LLM Group 1 Constitutional and Administrative Law Administrative Process and Judicial Control 2019 Question Paper with Solutions

For full marks, cover: what "the administrative process" means, which is broader than administrative law and is the point of the question; the reasons for its growth in common law countries; the four kinds of function the administration performs, rule making, adjudication, discretionary decision and ministerial action, which is the organising structure of the answer; the constitutional difficulty each creates in a system built on the separation of powers; the mechanisms of control that have grown up in response; and a critical assessment of where the process now stands, including the tribunal question.

What the administrative process is

The administrative process is the whole activity by which the modern State carries its policies into effect: making rules, granting and refusing licences, allocating benefits, inspecting, investigating, adjudicating disputes and enforcing decisions. It is wider than administrative law, which is the body of principle by which that activity is controlled. The distinction matters because a question on the process asks what the administration does; a question on administrative law asks how the courts check it.

Function one: rule making

The administration legislates. Rules, regulations, byelaws, orders and notifications are made under statutory powers, and they bind exactly as a statute does. Article 13(3)(a) treats them as "law". The volume is enormous: in any modern system the quantity of delegated legislation vastly exceeds the quantity of primary legislation.

The constitutional difficulty is that legislative power belongs to the legislature. The common law answer is the doctrine of excessive delegation: the legislature must lay down the policy and the standard and may leave only details to the delegate. In re Delhi Laws Act, 1912, AIR 1951 SC 332, held that the power to extend an existing law with modifications not touching its essential features is delegable but that the power to repeal or amend a law is an essential legislative function and is not; Hamdard Dawakhana v. Union of India, AIR 1960 SC 554, struck down a provision allowing the executive to add "any other disease or condition" to a list of prohibited advertisement subjects, because it supplied no standard at all.

The common law answer has two parts. The first is to require natural justice of them: after Ridge v. Baldwin, [1964] AC 40, and A.K. Kraipak v. Union of India, (1969) 2 SCC 262, the duty to act fairly attaches to any power to affect rights, and the earlier requirement of a superadded duty to act judicially is gone. The second is to keep them under the supervision of the ordinary courts: in L. Chandra Kumar v. Union of India, (1997) 3 SCC 261, a seven judge Bench held that judicial review under Articles 226, 227 and 32 is part of the basic structure, struck down Article 323A(2)(d) and Article 323B(3)(d) so far as they excluded it, and held that tribunal decisions are subject to scrutiny by a Division Bench of the High Court. The most recent chapter is Madras Bar Association v. Union of India, 2025 INSC 1330, decided on 19 November 2025, which struck down provisions of the Tribunals Reforms Act, 2021, that gave the executive control over the appointment, tenure and service conditions of tribunal members, held that re-enacting provisions already struck down is an impermissible legislative override, and directed the Union to establish a National Tribunals Commission within four months.

Function four: ministerial and investigative action

A great deal of administration involves no judgment at all: registering, recording, collecting, inspecting, issuing. These attract mandamus where a duty is refused, and little else. Investigation is a distinct category and is more troublesome, because an investigative decision affects reputation and liberty before any adjudication: Barium Chemicals Ltd. v. Company Law Board, AIR 1967 SC 295, holds that although the formation of the requisite opinion is subjective, the existence of the circumstances on which the opinion is founded is a condition precedent and is justiciable.

The scope of control: the mechanisms

The classical objection, and the reply

The nineteenth century view was that the administrative process ought not to exist at all, and Dicey is its author. In An Introduction to the Study of the Law of the Constitution (1885) he gave the rule of law three meanings: the absence of arbitrary power, so that a man may be punished only for a distinct breach of law established before the ordinary courts; equality before the law, all classes being subject to the ordinary law administered by the ordinary courts, which he framed expressly to exclude the French droit administratif; and the proposition that in England the constitution is the result of the ordinary law rather than its source. On that view a body with power to make rules, decide disputes and enforce its own decisions is a contradiction of each of the three.

The Indian statutory apparatus of the process

An Indian answer should show that the process is not an abstraction but a set of identifiable institutions. Rule making runs through the General Clauses Act, 1897, which supplies the rules for the exercise of statutory powers, and through the laying provisions of individual Acts. Adjudication runs through the Administrative Tribunals Act, 1985, the Consumer Protection Act, 2019, the Companies Act, 2013, with the National Company Law Tribunal, the Securities and Exchange Board of India Act, 1992, with the Securities Appellate Tribunal, and the Electricity Act, 2003. Regulation runs through the Reserve Bank, the Securities and Exchange Board of India, the Telecom Regulatory Authority of India, the Competition Commission and the Insurance Regulatory and Development Authority, each of which makes binding regulations, adjudicates disputes and enforces its own orders in the same institution.

Conclusion. The administrative process in common law countries is the whole apparatus by which the modern State gives effect to policy, and its nature is that it combines, in one set of hands, functions the classical separation of powers assigned to three. It legislates by rule making, and is controlled by the doctrine of excessive delegation from In re Delhi Laws Act and Hamdard Dawakhana. It adjudicates through tribunals and statutory authorities, and is controlled by the extension of natural justice to administrative action after Ridge v. Baldwin and Kraipak, and by the constitutional supervision that L. Chandra Kumar placed beyond legislative reach. It decides discretionary questions, and is controlled by a body of doctrine that examines whether the discretion was exercised and whether it was abused, without substituting the court's own choice. And it acts ministerially and investigates, where mandamus and the jurisdictional fact doctrine of Barium Chemicals apply. The scope of the process is therefore as wide as the functions of the State, and the scope of its control is as wide as the courts' willingness to insist that every one of those functions be exercised for a purpose, on relevant material, and for reasons that can be seen.

The claim that a court might control the legislature itself appears in Dr. Bonham's Case, (1610) 8 Co. Rep. 113b. The College of Physicians fined and imprisoned Bonham for practising without its licence, and kept half the fine, so that it was judge in its own cause. Coke CJ said that when an Act of Parliament is against common right and reason, or repugnant, or impossible to be performed, the common law will control it and adjudge such an Act to be void. The claim did not survive the settlement of 1688 and parliamentary sovereignty, but the sentence travelled.

The Government of India Act, 1935, added the second element. By distributing legislative power between the Centre and the Provinces through three lists, it necessarily made legislative competence a justiciable question, and the Federal Court decided such questions. What was missing was any bill of rights, so competence was the only ground and there was nothing corresponding to review for violation of rights.

Stage four: the Constitution of 1950

India did not have to derive review from the nature of a written constitution, because the Constitution conferred it expressly. Article 13 declares laws inconsistent with Part III void and defines "law" to include ordinances, orders, byelaws, rules, regulations and notifications. Article 32 guarantees the right to move the Supreme Court, and Dr. Ambedkar called it the very soul of the Constitution. Articles 226 and 227 confer wider powers on the High Courts. Articles 245 and 246 with the Seventh Schedule preserve the competence question.

I.C. Golak Nath v. State of Punjab, AIR 1967 SC 1643, reversed the position. Eleven judges held by six to five that an amendment is law within Article 13, that Parliament cannot abridge the fundamental rights, and, applying the doctrine of prospective overruling for the first time in India, that the ruling would not affect past amendments.

Kesavananda Bharati v. State of Kerala, (1973) 4 SCC 225, settled it. The head of the Edneer Mutt challenged Kerala land reform legislation, and the case became the vehicle for reconsidering Golak Nath. Thirteen judges sat for sixty eight days. By seven to six the Court overruled Golak Nath to the extent of holding that Parliament may amend any provision, including the fundamental rights, but held that it may not damage or destroy the basic structure of the Constitution. The basic features named across the opinions include the supremacy of the Constitution, the republican and democratic form of government, its secular and federal character, the separation of powers, judicial review itself, free and fair elections and the rule of law.

Stage six: the doctrine applied

The doctrine's significance is that it has been used, not merely stated, and the three principal applications should be given.

Supreme Court Advocates on Record Association v. Union of India, (2016) 5 SCC 1.* The Ninety Ninth Amendment and the National Judicial Appointments Commission Act, 2014, replaced the collegium with a six member Commission comprising the Chief Justice, two senior judges, the Law Minister and two eminent persons, any two of whom could veto a recommendation. The Court struck both down by four to one, holding that the primacy of the judiciary in appointments is part of the independence of the judiciary and therefore of the basic structure, and objecting in particular to the two member veto and to the presence of the Law Minister.

Two decisions extended the doctrine's reach beyond amendments. Indra Sawhney v. Union of India, 1992 Supp (3) SCC 217, and M. Nagaraj v. Union of India, (2006) 8 SCC 212, applied basic structure reasoning to reservation policy through the width and identity tests; and I.R. Coelho v. State of Tamil Nadu, (2007) 2 SCC 1, held that laws placed in the Ninth Schedule after 24 April 1973, the date of Kesavananda, are open to challenge if they violate the basic structure, so that the Schedule is no longer an absolute shelter.

A critical assessment

The case against the doctrine must be stated at full strength in an LLM answer. It has no textual foundation: Article 368 empowers Parliament to amend "this Constitution", and the doctrine reads in a limitation the framers did not write, on the authority of a seven to six majority. Its content is open ended, since a basic feature is whatever a majority of the Court holds to be one, and the list has grown by accretion. It is counter majoritarian in a strong sense, since it defeats an amendment passed by the special majority the Constitution itself prescribes and, in the NJAC case, ratified by a large number of State legislatures. And in that case the Court adjudicated on a matter, judicial appointments, in which it had an institutional interest.

Conclusion. The journey runs from a supervisory device to a constitutional limit on the power to amend the Constitution itself. It begins with the prerogative writs, by which the King's Bench kept inferior jurisdictions within their authority, and with Coke's claim in Bonham that the common law might control an Act of Parliament, a claim England abandoned and America adopted in Marbury. British India received the writ jurisdiction and, under the 1935 Act, the competence question; the Constitution of 1950 added the missing element by conferring review expressly in Articles 13, 32, 226 and 227 and by supplying a Part III to review against. The contest over the amending power then ran from Shankari Prasad through Sajjan Singh and Golak Nath to Kesavananda Bharati, where the Court reached the compromise that has held for over fifty years: the amending power is plenary in its reach and limited in its depth. What makes the doctrine more than a theory is that it has been applied, in Indira Nehru Gandhi to protect free and fair elections, in Minerva Mills to prevent a limited power making itself unlimited, in I.R. Coelho to open the Ninth Schedule, and in the NJAC case to protect judicial independence. It has no text behind it and its content is judicially determined, which are real objections; the answer to them is the record of what it has been used to prevent.

The premise: why it is inevitable

Four reasons make delegated legislation unavoidable and they should be stated in a sentence each. Pressure on legislative time: a legislature that debated every technical detail of every statute would pass a fraction of the law a modern State requires. Technicality: the rules for the design of a boiler, the tolerances of a drug, the classification of a security or the format of an electronic record cannot sensibly be settled on the floor of a House. Flexibility: conditions change faster than statutes can be amended, and a rule can be altered in weeks where an Act takes years. Emergency: war, epidemic, famine and financial crisis require immediate action under powers conferred in advance.

The Committee on Ministers' Powers (Donoughmore, 1932) in England accepted its necessity while warning against its abuse, and no serious Indian writer disputes it. The argument in this branch of the law is therefore not about whether delegation should exist but about how it is controlled, which is what the question asks.

Hamdard Dawakhana v. Union of India, AIR 1960 SC 554, is the striking down. Section 3(d) of the Drugs and Magic Remedies (Objectionable Advertisements) Act, 1954, prohibited advertisements relating to the diseases listed "and any other disease or condition which may be specified in the rules". The Court held the words invalid, because the Act supplied no principle or standard to guide the addition and therefore conferred uncontrolled and unguided power. The contrast is Edward Mills Co. Ltd. v. State of Ajmer, AIR 1955 SC 25, upholding a power to add employments to a schedule of the Minimum Wages Act, 1948, because the preamble and scheme supplied the standard, the prevention of sweated labour.

In taxation the requirement is stricter, and Devi Das Gopal Krishnan v. State of Punjab, AIR 1967 SC 1895, holds that a power to fix a rate of tax with no maximum and no guiding principle is excessive delegation.

A special case is the Henry VIII clause, empowering the executive to modify the parent Act itself to remove difficulties. It is read narrowly, on the authority of In re Delhi Laws Act, as permitting only modifications that do not touch policy or essential features; it is not a licence to legislate afresh.

Three related defects belong here. A rule made by an authority other than the one on whom the power was conferred, unless sub-delegation is authorised, delegatus non potest delegare. A retrospective rule where the Act does not expressly or by necessary implication permit retrospectivity, which Hukam Chand v. Union of India, AIR 1972 SC 2427, forbids. And a rule excluding natural justice where the Act does not exclude it, since the principles are implied into a statutory power unless displaced expressly or by necessary implication.

Ground four: the rule violates the Constitution

Article 13(3)(a) defines "law" to include any ordinance, order, byelaw, rule, regulation, notification, custom or usage having the force of law, so Article 13(2) applies to a rule exactly as to a statute. A rule imposing an unreasonable restriction on trade offends Article 19(1)(g); one drawing an irrational classification offends Article 14; one prescribing an unfair procedure affecting liberty offends Article 21 after Maneka Gandhi v. Union of India, (1978) 1 SCC 248. This ground is independent: a rule that is entirely within its parent Act is still void if it violates Part III.

Ground six: the manner of making

Where the Act attaches conditions to the making of a rule, non compliance may invalidate it, and the question is whether the condition is mandatory or directory. Raza Buland Sugar Co. Ltd. v. Municipal Board, Rampur, AIR 1965 SC 895, supplies the test: the object of the provision and the consequence of non compliance. There, the requirement of publication of a draft tax proposal was mandatory, its purpose being to allow ratepayers to object, but the requirement as to the particular manner of publication was directory, so substantial compliance sufficed. Common conditions are previous publication, consultation with a named body or with affected interests, prior sanction, and laying before the legislature.

What judicial control cannot reach

Two limits should be stated honestly. A court will not examine the wisdom or expediency of a rule, nor substitute its own view of what the rule should contain; the question is always vires and not merit, and a rule that is foolish but within power stands. And judicial control is reactive: it requires a person affected who has standing and the resources to litigate, which means that a rule bearing on people who cannot litigate is in practice never tested. That is why the parliamentary and procedural controls matter even though they are weaker: the Committee on Subordinate Legislation in each House examines whether rules exceed the power conferred, whether they impose taxation, whether they exclude the jurisdiction of courts and whether they were laid in time, and requirements of previous publication give those affected a chance to object before the rule exists rather than a remedy afterwards.

For full marks, cover: the deliberate omission of "due process" from the Indian Constitution and the Constituent Assembly history, since the question turns on it; A.K. Gopalan and the position that omission produced; Maneka Gandhi and the judicial reintroduction of due process in substance; then the two halves the question names, procedural fairness and substantive fairness, treated separately with their own authorities; the constitutional homes of natural justice in Articles 14 and 21; the limits and exceptions; and a critical assessment of a judiciary that has restored by construction what the framers deliberately left out.

The consequence: A.K. Gopalan

A.K. Gopalan v. State of Madras, AIR 1950 SC 27, gave the omission its full effect. A communist leader detained under the Preventive Detention Act, 1950, challenged it. The Supreme Court held, by a majority, that "procedure established by law" means procedure prescribed by a statute duly enacted by a competent legislature, and that the Court could not examine whether the procedure was fair or reasonable; and that Articles 19, 21 and 22 operate in separate compartments, so that a preventive detention law falls to be tested against Article 22 alone and the reasonableness required by Article 19 has no application. Fazl Ali J. dissented, holding that the Articles are not mutually exclusive and that the principles of natural justice are part of the general law of the land.

The result was that for twenty eight years an Indian citizen deprived of liberty could complain only that no statute authorised the deprivation, not that the statute was unfair.

Procedural fairness: the judiciary's role

Procedural fairness in India rests on the two limbs of natural justice, both now anchored in Articles 14 and 21.

Nemo judex in causa sua, the rule against bias. In A.K. Kraipak v. Union of India, (1969) 2 SCC 262, a candidate for selection to the Indian Forest Service sat on the selection board and was placed at the top of the list; the Supreme Court quashed the selection, holding that the line between quasi judicial and administrative functions is thin and being obliterated, and that natural justice applies to administrative action. In Ranjit Thakur v. Union of India, (1987) 4 SCC 611, a court martial in which the officer against whom the accused had complained participated was quashed, the test being whether a reasonable person would entertain a reasonable apprehension of bias.

The judiciary's contribution has been to make these requirements constitutional rather than merely implied. Before Maneka Gandhi a statute could displace natural justice and that was the end of the matter; after it a procedure that is not fair fails Article 21 itself, and the statute is open to challenge.

Substantive fairness: the judiciary's role

Substantive fairness asks not how the decision was reached but whether its content is one the Constitution permits, and India has developed three routes to it.

The first is arbitrariness under Article 14. E.P. Royappa v. State of Tamil Nadu, (1974) 4 SCC 3, held that equality and arbitrariness are sworn enemies and that an arbitrary act is by that fact unequal and violates Article 14. Shayara Bano v. Union of India, (2017) 9 SCC 1, took the doctrine to legislation itself, confirming manifest arbitrariness as a ground for striking down a law and applying it to instantaneous triple talaq.

The exceptions and limits

Natural justice is excluded or attenuated in cases of genuine urgency, where prior notice would frustrate the purpose, where the statute expressly excludes it, in legislative as distinct from administrative action, and under the narrow doctrine of useless formality, which M.C. Mehta v. Union of India, (1999) 6 SCC 237, confined to cases where on admitted or indisputable facts only one conclusion is possible. Where urgency justifies deferring a hearing, Swadeshi Cotton Mills v. Union of India, (1981) 1 SCC 664, holds that the rule may be moulded but not abrogated, and that a post-decisional hearing must be real, the authority being willing to reconsider and able to undo what it has done.

A critical assessment

The honest assessment is that the judiciary has restored by construction what the framers deliberately removed, and that this is both the achievement and the objection. Frankfurter's warning to Rau was that due process gives a small number of judges a veto over social and economic legislation; Indian courts have exercised something very like that veto, and Shayara Bano shows manifest arbitrariness being used against legislation itself, which is precisely the power the Assembly declined to confer.

The Indian authorities the examiner expects

Three Indian decisions carry this branch of the subject, and an answer resting only on the English cases is incomplete. In State of Orissa v. Dr. (Miss) Binapani Dei, AIR 1967 SC 1269, an Assistant Surgeon in the Orissa Medical Service was compulsorily retired after the State unilaterally altered her recorded date of birth on an enquiry at which she had no real opportunity to meet the material against her. The Supreme Court set the order aside and held that even an administrative order which involves civil consequences must be made consistently with the rules of natural justice. That proposition preceded A.K. Kraipak by two years and is the true Indian starting point for everything said above.

Conclusion. The Indian Constitution does not contain a due process clause because the framers, on Frankfurter's advice through B.N. Rau, deliberately replaced it with "procedure established by law". A.K. Gopalan gave that choice its literal effect and left the citizen with nothing but the question whether a statute existed. The judiciary's role since 1978 has been to restore, through construction, both halves of what was omitted. Procedural fairness came back through Maneka Gandhi, which held that the procedure under Article 21 must be right, just and fair, read audi alteram partem into a silent statute, and fused Articles 14, 19 and 21; it is applied through the two limbs of natural justice, most recently in Krishnadatt Awasthy in 2025. Substantive fairness came back through the arbitrariness doctrine of Royappa and Shayara Bano, through proportionality after Om Kumar and Modern Dental College, and through the expansion of Article 21 in the legal aid, speedy trial, prison and privacy cases. The result is a jurisdiction wider than the text suggests and narrower than American due process in one important respect, that it has been used chiefly to enlarge protection rather than to defeat social legislation. Whether the courts were entitled to restore what the Assembly removed is a fair question; the reason they have not been seriously challenged on it is ADM Jabalpur, which showed the country exactly what the Gopalan position was worth when it mattered.

A pre-decisional hearing serves all three. A post-decisional hearing serves the first two imperfectly and the third scarcely at all, because by then the reasons are formed and the authority is defending a position rather than reaching one. That is the difficulty Swadeshi Cotton Mills had to resolve.

The law before the case

Maneka Gandhi v. Union of India, (1978) 1 SCC 248, had held three things three years earlier. That Articles 14, 19 and 21 form an integrated scheme; that the procedure under Article 21 must be right, just and fair, and not arbitrary, fanciful or oppressive; and that audi alteram partem must be read into a statute that does not exclude it expressly or by necessary implication. But it had also, on the Attorney General's undertaking that the petitioner would be heard, disposed of the case without quashing the order, Bhagwati J. observing that where immediate action is necessary a post-decisional hearing may satisfy fairness. That observation was made in a case where the Court did not have to decide the point, and it left the law unsettled.

Proposition one: the audi alteram partem rule is flexible but not dispensable. The rule is not a rigid or cast iron formula; it may be moulded to fit the situation, the nature of the power, the framework of the statute and the interests at stake. But it may not be abrogated unless the statute excludes it expressly or by necessary implication, and exclusion is not to be inferred merely from the absence of a provision for a hearing.

Proposition two: "immediate action" does not exclude natural justice. The words in Section 18AA relate to the urgency of taking over management, not to the exclusion of the person affected from the process. A power to act at once is not a power to act without ever hearing; at most it justifies postponing the hearing.

Proposition three: where urgency is genuine, a pre-decisional hearing may be replaced by a post-decisional one. The majority accepted the Maneka Gandhi concession and gave it a frame: the urgency must be real, and it must appear from the material, not merely be asserted.

The criticism: K.I. Shephard

K.I. Shephard v. Union of India, (1987) 4 SCC 431, is the strongest Indian statement against the device and should be set against Swadeshi Cotton Mills. Three private banks were amalgamated with nationalised banks under schemes framed under Section 45 of the Banking Regulation Act, 1949, and certain employees were excluded from employment without being heard; they were offered an opportunity to represent afterwards. The Supreme Court held the exclusion bad, observing that once a decision has been taken there is a tendency to uphold it, that a representation may not yield any fruitful purpose, and that a post-decisional hearing is no substitute for a pre-decisional one where there was no reason why a hearing could not have been given first.

The two decisions are not in conflict, and an answer should say why. Swadeshi Cotton Mills says that where urgency is genuine, the later hearing may replace the earlier one and must be real. K.I. Shephard says that where there was no urgency, the later hearing does not save the decision at all. Read together, the rule is that the necessity must be established before the concession is available.

The most recent authority reinforces the primacy of the earlier hearing. In Krishnadatt Awasthy v. State of M.P., 2025 INSC 126, decided on 29 January 2025, the appointments of fourteen Shiksha Karmi teachers were cancelled for bias and nepotism without hearing them, and it was argued that a later revision cured the defect. The Court held that audi alteram partem is the cornerstone of justice and that a defect at the initial stage cannot generally be cured at the appellate stage, because an authority reviewing a decision taken without a hearing has no properly reasoned decision before it. The appeal succeeded and the appointments stood, the Court setting aside the ex parte cancellation. The case had reached three judges because a two judge Bench split on 4 April 2024, Maheshwari J. holding the selection vitiated by bias and Vishwanathan J. holding the cancellation vitiated by the denial of hearing, so it is a rare instance of the two limbs of natural justice pulling against each other, with the fair hearing limb prevailing.

The Indian authorities the examiner expects

Three Indian decisions carry this branch of the subject, and an answer resting only on the English cases is incomplete. In State of Orissa v. Dr. (Miss) Binapani Dei, AIR 1967 SC 1269, an Assistant Surgeon in the Orissa Medical Service was compulsorily retired after the State unilaterally altered her recorded date of birth on an enquiry at which she had no real opportunity to meet the material against her. The Supreme Court set the order aside and held that even an administrative order which involves civil consequences must be made consistently with the rules of natural justice. That proposition preceded A.K. Kraipak by two years and is the true Indian starting point for everything said above.

Prejudice, and what a breach of the hearing rule actually costs

The modern Indian question is not only whether a hearing was denied but what follows from the denial, and two decisions settle it. In Managing Director, ECIL, Hyderabad v. B. Karunakar, (1993) 4 SCC 727, a Constitution Bench held that the enquiry officer's report must be furnished to the delinquent employee before the disciplinary authority acts on it, for the plain reason that a person cannot answer a report he has never seen. But the Court declined to make the consequence automatic: the order is not quashed as a matter of course, the court must ask whether the non supply caused prejudice, and the ordinary relief is to remit the proceeding to the stage at which the defect occurred rather than to order reinstatement with back wages. That is the practical answer to a candidate who assumes that every procedural breach destroys the decision.

Conclusion. Swadeshi Cotton Mills is the case in which Indian law fixed the frame for this whole subject. It held that the audi alteram partem rule is flexible but not dispensable, that it may be moulded and not abrogated unless the statute excludes it expressly or by necessary implication, that words permitting immediate action do not by themselves exclude it, that where genuine urgency exists a post-decisional hearing may replace a pre-decisional one, and, decisively, that such a hearing must be real, before an authority willing to reconsider and able to undo. K.I. Shephard added that where no urgency existed the later hearing saves nothing, and M.C. Mehta kept the useless formality escape narrow. Measured against the three purposes a hearing serves, the post-decisional hearing is plainly inferior, and the courts have treated it accordingly: not as an equivalent, but as the most that can be had where the alternative is a power that cannot be used at all. The significance of the distinction is therefore that the burden lies on the administration to show why it could not hear first, and Krishnadatt Awasthy in 2025 shows the Court still applying exactly that discipline.

The family they belong to

Three kinds of statutory provision attempt to insulate an administrative decision, and they work in three different ways. A finality clause says the decision "shall be final", and operates on the right of appeal. An ouster or exclusionary clause says the decision "shall not be called in question in any court", and operates on the jurisdiction of the court. A conclusive evidence clause says a document or finding "shall be conclusive evidence" of a stated fact, and operates on the law of evidence, leaving the court's jurisdiction untouched but removing the material on which it could act.

The courts' answer to all three rests on a single distinction: between an error made within jurisdiction and an act done without jurisdiction. A clause may protect the first. It cannot protect the second, because an act done without jurisdiction is a nullity, and a provision protecting "a decision", "a determination" or "a certificate" does not protect what in law is none of those things.

The danger, and the ground of control, is that such a clause can convert a jurisdictional fact into an unexaminable one. Where a statute makes a power exercisable only on the existence of a state of affairs, the existence of that state of affairs is a condition precedent to the power and is examinable by the court however the statute is framed. Barium Chemicals Ltd. v. Company Law Board, AIR 1967 SC 295, is the authority: an order for investigation into a company's affairs was set aside because the circumstances disclosed did not answer the statutory description on which the requisite opinion had to be founded, the Court holding that although the formation of the opinion is subjective, the existence of the circumstances is objective and justiciable. A certificate cannot therefore be conclusive of a state of affairs that the authority had no power to certify, because the clause presupposes a valid exercise of the power it protects.

Three further limits apply. The clause is confined to the precise fact declared conclusive and to no other. It cannot displace a challenge on the ground of mala fides or fraud, since fraud unravels everything. And it cannot exclude a challenge based on breach of natural justice or on violation of Part III, since after L. Chandra Kumar v. Union of India, (1997) 3 SCC 261, judicial review under Articles 32, 226 and 227 is part of the basic structure and cannot be excluded even by constitutional amendment.

(b) Finality clauses

A finality clause provides that the decision of a specified authority "shall be final", or "final and conclusive", or "shall not be liable to be questioned". Its object is to bring proceedings to an end within the statutory scheme, and its natural meaning is that no further appeal lies.

Kihoto Hollohan v. Zachillhu, 1992 Supp (2) SCC 651, shows the technique applied to a finality clause in the Constitution itself, and is the best illustration available. Paragraph 6(1) of the Tenth Schedule provides that the decision of the Speaker or Chairman on a question of disqualification for defection shall be final, and paragraph 7 barred the jurisdiction of all courts including under Articles 136, 226 and 227. The Supreme Court held paragraph 7 invalid for want of ratification by the States under the proviso to Article 368(2); and it held that the Speaker, in deciding such a question, acts as a tribunal, so that the finality in paragraph 6(1) does not exclude judicial review on the grounds of mala fides, perversity, violation of a constitutional mandate and violation of natural justice, though the Court will not interfere at an interlocutory stage before the decision is made.

A finality clause therefore does real work, and this should be said plainly. It removes the appeal, so a decision that is merely wrong stands; and it shifts the burden onto the challenger to establish a jurisdictional defect, a breach of natural justice or mala fides rather than error. An answer that says finality clauses "have no effect" overstates the position and misses what they achieve.

The three clauses compared

A clause is construed strictly and against the person relying on it, because it derogates from the ordinary right of access to a court, which the Supreme Court has repeatedly described as a valuable right. Any ambiguity is resolved in favour of preserving the jurisdiction, so a clause that speaks of an "appeal" is not read as excluding review, and one that makes a decision final "for the purposes of this Act" is confined to that Act.

A clause covers only what it names. A provision making a certificate conclusive of service of notice says nothing about whether the notice was one the authority had power to issue; a provision making an assessment final says nothing about whether the assessing officer had jurisdiction over the assessee. This is the commonest way in which such clauses are defeated, and it requires no constitutional doctrine at all, merely careful reading.

A clause cannot enlarge the power it protects. The protection presupposes a valid exercise; it cannot supply one. This is the proposition that unites Barium Chemicals on jurisdictional facts with Anisminic on wrong questions, and it explains why the two lines of authority reach the same result by different routes.

What a legislature should do instead

The practical lesson of the case law is worth a paragraph, because it is what a well drafted statute now does. Instead of attempting to exclude review, modern statutes channel it: they provide a specialist appellate tribunal, prescribe a short limitation period for challenges, require reasons to be recorded, and confine further challenge to questions of law. That approach survives judicial scrutiny because it does not exclude the constitutional courts; it merely ensures that the specialist body decides first and that the record is complete when the matter reaches them. L. Chandra Kumar v. Union of India, (1997) 3 SCC 261, expressly contemplates that structure, holding that tribunals may act as courts of first instance in their fields while their decisions remain subject to scrutiny before a Division Bench of the High Court concerned.

For full marks, cover: two notes of roughly equal length, and the point that unites them, that both are discretionary limits the courts impose on themselves rather than limits the Constitution imposes; for laches, the meaning, the rationale, the absence of any fixed period, the two questions a court actually asks, the exceptions, and the position in service matters; for res judicata, Section 11 CPC and its five conditions, constructive res judicata under Explanation IV, the extension to writ proceedings by Daryao, the exceptions, the habeas corpus position, and the distinctions from estoppel and stare decisis.

Its rationale is not administrative tidiness but the protection of third parties. During the delay, appointments may have been made, promotions given, contracts performed, buildings constructed and money spent. A court that grants relief long afterwards may unsettle more than it settles, and the injustice it does to those who acted on the decision may exceed the injustice it repairs.

No period is prescribed and the Limitation Act, 1963, does not apply to a writ petition. What a court actually asks is two questions. Is the delay explained? A petitioner who was pursuing a departmental remedy, was under a disability, or was unaware of the facts through no fault of his own, may be excused years. Would relief now be unjust to others? Where rights have accrued to third parties, even a short delay may be fatal.

(b) Res judicata

Res judicata means that a matter finally adjudicated between the same parties by a competent court cannot be litigated again. It rests on three maxims of public policy: interest reipublicae ut sit finis litium, it is in the interest of the State that there be an end to litigation; nemo debet bis vexari pro una et eadem causa, no one should be vexed twice for the same cause; and res judicata pro veritate occipitur, a decided matter is accepted as true.

Section 11 of the Code of Civil Procedure, 1908, states the rule for suits and requires five conditions. The matter must have been directly and substantially in issue in the former suit; between the same parties or those claiming under them; litigating under the same title; before a court competent to try the subsequent suit; and it must have been heard and finally decided.

Explanation IV contains constructive res judicata: any matter which might and ought to have been made a ground of attack or defence in the former proceeding shall be deemed to have been directly and substantially in issue. Its effect is that a litigant must bring his whole case at once and cannot keep a ground in reserve.

Constructive res judicata applies to writ proceedings. In Direct Recruit Class II Engineering Officers' Association v. State of Maharashtra, (1990) 2 SCC 715, the Supreme Court applied it in service litigation, holding that a party who has litigated and lost cannot return with a ground that was available and was not taken.

Habeas corpus is the principal exception, and the reason is the same reason that laches does not apply to it. In Lallubhai Jogibhai Patel v. Union of India, (1981) 2 SCC 427, the Court held that constructive res judicata does not apply to a petition for habeas corpus, so a ground available but not taken earlier may be taken later, though a successive petition on the same ground is not entertained. The justification is that the legality of a detention is tested as at the date of the return and that liberty is a continuing matter, so a rule requiring every argument to be exhausted at once would be inconsistent with the nature of the writ.

How a court actually applies them

Both doctrines are applied through a sequence of questions rather than by a rule, and setting the sequence out shows that the answer is understood rather than memorised.

On laches the court asks four things in order. When did the cause of action arise, which is not always the date of the order, since a person may learn of it later or the injury may accrue afterwards? What did the petitioner do in the interval, a departmental representation or a pending appeal being an explanation and mere inaction being none? Have third party rights accrued, appointments made, contracts performed, construction completed? And can the injustice be avoided by moulding the relief, granting the declaration but confining its operation prospectively, which is the standard course in service matters and allows a court to correct the principle without unsettling those who acted on the old position.

Where the two doctrines meet

They meet in the rule that a dismissal for laches does not operate as res judicata, which Daryao states expressly. The reason repays attention: a dismissal for delay is not an adjudication of the rights in dispute, so the finality that res judicata protects never arose. It follows that a petitioner dismissed for laches in the High Court is not, in principle, barred from the Supreme Court, though he will ordinarily meet the same objection there. They also meet in their treatment of a void order: neither doctrine easily protects a nullity, since time cannot validate what was never valid and a decision by a court without jurisdiction binds nobody, though relief in both cases remains discretionary.

Forward Construction Co. v. Prabhat Mandal (Regd.), Andheri, (1986) 1 SCC 100, confirms the position for both Articles 32 and 226, holding that a matter which might and ought to have been raised in the earlier proceeding is barred in the later one, and that the law discourages a litigant who withholds a ground for future use. Between them these two decisions, and not Direct Recruit alone, are what an examiner expects on constructive res judicata in this subject.

Laches in the Indian cases, and the spelling this paper uses

The paper prints "Latches". The doctrine is laches, and the two are the same thing, the spelling being the paper's own; a candidate should answer on laches and not be thrown by it. The Indian authorities are settled and they are worth naming, because this is the one head in the subject where answers habitually cite nothing at all.

Conclusion. Laches and res judicata are the two ways in which a court refuses relief to a petitioner who may be entirely right on the merits, and both are justified by an interest that is not the petitioner's. Laches protects those who have acted in reliance on the decision, which is why it is applied strictly in service and tender matters, why it is answered by an explanation rather than by a rule, and why relief is often granted prospectively rather than refused outright. Res judicata protects finality, and Daryao is the decision that matters because it rests the doctrine on public policy rather than on the terms of Section 11, which is what allowed it to be applied to a jurisdiction Section 11 does not reach, while preserving exceptions for every case in which the earlier proceeding never reached the merits. The most revealing feature of both is where they stop. Neither is applied with any rigour to habeas corpus: delay does not defeat it, because an unlawful detention is a wrong renewed every hour, and constructive res judicata does not bind it, because Lallubhai Jogibhai Patel holds that a detenu may raise a ground he did not raise before. The interests these doctrines protect, settled expectations and finality, are ones the common law has never been willing to weigh against personal liberty.

For full marks, cover: the question has two halves joined by "with special reference to writs", so organise it as origin and development first, then the writs in India, then the two together; under origin, the prerogative writs and their character, Bonham, Marbury and British India; under the constitutional perspective, Articles 13, 32, 226, 227 and 136; then each writ separately, since the question says "with special reference to writs"; then the judicial perspective, meaning how the courts have enlarged the writs beyond their English form; then the entrenchment of the jurisdiction and a critical note.

The claim that a court might control the legislature appears in Dr. Bonham's Case, (1610) 8 Co. Rep. 113b, where the College of Physicians fined and imprisoned Bonham for practising without its licence and kept half the fine, so that it was judge in its own cause; Coke CJ said that when an Act of Parliament is against common right and reason the common law will control it and adjudge such an Act to be void. England abandoned the claim after 1688; Marbury v. Madison, 5 US 137 (1803), adopted it, Marshall CJ holding that Section 13 of the Judiciary Act, 1789, in purporting to give the Supreme Court original jurisdiction to issue mandamus, exceeded Article III and was void, and that it is emphatically the province and duty of the judicial department to say what the law is.

In British India the writ jurisdiction came with the Supreme Courts at Calcutta, Madras and Bombay under the Regulating Act, 1773, and its successors, and passed to the High Courts under the Indian High Courts Act, 1861. It was confined to the Presidency towns, which is why the framers of the Constitution extended it to the whole country and to every High Court. The Government of India Act, 1935, added the second element by distributing legislative power in three lists, so that legislative competence became justiciable and the Federal Court decided such questions; what was missing was a bill of rights.

The five writs in India

Habeas corpus produces a detained person and tests the legality of the detention. Standing is relaxed so that a relative, a friend or a stranger may apply and a letter may be treated as a petition; the burden lies on the detainer; and legality is examined as at the date of the return. It lies against a private person as well as the State. In Sunil Batra (II) v. Delhi Administration, (1980) 3 SCC 488, a letter from a convict alleging that a warder had tortured a fellow prisoner to extort money from his visitors was converted into a habeas corpus petition, and the Court held that the writ protects a prisoner against illegality inside a lawful custody.

Prohibition stops an inferior court or tribunal proceeding beyond its jurisdiction, and is preventive; certiorari quashes a decision already made, and is corrective. Syed Yakoob v. K.S. Radhakrishnan, AIR 1964 SC 477, defines the scope of certiorari: errors of jurisdiction, breaches of natural justice and errors of law apparent on the face of the record are correctable, but a finding of fact, however erroneous, is not, the jurisdiction being supervisory and not appellate; and the error must be self evident and must not require elaborate argument. Radhey Shyam v. Chhabi Nath, (2015) 5 SCC 423, holds that certiorari does not lie against a judicial order of a civil court, which is reachable only under Article 227.

Quo warranto calls on a person to show by what authority he holds a public office of a substantive character created by statute or by the Constitution. Its distinguishing feature is that any person may apply, whether or not personally aggrieved, because the object is to remove a usurper rather than to redress a private wrong.

The judicial perspective: how the writs were enlarged

The Indian courts have taken the writs well beyond their English form, and four developments should be given.

The reach of the writs was extended by the expansion of Article 12. Ramana Dayaram Shetty v. International Airport Authority of India, (1979) 3 SCC 489, and Pradeep Kumar Biswas v. Indian Institute of Chemical Biology, (2002) 5 SCC 111, brought within the writ jurisdiction bodies that are not departments of government at all, the test being whether the body is financially, functionally and administratively dominated by or under the control of the Government.

The entrenchment, and the one failure

L. Chandra Kumar v. Union of India, (1997) 3 SCC 261, holds that judicial review under Articles 32, 226 and 227 is part of the basic structure and cannot be excluded even by constitutional amendment; it struck down Article 323A(2)(d) and Article 323B(3)(d) so far as they excluded that jurisdiction over tribunals. Madras Bar Association v. Union of India, 2025 INSC 1330, decided on 19 November 2025, applies the same principle, striking down provisions of the Tribunals Reforms Act, 2021, that re-enacted invalidated provisions and directing the Union to establish a National Tribunals Commission within four months.

Conclusion. The notion of judicial review reached India as a technique before it arrived as a doctrine. The technique was the prerogative writ, by which the King's Bench kept inferior jurisdictions within their authority, supervisory in nature and discretionary in exercise, and it came to the Presidency towns with the Supreme Courts of 1773 and passed to the High Courts in 1861. The doctrine, that a court may hold a law void for inconsistency with a superior law, was Coke's claim in Bonham, developed in Marbury, and it entered Indian law properly only with the Constitution, which supplied both a Part III to review against and an express conferment in Articles 13, 32, 226, 227 and 136. The writs are the point at which the two meet, and the Indian courts have transformed them: standing relaxed to admit a letter from a stranger, procedure remade to include commissioners of inquiry, relief extended to compensation, and the class of respondents widened to every body performing a public duty. What holds the whole structure together is L. Chandra Kumar, which placed the jurisdiction in the basic structure; and what shows why that mattered is ADM Jabalpur, where the writ survived intact and was worth nothing because the right behind it had been suspended.

The tension is that the same feature which makes discretion useful makes it dangerous. Lord Halsbury put the answer in Sharp v. Wakefield, [1891] AC 173: when it is said that something is to be done within the discretion of the authorities, that something is to be done according to the rules of reason and justice, not according to private opinion, according to law and not humour, and it must be regular, not arbitrary, vague and fanciful. In S.G. Jaisinghani v. Union of India, AIR 1967 SC 1427, the Supreme Court held that the absence of arbitrary power is the first essential of the rule of law, that discretion must be confined within clearly defined limits, and that a decision should be predictable so that the citizen knows where he is.

A court reviewing discretion therefore has a narrow path: it must prevent the range of choice being exceeded or abused, without making the choice itself. Every ground below is a technique for doing that.

Head one: failure to exercise the discretion

Sub-delegation. A power conferred on a named authority must be exercised by that authority, delegatus non potest delegare, subject to the practical relaxation in Carltona Ltd. v. Commissioners of Works, [1943] 2 All ER 560, that a Minister acts through departmental officials.

Head two: abuse of the discretion

Improper purpose. Padfield v. Minister of Agriculture, Fisheries and Food, [1968] AC 997: the Minister refused to refer a milk price complaint to a committee of investigation, partly because a favourable report would politically embarrass him. The House of Lords held that a discretion must be used to promote the policy and objects of the Act, which are to be determined by construing the Act as a whole, and ordered him to consider the complaint according to law. In State of Punjab v. Gurdial Singh, (1980) 2 SCC 471, land acquisition ostensibly for a grain market was pursued to injure a political opponent; Krishna Iyer J. held it a fraud on power, legal malice being the exercise of statutory power for purposes foreign to the statute.

Irrelevant considerations. The first limb of Wednesbury, [1948] 1 KB 223, requires an authority to take into account matters it is bound to consider and to exclude the irrelevant. In Ram Manohar Lohia v. State of Bihar, AIR 1966 SC 740, a detention order made to prevent a breach of "law and order" was quashed because the rule authorised detention only to prevent a breach of "public order", the Court describing law and order, public order and the security of the State as three concentric circles of decreasing width.

The first is the jurisdictional fact. Barium Chemicals Ltd. v. Company Law Board, AIR 1967 SC 295: an order for investigation into a company's affairs required the Board to be of opinion that specified circumstances existed. The Court held that although the formation of the opinion is subjective, the existence of the circumstances is a condition precedent and is objectively examinable, so that if there is no material on which a reasonable person could form the opinion the order falls. Rohtas Industries Ltd. v. S.D. Agarwal, AIR 1969 SC 707, confirmed it.

The second is the repudiation of Liversidge. In Liversidge v. Anderson, [1942] AC 206, the majority held that "if the Secretary of State has reasonable cause to believe" meant only that he must think he had, and that the grounds were not examinable. Lord Atkin dissented, saying he knew of only one authority for such a construction and quoting Humpty Dumpty's claim that a word means what he chooses it to mean, and that amid the clash of arms the laws are not silent. His view is now accepted; and in Khudiram Das v. State of West Bengal, (1975) 2 SCC 81, the Indian Court held that even under a statute resting on subjective satisfaction it may examine whether the satisfaction was reached on relevant material and whether the mind was applied.

Krishnadatt Awasthy v. State of M.P., 2025 INSC 126, decided on 29 January 2025, is the recent natural justice authority: appointments of fourteen Shiksha Karmi teachers, selected in 1998 by a committee containing their own close relatives, were cancelled by the Collector for nepotism without hearing them. The Court held that audi alteram partem* is the cornerstone of justice and that a defect at the initial stage cannot generally be cured at the appellate stage. The appeal succeeded and the appointments stood, the Court setting aside the ex parte cancellation. The case had reached three judges because a two judge Bench split on 4 April 2024, Maheshwari J. holding the selection vitiated by bias and Vishwanathan J. holding the cancellation vitiated by the denial of hearing, so it is a rare instance of the two limbs of natural justice pulling against each other, with the fair hearing limb prevailing.

K. Purushottam Reddy v. Union of India, 2025 INSC 894*, holds that the doctrine of legitimate expectation cannot be invoked to claim an enforceable right contrary to the constitutional arrangement, on a petition seeking an increase in the Assembly seats of Andhra Pradesh and Telangana in the teeth of Article 170(3).

An assessment

Two honest qualifications should close. The grounds are established almost entirely from the record, so an authority that recites the relevant considerations and gives intelligible reasons is largely secure whatever its actual motive; the requirement to give reasons, made general by S.N. Mukherjee v. Union of India, (1990) 4 SCC 594, therefore does more work than any substantive ground. And the intensity of review still turns on classification: after Om Kumar the same decision attracts primary review if characterised as a restriction on a fundamental right and secondary review if characterised as a service penalty, so the characterisation frequently decides the case.

For full marks, cover: why the distinction mattered before 1969, since a question asking whether two things are the same must first explain why anyone cared; the classical tests for classifying an action; A.K. Kraipak in full, with its facts, the two arguments made, the holding and the exact words about the dividing line, because the question is built on it; what survived the case, since the distinction is not wholly abolished and an answer that says it is loses marks; the consequences for certiorari, for reasons and for the standard of review; the later Indian development; and a reasoned answer to the question actually asked.

Why the distinction mattered

Before 1969 the classification decided three things. Whether natural justice applied at all: it applied to quasi judicial functions and not to administrative ones. Whether certiorari would issue: it issued only where there was a duty to act judicially. And whether reasons had to be given.

The English turn came with Ridge v. Baldwin, [1964] AC 40. Charles Ridge, Chief Constable of Brighton, was dismissed by the Watch Committee, without notice or hearing, under a power to dismiss any constable thought negligent in his duty or otherwise unfit. The House of Lords held the dismissal void, Lord Reid holding that the supposed requirement of a superadded duty to act judicially was a misreading, and that the duty to act judicially is an inference from the nature of the power to affect rights.

A.K. Kraipak v. Union of India, (1969) 2 SCC 262

The facts. The Indian Forest Service was to be constituted in Jammu and Kashmir, and officers of the State forest service were to be selected into it. A Special Selection Board was constituted, and one of its members was Naqishbund, who was then officiating as Chief Conservator of Forests in the State. Naqishbund was himself a candidate for selection into the Service, and he was in direct competition with several of the other candidates, in particular Basu, Baig and Kaul, who were senior to him in the State service and whose promotion to the substantive post of Chief Conservator would have affected his own position.

On the second argument it held that it was not necessary to establish actual bias, and that Naqishbund's withdrawal when his own case was considered did not cure the defect, because he sat when the claims of his rivals were considered and every mark he awarded or withheld against them affected his own relative position. The whole selection was therefore vitiated, including in respect of candidates against whom no personal interest could be shown, because the Board itself was improperly constituted.

What survived the case

This is the part that distinguishes a good answer, because Kraipak did not abolish the distinction; it removed its principal consequence. The classification still matters in four respects.

First, the content of natural justice varies with the function. Kraipak itself says that what rule applies depends on the framework of the law, the constitution of the body and the nature of the rights affected. A quasi judicial body determining a lis on evidence may owe an oral hearing, disclosure of documents and cross examination; an administrative authority may owe no more than notice of the substance and an opportunity to make written representations.

The later development

Maneka Gandhi v. Union of India, (1978) 1 SCC 248, completed what Kraipak began by making procedural fairness a constitutional requirement: the procedure under Article 21 must be right, just and fair, and not arbitrary, fanciful or oppressive, and audi alteram partem is read into a statute that does not exclude it expressly or by necessary implication. After that, the question is not what kind of function is being performed but whether rights are being affected.

Swadeshi Cotton Mills v. Union of India, (1981) 1 SCC 664, applied the same reasoning to a plainly administrative power, the takeover of the management of six textile undertakings under Section 18AA(1)(a) of the Industries (Development and Regulation) Act, 1951, holding that the rule may be moulded but not abrogated and that where urgency justifies a post-decisional hearing that hearing must be real.

Mohinder Singh Gill v. Chief Election Commissioner, (1978) 1 SCC 405, added two things that recur constantly. The first is that fairness is required of every action, judicial, quasi judicial, administrative or quasi administrative alike. The second, and the more useful in practice, is that an order must be judged by the reasons it bears on its face and cannot be supplemented by fresh reasons in an affidavit, the Court approving Gordhandas Bhanji and observing that an order is not like old wine, becoming better as it grows older. The Court also gave "civil consequences" a deliberately wide meaning, covering not merely property and personal rights but civil liberties, material deprivation and non pecuniary loss, which is what brings so much administrative action within the rule at all.

Union of India v. Tulsiram Patel, AIR 1985 SC 1416, supplies the other side and is the authority for the exceptions. Government servants dismissed without enquiry under the second proviso to Article 311(2) challenged their dismissals. The Constitution Bench held that where that proviso genuinely applies, natural justice stands excluded by the Constitution itself, and that the content of natural justice is what good conscience requires in the given situation, nothing more and nothing less. It is the case to cite whenever an answer turns to when a hearing may be dispensed with.

For full marks, cover: the two doctrines separately and then together, since the question pairs them and the pairing is not accidental; for bias, the maxim, the rationale that appearance matters independently of outcome, the three forms each with a worked case, the test applied in India as against the English formulations, the doctrine of necessity and waiver; for reasoned decision, why reasons matter, the three functions the Supreme Court has identified, the leading cases, what an adequate reason looks like, and the exceptions; then the connection, which is that reasons are the mechanism by which bias and every other defect become visible; and the 2025 authority.

The rule against bias

Nemo judex in causa sua, no one may be a judge in his own cause, is the first limb of natural justice. Its rationale is distinctive and must be stated, because it explains every peculiarity of the doctrine: the rule protects confidence in the decision, not the correctness of the decision. That is why a decision is set aside for bias even where it is shown to be right, and even where the decision maker is shown to have been in fact impartial.

Personal bias covers relationship, friendship, hostility or professional connection. Manak Lal v. Dr. Prem Chand, AIR 1957 SC 425: the chairman of a Bar Council disciplinary tribunal had earlier appeared as counsel for the complainant; the proceedings were quashed although no actual prejudice was shown, the Court holding that the test is whether there is a reasonable ground for believing that the member was likely to be biased. Ranjit Thakur v. Union of India, (1987) 4 SCC 611: a soldier who had complained against a superior officer was tried by a court martial in which that officer participated, and was sentenced to rigorous imprisonment and dismissal for refusing food; the Court quashed the proceedings, holding that the test is whether a reasonable person in the position of the party would have a reasonable apprehension of bias.

The test applied in India should be stated with care. England moved from a real likelihood of bias, through the "real danger" test of R v. Gough, [1993] AC 646, to the formulation in Porter v. Magill, [2002] 2 AC 357, that the question is whether a fair minded and informed observer would conclude that there was a real possibility of bias. Indian courts have not adopted a single verbal formula; the working test from Manak Lal and Ranjit Thakur is whether a reasonable person would entertain a reasonable apprehension of bias, which is closer to the reasonable suspicion end than the Gough test was and is more protective of the litigant. A vague or fanciful suspicion is not enough; the apprehension must be one a reasonable person would form on the facts.

Two qualifications. The doctrine of necessity allows a disqualified person to act where no one else is legally competent, since the alternative is that no decision is taken; in Election Commission of India v. Dr. Subramaniam Swamy, (1996) 4 SCC 104, the Supreme Court applied it where the Chief Election Commissioner was said to be biased and no other authority could tender the required opinion, devising a procedure by which he would act on the opinion of the other Commissioners. Waiver is possible where a party with knowledge of the disqualifying interest proceeds without objection, but knowledge must be established and the rule is applied cautiously.

Union of India v. Mohan Lal Capoor, (1974) 2 SCC 836, shows what an inadequate reason looks like. Officers were superseded for selection to the Indian Administrative Service with the expression "not found suitable" recorded against their names. The Supreme Court held that reasons are the links between the materials on which certain conclusions are based and the actual conclusions, that they must disclose how the mind was applied, and that "not found suitable" is a conclusion and not a reason at all.

Kranti Associates (P) Ltd. v. Masood Ahmed Khan, (2010) 9 SCC 496, summarises the modern law. An adjudicator must record reasons; reasons are the link between the material and the conclusion; recording of reasons excludes the chances of arbitrariness; reasons are the heartbeat of every conclusion; and reasons introduce transparency, which is a requirement of the rule of law. The Court cautioned against a mechanical recital, holding that reasons must be intelligible and must deal with the substantial points raised, and it noted the growing insistence on reasons in international human rights jurisprudence.

Reasons are the mechanism by which bias becomes visible. Bias is almost never admitted and is rarely provable directly; what a litigant has is the decision and its reasons. A decision that gives no reasons, or gives a formula, leaves the court with nothing but the surrounding circumstances, and courts have therefore been willing to infer from unexplained decisions what they could not find on the record. Lord Upjohn's observation in Padfield v. Minister of Agriculture, Fisheries and Food, [1968] AC 997, that if a Minister gives no reason in circumstances where the decision appears inexplicable the court may infer that he had none, is the general principle.

Conversely, the rule against bias is what makes reasons trustworthy. Reasons given by an interested decision maker are a rationalisation, not a justification; Ranjit Thakur makes the point from the other direction, holding that a strikingly disproportionate sentence is itself evidence of the mind that imposed it. A system that required reasons but tolerated interested decision makers would obtain elaborate reasons for predetermined results.

The Indian authorities the examiner expects

Three Indian decisions carry this branch of the subject, and an answer resting only on the English cases is incomplete. In State of Orissa v. Dr. (Miss) Binapani Dei, AIR 1967 SC 1269, an Assistant Surgeon in the Orissa Medical Service was compulsorily retired after the State unilaterally altered her recorded date of birth on an enquiry at which she had no real opportunity to meet the material against her. The Supreme Court set the order aside and held that even an administrative order which involves civil consequences must be made consistently with the rules of natural justice. That proposition preceded A.K. Kraipak by two years and is the true Indian starting point for everything said above.

Conclusion. The rule against bias and the duty to give reasons are the two halves of what makes an administrative decision defensible. Bias goes to the decision maker, and its distinctive feature is that appearance is enough: Dimes set aside a Lord Chancellor's decrees for a shareholding nobody thought had influenced him, Manak Lal and Ranjit Thakur fix the Indian test as reasonable apprehension rather than proof, and Kraipak and Gullapalli Nageswara Rao show it applied to selection boards and to departments hearing objections to their own schemes. The reasoned decision goes to the decision, and S.N. Mukherjee establishes it as a general requirement while Mohan Lal Capoor shows what will not do and Kranti Associates what will. Their significance for fair administration is that neither works alone. Without reasons, bias is unprovable and review is guesswork; without the rule against bias, reasons are the elaborate justification of a result already fixed. Taken together they produce the only thing a citizen can actually use: a decision by someone with no stake in it, accompanied by an account of why it was made, which is what makes it possible to tell whether the power was exercised for the purpose for which it was given.

The need

Five reasons are conventionally given and each should carry a sentence of explanation. Pressure on legislative time: a House that debated every technical detail would pass a fraction of the law required. Technicality: the tolerances of a drug, the design of a boiler, the classification of a security or the format of an electronic record cannot be settled on the floor of a House. Flexibility: conditions change faster than statutes can be amended, and a rule can be altered in weeks. Emergency: war, epidemic, famine and financial crisis require immediate action under powers conferred in advance. Experiment: a rule that proves unworkable can be withdrawn, where an Act cannot.

Two committee reports are the standard authority. The Committee on Ministers' Powers (Donoughmore), 1932, in England concluded that delegation is inevitable and that the safeguard lies in its control rather than its prohibition. In India the Fourteenth Report of the Law Commission (1958) and the Committees on Subordinate Legislation of both Houses have reached the same conclusion while criticising the drafting of enabling provisions.

The planned economy and the welfare State entrenched the practice. Industrial licensing under the Industries (Development and Regulation) Act, 1951, import and export control, foreign exchange regulation, land reform, price control and food distribution all operated through rules, orders and notifications rather than through statutes. The modern phase is regulatory: the Securities and Exchange Board of India, the Reserve Bank, the Telecom Regulatory Authority, the Competition Commission and the electricity regulatory commissions each make binding regulations, and the substantive law in those fields is now overwhelmingly regulatory rather than statutory.

The measure of the development is volume. In any modern legal system the quantity of subordinate legislation vastly exceeds the quantity of primary legislation, and India is no exception; the number of statutory instruments made in a year runs into thousands against a few dozen Acts. That disproportion is the fact from which every criticism starts.

The critical analysis

Four criticisms should be made, and each answered where there is an answer.

The doctrine of excessive delegation has produced few striking downs. Since Hamdard Dawakhana v. Union of India, AIR 1960 SC 554, the instances are rare, because a court that looks for a standard will usually find something in the scheme of the Act. The requirement of a policy has therefore become easy to satisfy, and the constitutional limit is less demanding in practice than it appears on paper.

The answer to all four is that the alternative is unavailable. A State charged by the Directive Principles with securing a living wage, distributing material resources to subserve the common good and providing public assistance cannot administer through primary legislation alone. The real question is the quality of control, which is the second limb of this question.

The grounds of judicial review

Ground one: the parent Act is invalid. If the enabling Act is beyond legislative competence or violates Part III, the rules made under it fall with it.

Ground three: substantive ultra vires. General Officer Commanding in Chief v. Dr. Subhash Chandra Yadav, (1988) 2 SCC 351, requires a rule to conform to the statute and to fall within the scope of the rule making power; Kunj Behari Lal Butail v. State of H.P., (2000) 3 SCC 40, struck down a rule that cut down an exemption the Act itself had granted, holding that a general power to make rules for carrying out the purposes of an Act does not permit a rule which travels beyond the Act or is inconsistent with it. Retrospectivity without authority falls here too, under Hukam Chand v. Union of India, AIR 1972 SC 2427.

Ground four: violation of the Constitution. Article 13(3)(a) makes a rule "law", so Article 13(2) applies directly, and a rule is void to the extent that it abridges a fundamental right, independently of whether it is within its parent Act.

The most recent demonstration that the control retains force is Madras Bar Association v. Union of India, 2025 INSC 1330, decided on 19 November 2025, striking down key provisions of the Tribunals Reforms Act, 2021, for giving the executive control over the appointment, tenure and service conditions of tribunal members and for re-enacting almost verbatim provisions already invalidated, which the Court held to be an impermissible legislative override, and directing the Union to establish a National Tribunals Commission within four months.

For full marks, cover: note (a) as the paper sets it, together, since this paper pairs the two doctrines that the other paper in this volume separates, and the pairing itself is the idea: a claim must be brought once and brought promptly; give each its rule, its authority and its exceptions, then the two things they have in common and the one place where both stop; note (b) with the definition, the procedural and substantive division, the sources, the Indian decisions, the distinction from promissory estoppel and the 2025 constitutional limit.

Daryao v. State of U.P., AIR 1961 SC 1457, extended the doctrine to writ proceedings. Petitions under Article 32 were filed after petitions on the same facts had been dismissed by the High Court under Article 226, and it was argued that a fundamental right cannot be defeated by a rule of procedure. The Supreme Court held that res judicata is not a technical rule but a rule of public policy founded on the need for finality, that it is therefore not confined to Section 11, and that a petition under Article 32 is barred where a petition under Article 226 on the same facts was dismissed on the merits after contest and no appeal was taken. The exceptions were carefully preserved: a dismissal in limine, or on the ground of laches or of an alternative remedy, does not bar, because nothing was adjudicated; a decision without jurisdiction binds nobody; and a change in the law or a continuing cause of action presents a fresh matter. Direct Recruit Class II Engineering Officers' Association v. State of Maharashtra, (1990) 2 SCC 715, applies constructive res judicata to service litigation.

What they have in common, and where both stop. Both protect an interest that is not the petitioner's: res judicata protects finality, laches protects reliance. Both are applied more strictly in service and property matters, where third party rights accumulate, than in matters of personal liberty. And both stop at habeas corpus. Delay does not defeat a habeas corpus petition, because an unlawful detention is a wrong renewed every hour and there are no settled expectations to protect; and in Lallubhai Jogibhai Patel v. Union of India, (1981) 2 SCC 427, the Supreme Court held that constructive res judicata does not apply to habeas corpus, so that a ground available but not taken earlier may be taken later, though a successive petition on the same ground is not entertained. That common exception is the most instructive feature of both doctrines: what is being weighed is never delay or repetition as such, but the interest on the other side, and personal liberty outweighs both.

Two distinctions complete the note. Res judicata is not estoppel, which proceeds from the conduct of a party rather than the judgment of a court; and it is not stare decisis, which binds courts as to a proposition of law rather than parties as to a matter decided.

Union of India v. Hindustan Development Corporation, (1993) 3 SCC 499, is the leading Indian exposition. The Railway Board changed its policy for procuring cast steel bogies, allocating part of its requirement to smaller manufacturers at a different rate; the established suppliers complained. Relief was refused, but the Court held that an expectation may arise from an express promise or an established practice, that it confers locus standi and is a ground for attacking a decision as arbitrary, but that it is not a right in the conventional sense, that mere disappointment founds no claim, and that it yields to an overriding public interest and to a bona fide change of policy. The Court warned against pressing the doctrine so far as to fetter the administration's power to change policy in the public interest.

Navjyoti Coop. Group Housing Society v. Union of India, (1992) 4 SCC 477, is the procedural form succeeding: housing societies allotted land for years on a seniority fixed by the date of registration were held entitled to be heard before the Delhi Development Authority applied a new criterion to them.

Its distinction from promissory estoppel is regularly examined. Promissory estoppel entered Indian public law through Motilal Padampat Sugar Mills v. State of U.P., (1979) 2 SCC 409, where a State assurance of a three year sales tax exemption, acted upon by establishing a vanaspati factory, was held binding notwithstanding the absence of consideration or compliance with Article 299; it requires a clear promise and detrimental reliance. Legitimate expectation requires neither: a settled practice suffices, and no change of position is necessary. Promissory estoppel ordinarily yields the benefit; legitimate expectation ordinarily yields only a hearing. The Supreme Court has emphasised that legitimate expectation belongs to public law and cannot govern private dealings, where estoppel holds the field.

The Indian authorities on res judicata in writ proceedings

Two decisions complete what Daryao began, and both are regularly set. In Devilal Modi v. Sales Tax Officer, Ratlam, AIR 1965 SC 1150, an assessment was challenged under Article 226 and the petition failed up to the Supreme Court. A second petition was then brought against the same assessment, raising grounds that had been available but had not been taken the first time. The Court held it barred by constructive res judicata, reasoning that if the rule were otherwise a litigant could file successive petitions taking one or two points at a time, which would defeat the public policy on which res judicata rests and would expose the other side to indefinite harassment. It is the leading authority for applying constructive res judicata to the writ jurisdiction, and it does so notwithstanding that Section 11 of the Code does not in terms apply.

In State of M.P. v. Nandlal Jaiswal, (1986) 4 SCC 566, the petitioners challenged the grant of distillery licences, but did so only after standing by while the successful applicants acted on the grants and invested substantial sums in setting up their plants. The Supreme Court refused relief, holding that the jurisdiction under Article 226 is discretionary and equitable, that a petitioner guilty of unexplained delay cannot demand it as of right, and that the court will not assist a person who sleeps over his rights while others alter their position on the faith of the decision. That is the rationale for the whole doctrine stated in a single case.

In Chennai Metropolitan Water Supply and Sewerage Board v. T.T. Murali Babu, (2014) 4 SCC 108, the Court held that delay is not a mere technicality: it reflects on the bona fides of the claim, the writ court must examine whether the explanation offered is satisfactory rather than apply any fixed period, and a stale claim is not to be revived merely because it may have substance. Together the two supply what the general statement of the doctrine cannot, which is an Indian authority for each half of it.

For full marks, cover: for the finality clause, what it says, what it achieves, what it cannot achieve and why, the distinction from the ouster and conclusive evidence clauses, the decisions that settle its treatment, and an honest statement of the practical effect, since the marks lie in saying what such a clause does achieve rather than dismissing it; for the French legal system, that the question means the administrative side of it, the two orders of courts, the Conseil d'Etat, the grounds of annulment, State liability from Blanco, Dicey's misdescription, and the comparison with India.

(a) Finality clause

A finality clause provides that the decision of a specified authority "shall be final", or "final and conclusive", or "shall not be liable to be questioned". Legislatures insert them to bring proceedings to an end within the statutory scheme, and their natural meaning is that no further appeal lies.

What survives such a clause is the supervisory jurisdiction, not an appellate one, and its scope is fixed by Syed Yakoob v. K.S. Radhakrishnan, AIR 1964 SC 477: certiorari corrects errors of jurisdiction and breaches of natural justice, and an error of law apparent on the face of the record, but a finding of fact, however erroneous, cannot be interfered with; and the error of law must be self evident and must not require an elaborate argument to establish.

Kihoto Hollohan v. Zachillhu, 1992 Supp (2) SCC 651, is the best illustration available, because there the finality clause was in the Constitution itself. Paragraph 6(1) of the Tenth Schedule provides that the decision of the Speaker or Chairman on a question of disqualification for defection shall be final, and paragraph 7 barred the jurisdiction of all courts including under Articles 136, 226 and 227. The Supreme Court held paragraph 7 invalid for want of ratification by the States under the proviso to Article 368(2); and it held that the Speaker deciding such a question acts as a tribunal, so that the finality in paragraph 6(1) does not exclude judicial review on the grounds of mala fides, perversity, violation of a constitutional mandate and violation of natural justice, though the Court will not interfere at an interlocutory stage.

(b) The French legal system

The question, set in a paper on judicial control of the administrative process, means the French system of administrative justice, droit administratif. Its two defining features are a distinct body of substantive law governing the relations between administration and citizen, and a distinct hierarchy of courts to apply it.

The institutions. At first instance sit the tribunaux administratifs; above them the cours administratives d'appel; and at the apex the Conseil d'Etat, which is both the supreme administrative court and the standing adviser of the Government on draft legislation and decrees. The Tribunal des Conflits, composed of members of both orders, decides whether a case belongs to the ordinary or the administrative jurisdiction.

The history explains the structure. The revolutionary law of 16 to 24 August 1790 forbade the ordinary courts to disturb the operations of administrative bodies, the framers remembering how the pre-revolutionary parlements had used judicial power to obstruct royal reform. Judicial control was therefore relocated rather than abolished. The Conseil d'Etat, created in 1799, at first exercised only justice retenue, proposing decisions for the head of State; the law of 24 May 1872 conferred justice déléguée, the power to decide in its own name.

State liability is wider than at common law. The Blanco decision of the Tribunal des Conflits in 1873, arising from a child injured by a wagon of a State tobacco factory at Bordeaux, held that the liability of the State for damage caused by those it employs in the public service is neither general nor absolute and is governed by special rules distinct from the Civil Code. From it grew faute de service, fault of the service, under which the State is liable for the defective functioning of a service without the claimant having to identify the officer at fault.

Dicey's misdescription must be stated and corrected. In An Introduction to the Study of the Law of the Constitution (1885) he made equality before the ordinary law administered by the ordinary courts the second meaning of the rule of law, and treated droit administratif as its antithesis, on the view that it gave officials special privileges and removed them from the ordinary courts. He was wrong, and acknowledged as much in later editions: the Conseil d'Etat is an independent court, its procedure is quick and cheap, its members understand the administration they control, and its record of annulment is at least as good as that of the English courts. The damage was done in England and in India, by delaying acceptance that a distinct administrative law and specialist adjudication were legitimate at all.

Conclusion. The two notes describe an attempt to limit judicial control and a system that organises it differently. A finality clause is the legislature's cheapest instrument: it ends the appeal and leaves a wrong decision standing, and it obliges the challenger to prove something worse than error. What it cannot do is displace a jurisdiction the Constitution confers, which is Bharat Bank; and where a clause goes further and purports to oust the court altogether, Anisminic holds that it protects error and never nullity, while L. Chandra Kumar places the jurisdiction in the basic structure and Kihoto Hollohan shows even a constitutional finality clause read down to leave mala fides, perversity and natural justice examinable. The French system shows the alternative arrangement: a separate order of courts headed by the Conseil d'Etat, an inquisitorial procedure suited to a citizen with no documents, a cheap action for annulment on grounds almost identical to our own, and a law of State liability developed from Blanco that is wider than the common law produced unaided. India has taken the unitary structure and borrowed only the specialist forum, and it has fixed the boundary exactly where the French model would begin: tribunals may specialise, but they may never be placed beyond the constitutional courts.