singh creates is not a bar on legislative action; it is the absence of any presumption that consequences survive without legislative action. Parliament or a State Legislature may enact an Act that adopts the substance of a lapsed ordinance and saves, expressly, the acts done and rights accrued under it, subject to the ordinary limits on validating legislation, including the requirement that the legislature remove the defect that caused any invalidity rather than merely declare the invalid valid. That route is the constitutionally proper one the judgment points to: durability for what an ordinance did is available, but it must be conferred by the body the Constitution charges with making law.

Q: Can an ordinance be challenged in court after it has ceased to operate?

Yes, where the challenge carries live consequences, and Krishna Kumar Singh is itself the authority for deciding such a challenge on the merits. The Sanskrit schools ordinances had long ceased to operate when the seven judges ruled on their constitutional character, because the ruling determined the petitioners’ claimed status and the validity of the practice. Courts do decline to decide challenges that have become purely academic, but a challenge that governs surviving claims, pending proceedings, or the lawfulness of a continuing practice of re-promulgation presents live questions. The survival enquiry under the fourth holding presupposes that courts will rule on lapsed ordinances whenever the fate of acts done under them is in dispute.