LLM Group 1 Constitutional and Administrative Law Administrative Process Nature and Scope 2015 Question Paper with Solutions

(a) Role of Civil Services

The civil service is the permanent executive. Ministers decide policy and leave office; the services advise on that policy, translate it into rules and orders, run the machinery that delivers it, and remain. Part XIV of the Constitution, Articles 308 to 323, is built on that division. Article 309 allows the appropriate legislature to regulate recruitment and conditions of service and, until it does, allows the President or Governor to make rules, which is why most service law in India is subordinate legislation rather than statute. Article 310 retains the doctrine of pleasure, and Article 311 cuts it down by two guarantees: no dismissal or removal by an authority subordinate to the appointing authority, and no dismissal, removal or reduction in rank without an inquiry in which the servant is informed of the charges and given a reasonable opportunity of being heard.

What the service actually does is wider than implementation. It supplies the institutional memory and the technical advice on which a minister depends; it drafts the delegated legislation under which most Indian regulation is actually carried on; it exercises statutory discretion in thousands of individual decisions daily, from a licence to a compassionate appointment; it adjudicates, because a large body of first instance decision making under Indian statutes is entrusted to officers acting quasi-judicially; and it collects the revenue and delivers the services. In the language of this syllabus, the civil service is the administrative process for most citizens most of the time.

Prakash Singh v. Union of India, (2006) 8 SCC 1 does the same work for the police, which is the largest and least insulated part of the field administration. The Court directed the constitution of State Security Commissions, the selection of the Director General of Police from a panel prepared by the Union Public Service Commission with a minimum two-year tenure, a minimum tenure for operational officers, the separation of investigation from law and order, and Police Establishment Boards and Police Complaints Authorities. Compliance is partial two decades later, and the gap between direction and compliance is itself the most useful thing a candidate can say about civil service reform in India.

(b) Regulation and De-regulation Policy in India

India has run three regulatory regimes in seventy-five years and the second is still being dismantled. The first, from 1947 to 1991, was the licence and permit system: the Industries (Development and Regulation) Act, 1951 required a licence to establish, expand or relocate an industrial undertaking; the Monopolies and Restrictive Trade Practices Act, 1969 required large houses to obtain approval for expansion; and the Foreign Exchange Regulation Act, 1973 treated every foreign exchange transaction as prohibited unless permitted. The instrument throughout was prior permission granted case by case in the discretion of an officer, which is the form of regulation most productive of delay, rent seeking and litigation.

That is a transfer of legislative, executive and judicial power to a single body, and Indian courts have accepted it while insisting on standards. Clariant International Ltd. v. SEBI, (2004) 8 SCC 524 treated the regulator's determination as an adjudicatory function attracting the ordinary controls of reasons and fairness. Competition Commission of India v. Steel Authority of India Ltd., (2010) 10 SCC 744 held that the Commission's direction for investigation under Section 26(1) is administrative rather than adjudicatory, so no hearing is required at that stage, but reasons must be recorded, which is a good illustration of how the courts calibrate procedure to function. Competition Commission of India v. Bharti Airtel Ltd., (2019) 2 SCC 521 held that where a sectoral regulator and the competition regulator both have jurisdiction, the sectoral regulator's findings on the technical questions come first, which is the Indian answer to regulatory overlap.

Deregulation does not suspend administrative law, and the Supreme Court has said so. Centre for Public Interest Litigation v. Union of India, (2012) 3 SCC 1, the 2G licences case, cancelled 122 licences allotted on a first-come-first-served basis because the method was arbitrary and the resource was public. The advisory opinion in Natural Resources Allocation, In re Special Reference No. 1 of 2012, (2012) 10 SCC 1 then corrected the wider reading: auction is not the only constitutionally permissible method of alienating natural resources, and the test is whether the method chosen serves the common good and is not arbitrary. Read together, the two decisions state the constitutional standard for the deregulated State: government may choose to withdraw, but the manner of withdrawal, being the disposal of public property, remains subject to Article 14.

Constitutional foundation. The Constitution (Forty-second Amendment) Act, 1976 inserted Articles 323A and 323B. Article 323A allows Parliament to create administrative tribunals for service matters, and the Administrative Tribunals Act, 1985 followed. Article 323B allows the appropriate legislature to create tribunals for the enumerated subjects, including taxation, foreign exchange, industrial and labour disputes, land reform, ceiling on urban property, elections and essential goods.

The constitutional case law runs in one direction. S.P. Sampath Kumar v. Union of India, (1987) 1 SCC 124 upheld the exclusion of High Court jurisdiction over service matters on the footing that the tribunal was an effective substitute, and directed changes to the mode of appointment to make it so. L. Chandra Kumar v. Union of India, (1997) 3 SCC 261, a bench of seven, went the other way on the essential point: the power of judicial review vested in the High Courts under Article 226 and 227 and in the Supreme Court under Article 32 is part of the basic structure, and clause 2(d) of Article 323A and clause 3(d) of Article 323B, to the extent that they excluded it, are unconstitutional. Tribunals are supplemental, not substitutional; their decisions are subject to scrutiny before a Division Bench of the High Court within whose jurisdiction the tribunal falls; and a litigant may not approach the Supreme Court directly.

The most recent decision is the most important and it post-dates every paper in this folder. In Madras Bar Association v. Union of India, decided on 19 November 2025, the Supreme Court struck down the core provisions of the Tribunals Reforms Act, 2021 dealing with appointment, tenure and service conditions, holding that Parliament had re-enacted provisions already held unconstitutional in the earlier Madras Bar Association decisions of 2020 and 2021 without curing the defects, and that this violated judicial independence and the separation of powers. The Court directed the Union to constitute a National Tribunals Commission within four months, a reform recommended since R. Gandhi in 2010 and repeated in Rojer Mathew.

Assessment. Tribunals now decide more disputes in India than the High Courts, and the constitutional question is no longer whether they may exist but on what terms. The settled terms are: judicial review by the High Court cannot be ousted; the selection process must be dominated by the judiciary; tenure must be long enough that a member is not seeking the executive's favour for reappointment; and the parent ministry must not be the administrative controller of the tribunal that hears appeals against it. What is still missing is the institution that would enforce those terms without fresh litigation every few years, and that is exactly what the National Tribunals Commission directed in November 2025 is meant to be.

For full marks, cover: a stated basis of comparison rather than five country descriptions in a row; each system's source of administrative law, the institution that controls the administration, the grounds and intensity of review, the remedy and the position on State liability; the two great families, the common law single-hierarchy model and the continental separate administrative jurisdiction; the leading authority in each system, worked and not merely named; a comparative table; and a critical conclusion on where the systems are converging and where they have just moved apart. The single most important point of currency is that the United States overruled Chevron in June 2024, so any answer describing American law as a system of deference to agency interpretation is now out of date.

The modern law grew out of the prerogative writs, and three decisions built it. Ridge v. Baldwin, [1964] AC 40 revived natural justice after decades in which it had been confined to bodies with a duty to act judicially: a chief constable dismissed without notice or hearing had his dismissal declared void, and the House of Lords held the duty to act fairly attaches to the effect of a decision on the individual, not to the label on the body. Anisminic Ltd. v. Foreign Compensation Commission, [1969] 2 AC 147 held that a determination reached on a wrong view of the law is a nullity and therefore not a "determination" protected by an ouster clause, which collapsed the distinction between jurisdictional and non-jurisdictional error and left ouster clauses very little to do. Council of Civil Service Unions v. Minister for the Civil Service, [1985] AC 374, the GCHQ case, held that prerogative power is reviewable by reference to its subject matter rather than its source, and Lord Diplock classified the grounds as illegality, irrationality and procedural impropriety, adding that proportionality might in time be added.

Delegation is controlled by the intelligible principle test, stated in J.W. Hampton, Jr. & Co. v. United States, 276 US 394 (1928) and applied to strike down statutes only twice, in Panama Refining Co. v. Ryan, 293 US 388 (1935) and A.L.A. Schechter Poultry Corp. v. United States, 295 US 495 (1935). Immigration and Naturalization Service v. Chadha, 462 US 919 (1983) struck down the legislative veto as a violation of bicameralism and presentment, which is the clearest modern illustration of separation of powers applied to the administrative State.

France

France has a separate body of administrative law applied by a separate hierarchy of administrative courts, and the arrangement is older than the Republic's present constitution. The Law of 16 to 24 August 1790 and the decree of 16 Fructidor Year III forbade the ordinary courts from interfering with the administration, and the Conseil d'État, created in 1799, was given the resulting work. Its jurisdiction became judicial rather than advisory in substance in 1872, and the modern structure is three tiered: tribunaux administratifs at first instance, cours administratives d'appel, and the Conseil d'État at the apex, with a Tribunal des conflits to allocate jurisdiction between the two orders.

The remedies are the recours pour excès de pouvoir, an action to annul an administrative act, open on the grounds of lack of jurisdiction, defect of form, misuse of power (détournement de pouvoir) and violation of law, and the recours de plein contentieux, in which the court may substitute its own decision and award damages. Benjamin (CE, 19 May 1933) applied what is in substance a proportionality test to a ban on a public lecture, holding that the mayor could not prohibit a meeting where a lesser measure would preserve order. Nicolo (CE, 20 October 1989) accepted that a treaty prevails over a later statute. The Law of 30 June 2000 created the référé-liberté, under which a judge must rule within forty-eight hours on a serious and manifestly unlawful interference with a fundamental freedom, and the Code des relations entre le public et l'administration, 2016 codified the citizen's procedural rights.

Dicey's charge was therefore the reverse of the truth. Droit administratif is not a shield for officials but a body of rules more favourable to the citizen in important respects than the common law was: liability without fault was recognised in Cames (CE, 1895) long before English law reached the same result by statute, and the Conseil d'État reviews for erreur manifeste d'appréciation, a control of evaluation that English law only reached through Wednesbury.

Germany is also the source of the proportionality doctrine that India now uses. The Apotheken-Urteil or Pharmacy judgment (BVerfGE 7, 377, decided 11 June 1958) struck down a Bavarian law restricting new pharmacies and set out the structured test of legitimate aim, suitability, necessity and proportionality in the narrow sense, which is the ancestor of the four-prong test adopted in Modern Dental College and Research Centre v. State of Madhya Pradesh, (2016) 7 SCC 353 and applied in K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1.

Two Indian features have no counterpart in the other four. The first is the absence of a general Administrative Procedure Act, so notice, comment, reasons and hearing rest on case law rather than statute, which the Law Commission's Fourteenth Report noticed as long ago as 1958. The second is public law compensation: from Rudul Sah v. State of Bihar, (1983) 4 SCC 141, where a man was held in jail for fourteen years after his acquittal, and Nilabati Behera v. State of Orissa, (1993) 2 SCC 746, Indian courts award damages in writ proceedings for the violation of fundamental rights, a remedy the English courts declined to create and the French reached through their own liability rules. Tort liability otherwise still labours under the sovereign function distinction of Kasturilal Ralia Ram Jain v. State of Uttar Pradesh, AIR 1965 SC 1039, narrowed but not overruled in N. Nagendra Rao and Co. v. State of Andhra Pradesh, (1994) 6 SCC 205.

The comparison in a table

Conclusion. The two families remain distinct in institutional design, one hierarchy in the common law world and two on the continent, but they have converged remarkably on substance. Proportionality began in Germany, was received into European Union and Convention law, entered the United Kingdom through the Human Rights Act and India through Om Kumar and Modern Dental College, and now supplies the working standard wherever a fundamental right is engaged. Codified procedure has spread the same way, from the American APA of 1946 to the German Act of 1976 to the French Code of 2016, and India is the outlier that still has none. Against that convergence stands the most important recent development, which runs the other way: by overruling Chevron in Loper Bright in June 2024, the United States has repudiated deference on questions of law and returned statutory interpretation to the courts. For an Indian student the comparison yields one practical lesson. India has borrowed its grounds of review from England, its proportionality from Germany through England, and its written guarantee of review from the logic of a written constitution; what it has not borrowed, and most needs, is the one American and German idea that regulates the administration before the dispute arises rather than after, a general statute of administrative procedure.

First, the absolute supremacy of regular law as opposed to arbitrary power. No one may be punished except for a distinct breach of law established before the ordinary courts, and wide discretionary authority in the government is inconsistent with the rule of law.

Second, equality before the law, meaning the equal subjection of all classes to the ordinary law administered by the ordinary courts, so that officials have no special exemption and no special tribunals.

Third, that the general principles of the constitution are the result of judicial decisions determining the rights of private persons in particular cases, rather than of a written guarantee, so that in England the constitution is the consequence of the ordinary law and not its source.

The modern reformulations

Later writers replaced Dicey's propositions with lists of qualities that a legal system must have. Joseph Raz's formulation is the most rigorous: laws should be prospective, open and clear; relatively stable; the making of particular orders should be guided by open, stable, clear and general rules; the independence of the judiciary must be guaranteed; natural justice must be observed; the courts should have review powers; they should be accessible; and the discretion of crime preventing agencies should not be allowed to pervert the law. Lord Bingham's eight sub-rules add substance, requiring accessibility, that questions of legal right be resolved by law and not discretion, equality before the law, the lawful and reasonable exercise of power, protection of fundamental human rights, means of resolving civil disputes without prohibitive cost, fair adjudicative procedures, and compliance by the State with its international obligations. The Delhi Declaration of the International Commission of Jurists, 1959 made the same move in a different register, declaring that the rule of law includes the creation of the social, economic and cultural conditions in which a person may live in dignity, which is the formulation Indian courts have found most congenial.

The decisions

S.G. Jaisinghani v. Union of India, AIR 1967 SC 1427* contains the most quoted Indian sentence on the subject and it should be used as the spine of the answer. Dealing with a seniority quota rule for income tax officers, the Court said that the absence of arbitrary power is the first essential of the rule of law, that where discretion is conferred it must be confined within clearly defined limits, and that a decision should be predictable and the citizen should know where he is. That is Dicey's first proposition rebuilt for a State that must have discretion.

Kesavananda Bharati v. State of Kerala, (1973) 4 SCC 225 placed the rule of law beyond the reach of the amending power by holding that Article 368 does not permit the destruction of the basic structure, and several judges named the rule of law among the features so protected.

A.D.M. Jabalpur v. Shivkant Shukla, (1976) 2 SCC 521 is the decision every answer must confront, because it is where the doctrine failed. During the Emergency the majority held that on the suspension of Article 21 no person had any locus standi to move a writ of habeas corpus, so a detenu could not complain even that his detention was without authority of law or in bad faith. Khanna J. dissented alone, holding that Article 21 is not the sole repository of the right to life and personal liberty and that the State has no power to deprive a person of life or liberty without the authority of law even in an Emergency. He was superseded for the Chief Justiceship. The dissent has since been vindicated twice over: the Constitution (Forty-fourth Amendment) Act, 1978 provided that the right to move a court for the enforcement of Articles 20 and 21 can never be suspended, and in K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1 a bench of nine expressly overruled ADM Jabalpur, holding that its majority opinion was seriously flawed and that natural or inalienable rights are not the gift of the State.

The limits

Three limits must be stated or the answer reads as an advertisement. First, equality before the law coexists with constitutional immunities, in Article 361 for the President and Governors and in Articles 105 and 194 for legislators, and with practical inequality of access, which is why Bingham's requirement of resolution without prohibitive cost matters more in India than any other of his sub-rules. Second, the rule of law is a formal virtue: a clear, prospective and equally applied law may still be unjust, which is why Indian courts have preferred the substantive Delhi Declaration version. Third, and most concretely, the rule of law fails when adjudication is too slow to matter, and a system with tens of millions of pending cases delivers, in a great many disputes, a right without a remedy.

Three of these decisions worked in full

A proposition is only as good as the facts that produced it, and three of the cases above repay being set out.

Dwarka Prasad Laxmi Narain v. State of Uttar Pradesh, AIR 1954 SC 224. The Uttar Pradesh Coal Control Order required a licence to deal in coal, and clause 4(3) allowed the State Coal Controller to exempt any person from the licensing requirement in his discretion. The petitioners, coal dealers, challenged it. The Court struck the clause down as an unreasonable restriction on the freedom of trade under Article 19(1)(g), holding that a provision which leaves it entirely to the officer to decide who shall be subject to the law and who shall not, without any standard to control the choice, cannot be a reasonable restriction. This is Jaisinghani applied to a rule rather than to an order, and it is the cleanest Indian illustration that in India an unguided discretion is a defect in the law itself.

Conclusion. The rule of law began as Dicey's three propositions about England, and it survives none of them intact: discretion is now unavoidable, officials are subject to special tribunals, and India's constitutional rights are written rather than judge made. What survives, and what the decisions enforce, is the core idea in Jaisinghani, that arbitrary power is the enemy and confined, guided, reviewable power is the answer. Indian courts have given that idea three concrete forms: it is part of the basic structure, so it cannot be amended away, as Kesavananda and Indira Nehru Gandhi established; it is enforceable through Article 14 as a prohibition of arbitrariness, as Royappa, Maneka Gandhi and Shayara Bano established; and it is not suspended by an emergency, which took the Forty-fourth Amendment and forty-one years to settle, and was finally settled when Puttaswamy overruled ADM Jabalpur. The lesson of that sequence is the one worth ending on: the rule of law is only as strong as the institution willing to enforce it against the government of the day.

What administrative discretion is, and why it exists

Discretion is the power to choose between courses of action, all of which are lawful. A statute that says a licensing authority "may" grant a licence, or that an officer may act if he "is satisfied" or "has reason to believe", or that a measure may be taken "in the public interest", confers discretion. It is unavoidable in a State that regulates the economy, licenses activity, distributes benefits and manages resources, because no legislature can foresee every fact situation, many decisions require technical judgment, and individualised justice sometimes demands departure from a rule that would work unfairly in a particular case.

The classical definition is Lord Halsbury's in Sharp v. Wakefield, [1891] AC 173. Discretion, he said, means that something is to be done according to the rules of reason and justice and not according to private opinion, according to law and not humour, and it is to be not arbitrary, vague and fanciful but legal and regular. That sentence contains the whole of modern administrative law on the subject: discretion is a legal power with legal limits, and the court's task is to find the limits, not to substitute its own choice.

Acting under dictation. Commissioner of Police, Bombay v. Gordhandas Bhanji, AIR 1952 SC 16 is the classic. A licence to build a cinema was granted by the Commissioner and then cancelled by him on the direction of the Government. The Court held the cancellation void: the statute gave the discretion to the Commissioner, and an order purporting to be his but in truth the Government's is no order at all. The case also settled that public orders must be judged by what they say on their face, and cannot be supported by reasons supplied later in an affidavit. Purtabpur Co. Ltd. v. Cane Commissioner, Bihar, (1970) 1 SCC 139 applied the same principle where the Cane Commissioner acted on the Chief Minister's direction.

Self-imposed fetters. An authority may adopt a general policy to guide the exercise of discretion, but it may not convert the policy into an inflexible rule that forecloses consideration of the individual case. The distinction was drawn in Shri Rama Sugar Industries Ltd. v. State of Andhra Pradesh, (1974) 1 SCC 534, where the majority upheld a State policy of granting purchase tax exemption only to cooperative sugar factories, holding that an authority with statutory discretion may adopt a general policy consistent with the statute; the qualification for which the case is cited is that it must not then shut its ears to the individual application.

Irrelevant considerations and no material. Barium Chemicals Ltd. v. Company Law Board, AIR 1967 SC 295 is the leading Indian case on subjective satisfaction. The statute allowed an investigation where the Board was "of the opinion" that circumstances suggested fraud or mismanagement. The Court held that even a subjectively worded power is reviewable: the existence of circumstances is a condition precedent, and if there is no material, or the material is extraneous, or no reasonable person could form the opinion on it, the order falls. Rohtas Industries Ltd. v. S.D. Agarwal, (1969) 1 SCC 325 applied it.

Unreasonableness. Associated Provincial Picture Houses Ltd. v. Wednesbury Corporation, [1948] 1 KB 223 allows interference where a decision is so unreasonable that no reasonable authority could ever have come to it. Tata Cellular v. Union of India, (1994) 6 SCC 651 received the standard in India in the context of tender evaluation, holding that the court reviews the decision-making process and not the decision, and warning against sitting in appeal on technical judgments.

Structuring and confining discretion

K.C. Davis's programme is the standard answer and it is worth stating as a programme. Discretion should be confined by fixing its boundaries in the statute, structured by open plans, published policy statements, rules, findings of fact, stated reasons, open precedents and fair informal procedure, and checked by supervision within the administration and review outside it. Indian law has adopted the structuring instruments piecemeal through case law: reasons through Mohan Lal Capoor and S.N. Mukherjee, guidelines through Jaisinghani, publication through Harla v. State of Rajasthan, AIR 1951 SC 467, and openness through the Right to Information Act, 2005. What it has not adopted is a general statute requiring them, which is why each has had to be litigated.

The impact of technological development

Technology has changed discretion in four ways, and only the first is the one most answers mention.

First, it has removed discretion from the point of delivery. Where a decision can be reduced to rules, the rules can be coded and the officer's choice disappears. Faceless assessment under Section 144B of the Income-tax Act, 1961, introduced with effect from 1 April 2021, allocates a return to a randomly selected assessment unit anywhere in the country, removes the taxpayer's contact with a named officer, and routes review through separate units; the scheme has been carried forward into the Income-tax Act, 2025, which replaced the 1961 Act on 1 April 2026. Direct benefit transfer, the Government e-Marketplace for procurement, automated allotment in admissions and online registration under a great many statutes do the same thing. The gain is real: the commonest form of arbitrariness in Indian administration was not a perverse decision but a decision delayed, hidden or sold, and a system with no counter has no queue to be bought.

Fourth, technology has created a new subject matter of discretion, and the courts have had to build standards for it. K.S. Puttaswamy (Aadhaar) v. Union of India, (2019) 1 SCC 1 upheld Section 7 of the Aadhaar Act on proportionality reasoning while striking down Section 57, which had allowed private bodies to demand authentication, and the exclusion errors argued in that case are the clearest illustration of what goes wrong when eligibility is decided by a machine. Shreya Singhal v. Union of India, (2015) 5 SCC 1 struck down Section 66A of the Information Technology Act, 2000 for vagueness, which is a discretion holding: a provision whose terms allow an officer to decide for himself what is "grossly offensive" confers precisely the unconfined power Jaisinghani forbids; the Court upheld Section 69A because it is confined by grounds and hedged with procedure. Anuradha Bhasin required shutdown orders to be published, for the plain reason that an unpublished order cannot be challenged.

Conclusion. Administrative discretion is not an evil to be eliminated but a necessity to be governed, and the law's method has been consistent since Sharp v. Wakefield: insist that the power be used for its purpose, on relevant material, by the authority named, after hearing the person affected, and with reasons that expose the chain of reasoning to review. Technology has kept that method's ends and broken several of its assumptions. It has genuinely reduced the petty discretion that produced most Indian corruption, and it has concentrated the significant discretion in design decisions taken once and applied at scale, where there is no named officer to answer, no contemporaneous record of the choice, and no reasons in the sense Mohan Lal Capoor meant. The next stage of Indian administrative law will be worked out on that ground, and the instruments already exist: proportionality as the standard of review, the duty to give reasons as the discipline, and publication as the precondition of both. What is missing is a statute that requires them before the litigation rather than after it.

What is being controlled

Delegated legislation is law made by an authority other than the legislature under power conferred by the legislature. In India it accounts for far more law by volume than the statute book: rules, regulations, orders, bye-laws, notifications, schemes and directions. The reasons are those the Committee on Ministers' Powers gave in England in 1932 and every Indian court has repeated since: pressure on legislative time, the technicality of modern subject matter, the need for flexibility and for rapid amendment, the demands of emergency, and the value of experimentation. Krishna Iyer J. put it at its bluntest in Avinder Singh v. State of Punjab, (1979) 1 SCC 137, saying that delegation is a necessary evil, that the complexities of modern administration are so bafflingly intricate that Parliament cannot legislate on every detail, and that the legislature is not an ideal instrument for the purpose.

Stage two is laying. Most Indian statutes require rules to be laid before Parliament, and three forms are used. Simple laying requires no more than that the rules be placed before the House. Laying subject to negative resolution allows the House to annul or modify them within a stated period, commonly thirty days over one or more sessions, without affecting the validity of anything already done. Laying subject to affirmative resolution requires the approval of the House before the rules take effect, and is used rarely, generally where the power is unusually wide.

The legal effect of a failure to lay is limited, and this is where the marks are. In Atlas Cycle Industries Ltd. v. State of Haryana, (1979) 2 SCC 196 the Supreme Court held that the laying requirement in Section 3(6) of the Essential Commodities Act, 1955 was directory and not mandatory, so an order not laid before Parliament was not thereby invalid; the Court reasoned that the provision contained no consequence for non-compliance and no provision for annulment. Jan Mohammad Noor Mohammad Begban v. State of Gujarat, AIR 1966 SC 385 is to the same effect. The result is that the ordinary form of laying in Indian statutes is an obligation without a sanction.

Judicial control: substantive ultra vires

Delegated legislation may be challenged on five substantive grounds.

That the parent Act is itself unconstitutional, including for excessive delegation, in which case everything made under it falls with it, as in Hamdard Dawakhana.

That the rule travels beyond the enabling provision. Kunj Behari Lal Butail v. State of Himachal Pradesh, (2000) 3 SCC 40 struck down a rule made under a general rule-making power because it dealt with a matter on which the Act itself was silent, holding that a general power to make rules to carry out the purposes of the Act does not enable the making of a rule that goes beyond the Act or deals with a field the Act has not entered. State of Karnataka v. H. Ganesh Kamath, (1983) 2 SCC 402 is to the same effect: a rule cannot supplant a provision of the parent Act, only supplement it.

That the rule conflicts with the parent Act or with another statute, which is a simple question of construction and the commonest ground in practice.

Judicial control: procedural ultra vires

Where the parent Act prescribes a procedure for making the rules, non-compliance may invalidate them, and the court asks whether the requirement is mandatory or directory.

Publication is the leading example, and it is a rule-of-law requirement rather than a technicality. Harla v. State of Rajasthan, AIR 1951 SC 467 set aside a conviction under a law that had never been published, holding that natural justice requires that before a law can be operative it must be promulgated, and that it would be against the principles of natural justice to permit the subjects of a State to be punished by a law of which they had no knowledge. Govind Lal Chhaganlal Patel v. Agricultural Produce Market Committee, (1975) 2 SCC 482 held that where the statute prescribes a mode of publication, that mode is mandatory: publication in a newspaper as required could not be replaced by publication in the Gazette, because the object was to reach the persons affected. B.K. Srinivasan v. State of Karnataka, (1987) 1 SCC 658 stated the general rule that subordinate legislation must be published or promulgated in some suitable manner, though not necessarily in the manner prescribed for a statute.

The pattern in recent decisions

The most instructive modern examples concern rules made about tribunals, and they show judicial control at full stretch. Rojer Mathew v. South Indian Bank Ltd., (2019) 6 SCC 1 struck down the rules made under Section 184 of the Finance Act, 2017 governing appointment, tenure and service conditions across nineteen tribunals, for excessive executive control, and directed fresh rules. When the substance was re-enacted, the Supreme Court in Madras Bar Association v. Union of India, decided 19 November 2025, struck down the corresponding provisions of the Tribunals Reforms Act, 2021 on the ground that Parliament had reproduced provisions already declared unconstitutional without curing the defects, and directed the constitution of a National Tribunals Commission within four months. That sequence is the best available answer to the examiner's implicit question, which control actually bites.

For full marks, cover: the origin in Locke and Montesquieu and the purpose the doctrine serves, which is liberty and not tidiness; the three propositions it contains; how the four systems this syllabus compares have implemented it, rigidly in the United States, hardly at all in the United Kingdom, and in France in the opposite direction to keep the ordinary courts out of administration; the Indian position with Ram Jawaya Kapur and the basic structure cases; the textual provisions and the deliberate overlaps; and the modern problem, which is that the administrative State combines all three functions in one body. Finish with the most recent Indian application, the Presidential Reference of November 2025, which is the best contemporary illustration available.

How the four systems apply it

The United States is the only rigid model. The Constitution vests legislative power in Congress, executive power in the President and judicial power in the courts by three separate vesting clauses, and reinforces the separation by checks: the veto and its override, the Senate's role in appointments and treaties, impeachment, and judicial review since Marbury v. Madison, 5 US 137 (1803). Youngstown Sheet and Tube Co. v. Sawyer, 343 US 579 (1952) is the classic application. To avert a strike during the Korean War, President Truman ordered the Secretary of Commerce to seize and operate the country's steel mills. The Supreme Court held the seizure unconstitutional: the President's power must stem from an act of Congress or from the Constitution itself, and Congress had considered and rejected seizure as a remedy. Jackson J.'s concurrence supplied the framework still used, of presidential power at its maximum when acting with congressional authority, in a twilight zone when Congress is silent, and at its lowest ebb when acting against the expressed will of Congress. Immigration and Naturalization Service v. Chadha, 462 US 919 (1983) struck down the one-House legislative veto for evading bicameralism and presentment.

The Indian position

India follows the British model of fused executive and legislature within a written constitution that provides for judicial review. Ram Jawaya Kapur v. State of Punjab, AIR 1955 SC 549 is the starting authority. Mukherjea C.J. held that the Indian Constitution has not indeed recognised the doctrine of separation of powers in its absolute rigidity, but the functions of the different parts or branches of the government have been sufficiently differentiated, and consequently it can very well be said that our Constitution does not contemplate assumption by one organ of functions that essentially belong to another.

The text provides for the separation in specific places. Article 50 directs the State to separate the judiciary from the executive in the public services. Articles 121 and 211 forbid discussion in the legislature of the conduct of a judge in the discharge of duty. Articles 122 and 212 bar courts from inquiring into the validity of legislative proceedings on the ground of irregularity of procedure. Articles 105 and 194 confer legislative privileges, and Article 361 confers immunity on the President and Governors.

The administrative State is the modern problem

The doctrine assumes three functions in three hands, and the regulatory agency holds all three. A body such as the Securities and Exchange Board of India makes regulations, investigates breaches of them, adjudicates and imposes penalties. So does a tribunal established by the very ministry whose orders it reviews. The law's answer has not been to unwind the combination, which is impracticable, but to separate the functions inside the body and to secure the independence of the adjudicating part. That is precisely the reasoning of the tribunal cases, from L. Chandra Kumar v. Union of India, (1997) 3 SCC 261, which held judicial review under Articles 226 and 32 to be part of the basic structure and tribunals to be supplemental rather than substitutional, to Madras Bar Association v. Union of India, decided 19 November 2025, which struck down the appointment and tenure provisions of the Tribunals Reforms Act, 2021 for violating judicial independence and the separation of powers, and directed the constitution of a National Tribunals Commission within four months.

Conclusion. Montesquieu's doctrine has never been implemented in the form he stated it, and no modern State could function if it were. What survives is the purpose: no organ should hold enough power to be unchecked, and the function of adjudication in particular must be kept out of the hands of the person with an interest in the outcome. India's constitutional settlement follows that purpose rather than the form. It permits a fused executive and legislature, permits the executive to legislate by ordinance and by rule, and permits adjudication by tribunals; and it then insists, as basic structure, on the two things the purpose actually requires, that the legislature may not decide a case and that judicial review may not be taken away. Indira Nehru Gandhi enforced the first, L. Chandra Kumar the second, and the tribunal litigation running from R. Gandhi in 2010 to Madras Bar Association in November 2025 shows that the second is still being enforced against successive attempts to work around it.

Its nature is best understood historically. Under the laissez-faire State the administration did three things, defence, policing and revenue, and the citizen's dealings with it were few. The welfare State, and in India the Directive Principles and the planning apparatus, made the administration the provider of education, health, food, housing, employment and social security. The regulatory State that followed liberalisation in 1991 changed the technique again, from owning and permitting to rule making and supervision by specialised regulators. At each stage the volume of official decisions affecting individual rights grew, and the legislature responded by conferring power in broad terms, leaving the detail to rules and the application to discretion.

Its scope covers four functions, and the classification has real consequences. Quasi-legislative functions, the making of rules and regulations, attract the controls of publication and ultra vires, but not natural justice, as Union of India v. Cynamide India Ltd., (1987) 2 SCC 720 held of price fixation. Quasi-judicial functions, the determination of disputes and of individual entitlements, attract natural justice, reasons and review. Purely administrative or ministerial functions attract the duty to act fairly. Ancillary functions, investigation, inspection, licensing, supervision and contracting, attract whichever of these the effect on the individual requires.

(b) Doctrine of Eminent power

Eminent domain, dominium eminens in Grotius, is the inherent power of the State to take private property for a public purpose without the owner's consent. It rests on two maxims, salus populi suprema lex, the welfare of the people is the supreme law, and necessitas publica major est quam privata, public necessity is greater than private. Its two classical limits are that the taking must be for a public purpose and that compensation must be paid.

The Indian constitutional history is a history of retreat and partial return. The Constitution originally guaranteed the right to acquire, hold and dispose of property in Article 19(1)(f) and protected against deprivation in Article 31. A sequence of amendments cut it down for land reform, including the First, Fourth, Seventeenth and Twenty-fifth, together with Articles 31A, 31B and the Ninth Schedule. The Constitution (Forty-fourth Amendment) Act, 1978 deleted Article 19(1)(f) and Article 31 and inserted Article 300A, which provides that no person shall be deprived of his property save by authority of law. Property thus ceased to be a fundamental right and became a constitutional right, enforceable in the High Court under Article 226 but not directly under Article 32.

The most important recent development is procedural and it post-dates every paper in this folder. In Kolkata Municipal Corporation v. Bimal Kumar Shah, decided 16 May 2024, the Supreme Court held that Article 300A is not a single right but comprises seven procedural sub-rights which the State must observe before depriving a person of property: the right to notice of the intention to acquire, in terms that are clear, cogent and meaningful; the right to be heard and to object; the right to a reasoned decision on those objections; the duty of the State to acquire only for a demonstrable public purpose; the right to fair compensation; the right to an efficient and expeditious process; and the right of conclusion, that is, final vesting. The acquisition before the Court, attempted unilaterally under Section 352 of the Kolkata Municipal Corporation Act, 1980, was set aside.

(c) Taxing Power

Taxation is the second inherent power of the State, and in India it is expressly conditioned. Article 265 provides that no tax shall be levied or collected except by authority of law. This means an Act of the competent legislature: an executive order, a departmental circular or an administrative instruction cannot impose a tax, and the requirement is one of the oldest rule of law guarantees in the Constitution.

Competence is distributed by Articles 245 and 246 read with the Seventh Schedule. The Union List, the State List and, since 2016, Article 246A inserted by the Constitution (One Hundred and First Amendment) Act, allocate taxing entries; Article 246A gives Parliament and the State legislatures simultaneous power to make laws on goods and services tax, and Article 279A creates the Goods and Services Tax Council. In Union of India v. Mohit Minerals Pvt. Ltd., (2022) 10 SCC 700 the Court held that the recommendations of the GST Council are persuasive and not binding, because both Parliament and the State legislatures hold simultaneous power under Article 246A, and described the arrangement as cooperative federalism in which the Union and the States are not always in a relationship of hierarchy.

Delegation of the taxing power is permitted within limits, and this is where administrative law meets tax. Corporation of Calcutta v. Liberty Cinema, AIR 1965 SC 1107 held that a levy which the Calcutta Municipal Act, 1951 called a licence fee on cinema houses was in substance a tax, because nothing was rendered in return for it, and then upheld the delegation of the power to fix its rate under Section 548(2) of that Act, because the legislature had supplied guidance by tying the levy to the needs of the Corporation. Devi Das Gopal Krishnan v. State of Punjab, AIR 1967 SC 1895 struck down an unguided power to fix the rate of sales tax at any figure the State chose, holding that fixing the rate is not an ancillary detail but part of the essential legislative function unless guidance is supplied. Avinder Singh v. State of Punjab, (1979) 1 SCC 137 upheld Section 90(4) of the Punjab Municipal Act, under which the State required municipalities to levy a tax of one rupee a bottle on foreign liquor, on the ground that the legislature had itself laid down the policy and had left only the selection of a rate within it.

Conclusion. The three notes describe the administrative State and the two inherent powers by which it reaches the citizen most directly, taking property and taking money. The pattern is the same in each. The power itself is not questioned: a modern State must administer, must sometimes acquire land, and must tax. What the law does is impose conditions of legality and procedure, and the movement in the last two years has been to tighten them. Bimal Kumar Shah in May 2024 converted Article 300A from a single sentence into seven procedural steps; Mineral Area Development Authority in July 2024 reallocated a large taxing power to the States and, in doing so, reminded every legislature that competence is a question of the text and not of practice; and the whole of the administrative process remains subject to the Kraipak principle that the label on a function decides nothing and the effect on the individual decides everything.

For full marks, cover: the definition and why discretion is unavoidable; the statutory language that creates it, which is the easiest way to organise the examples the question asks for; then a run of concrete Indian examples across licensing, preventive detention, investigation, acquisition, allocation of resources, contracts, clemency and travel documents, each attached to the case that controls it; and then the grounds on which the courts intervene, illustrated from those same examples rather than listed separately. The examiner has asked for examples, so a general essay on judicial review with no statute named in it will not score.

How statutes create discretion

Four drafting formulae account for almost all of it. A power expressed as "may", which imports a choice. A power exercisable where the authority "is satisfied" or "has reason to believe", which imports subjective satisfaction. A power to act "in the public interest", "as it thinks fit" or "if it considers it necessary", which supplies an open-textured standard. And a power to do what "may be prescribed", which delegates the making of the rule itself.

Preventive detention. The National Security Act, 1980 and its predecessors empower an officer to detain a person where he "is satisfied" that detention is necessary to prevent a person from acting in a manner prejudicial to the security of the State or the maintenance of public order. This is the widest discretion Indian law confers on an executive officer, and the courts control it procedurally rather than substantively: in Khudiram Das v. State of West Bengal, (1975) 2 SCC 81 the Court held that the subjective satisfaction must be based on material having a rational nexus with the object, that the grounds must be communicated so the detenu can make an effective representation, and that a ground which is vague, irrelevant or non-existent vitiates the order.

Investigation on subjective satisfaction. Section 237(b) of the Companies Act, 1956 allowed the Company Law Board to order an investigation if "in the opinion" of the Board there were circumstances suggesting fraud. In Barium Chemicals Ltd. v. Company Law Board, AIR 1967 SC 295 the Court held that even a subjectively worded power is reviewable: the existence of the circumstances is a condition precedent, and if there is no material, or the material is irrelevant, or no reasonable person could form the opinion, the order is bad.

Government contracts and tenders. Tata Cellular v. Union of India, (1994) 6 SCC 651 is the leading case: the government has freedom of contract and the court reviews the decision-making process, not the decision, but the process must be free of arbitrariness, bias and mala fides. Michigan Rubber (India) Ltd. v. State of Karnataka, (2012) 8 SCC 216 restated the test as whether the process was arbitrary or mala fide and whether the public interest was affected.

Clemency. The power under Articles 72 and 161 to pardon, remit or commute is the widest discretion in the Constitution, and it is not unreviewable. Maru Ram v. Union of India, (1981) 1 SCC 107 held that it must be exercised on the advice of the government and cannot be exercised arbitrarily or on irrelevant considerations, and Epuru Sudhakar v. Government of Andhra Pradesh, (2006) 8 SCC 161 set aside a remission granted on extraneous political grounds, holding that the exercise is subject to judicial review for non-application of mind, mala fides, extraneous considerations and the suppression of material.

What the courts do with these discretions

Read together, the examples show a single method. The court will not substitute its own choice, but it will insist that the discretion be exercised by the authority named and not on someone else's direction (Commissioner of Police, Bombay v. Gordhandas Bhanji, AIR 1952 SC 16, where a licence cancelled by the Commissioner on the Government's direction was held void, and the Court added that a public order must be judged by what it says on its face); for the purpose for which it was given (Padfield v. Minister of Agriculture, Fisheries and Food, [1968] AC 997, and in India State of Punjab v. Gurdial Singh, (1980) 2 SCC 471, where Krishna Iyer J. quashed an acquisition pursued for a political grudge, defining legal malice as the use of power for an alien purpose); on relevant material (Barium Chemicals); after hearing the person affected where his rights are at stake (A.K. Kraipak v. Union of India, (1969) 2 SCC 262, holding that the line between administrative and quasi-judicial power is thin and being obliterated); and with reasons (Union of India v. Mohan Lal Capoor, (1973) 2 SCC 836, reasons are the links between the material and the conclusion). Where a fundamental right is restricted, the intensity rises from Wednesbury to proportionality: Om Kumar v. Union of India, (2001) 2 SCC 386 and Modern Dental College and Research Centre v. State of Madhya Pradesh, (2016) 7 SCC 353.

The choice of punishment in a disciplinary proceeding. Once misconduct is proved, the penalty is chosen by the disciplinary authority, and this is the clearest example in Indian law of a discretion the courts deliberately refuse to exercise themselves. B.C. Chaturvedi v. Union of India, (1995) 6 SCC 749 held that the court does not sit as an appellate authority over the penalty and will not reappreciate the evidence, but that it may interfere where the punishment is shockingly disproportionate to the misconduct, and that the ordinary course on interference is to remit the matter to the authority rather than to substitute a penalty. Om Kumar v. Union of India, (2001) 2 SCC 386 placed that within the modern framework by classifying quantum of punishment as a matter for secondary Wednesbury review rather than for the primary proportionality review applied where a fundamental right is restricted.

For full marks, cover: this question is set in the same words as question 2 of Q.P. Code 12282, the first paper in this volume, and that answer takes the five systems one at a time. This answer takes the other route and compares them theme by theme, which is the structure that scores better, because an examiner asking for a comparative account is testing whether the candidate can compare rather than describe. Six themes carry the marks: the source of administrative law, the control of the power to delegate, whether there is a statutory code of procedure, the institution that supervises the administration, the grounds and intensity of review, and the remedy including damages against the State. Close with a table and with the two movements that matter, the long convergence on proportionality and codified procedure, and the sharp divergence of 2024, when the United States overruled Chevron. Either structure will pass; only this one is a comparison.

In France the source is neither constitution nor code but the case law of the Conseil d'État, which built the subject from Blanco (Tribunal des conflits, 8 February 1873) onwards, and only codified the citizen's procedural rights in 2016.

Theme two: the control on the power to delegate

Germany states the limit in the constitutional text. Article 80(1) of the Basic Law requires that where a statute authorises the making of subordinate legislation, the content, purpose and extent of the authority be determined in the statute itself. What India and America derived from case law, Germany wrote down.

The United States derives it as the intelligible principle test, from J.W. Hampton, Jr. and Co. v. United States, 276 US 394 (1928), applied to invalidate a federal statute only twice, in Panama Refining Co. v. Ryan, 293 US 388 (1935) and A.L.A. Schechter Poultry Corp. v. United States, 295 US 495 (1935).

Theme three: is there a code of administrative procedure

This is the theme on which India stands alone, and it is the most useful single point in the answer. The United States has had notice and comment rulemaking under Section 553 of the Administrative Procedure Act, 1946, formal adjudication before a separated hearing officer under Sections 554 and 556 to 557, and the statutory standards of review in Section 706, for eighty years. Germany codified the administrative act, the right to be heard, the duty to give reasons and the conditions of revocation in the Verwaltungsverfahrensgesetz of 1976, and court procedure in the Verwaltungsgerichtsordnung of 1960. France codified the citizen's procedural rights in the Code des relations entre le public et l'administration in 2016. The United Kingdom has no code but a developed common law of fairness and a unified tribunal system under the Tribunals, Courts and Enforcement Act, 2007.

India has neither a code nor a duty to consult, which is why the same principle has had to be established again in each new statutory setting: publication in Harla v. State of Rajasthan, AIR 1951 SC 467, reasons in Union of India v. Mohan Lal Capoor, (1973) 2 SCC 836, hearing in A.K. Kraipak v. Union of India, (1969) 2 SCC 262. The Law Commission's Fourteenth Report noticed the gap in 1958 and it has not been filled.

Theme five: grounds and intensity of review

England supplied the grammar. Ridge v. Baldwin, [1964] AC 40 revived natural justice by holding that the duty to act fairly attaches to the effect of a decision on the individual rather than to the label on the body: a chief constable dismissed by a watch committee without notice or hearing had his dismissal declared void. Anisminic Ltd. v. Foreign Compensation Commission, [1969] 2 AC 147 held that a determination reached on a wrong view of the law is a nullity, so an ouster clause protecting a determination protects nothing. Council of Civil Service Unions v. Minister for the Civil Service, [1985] AC 374 classified the grounds as illegality, irrationality and procedural impropriety, and proportionality arrived with the Human Rights Act, 1998 and R (Daly) v. Secretary of State for the Home Department, [2001] UKHL 26.

Germany supplied proportionality itself, structured in the Pharmacy judgment of the Federal Constitutional Court of 11 June 1958 into legitimate aim, suitability, necessity and balance, and it is that test which India later received.

India runs three standards at once and the answer should say so plainly. Wednesbury unreasonableness was received in Tata Cellular v. Union of India, (1994) 6 SCC 651, which also held that the court reviews the decision-making process and not the decision. Article 14 arbitrariness, from E.P. Royappa v. State of Tamil Nadu, (1974) 4 SCC 3 to Shayara Bano v. Union of India, (2017) 9 SCC 1, gives Indian courts a ground the others do not have, because it permits the invalidation of the rule and not merely of its application. And proportionality applies where a fundamental right is restricted, after Om Kumar v. Union of India, (2001) 2 SCC 386 and Modern Dental College and Research Centre v. State of Madhya Pradesh, (2016) 7 SCC 353.

On damages the systems separate sharply, and Blanco is the case to work. A child, Agnès Blanco, was injured at Bordeaux by a wagon pushed by workmen of the State tobacco factory. The Tribunal des conflits held that the liability of the State for damage caused to individuals by persons it employs in a public service cannot be governed by the principles of the Civil Code, that it is neither general nor absolute, that it has its own rules varying with the needs of the service, and that the administrative jurisdiction alone may decide it. From that beginning French law developed liability without fault, in Cames (Conseil d'État, 1895) and after. Germany provides for State liability constitutionally in Article 34 of the Basic Law read with the Civil Code; England abolished Crown immunity by the Crown Proceedings Act, 1947; America waived it in part by the Federal Tort Claims Act, 1946.

For full marks, cover: the doctrine in one paragraph, then the question actually asked, which is not what the doctrine says but how strictly it is applied. Treat rigidity and flexibility as two ends of a spectrum, give the American model as the rigid end with worked examples of what rigidity produces, the British and Indian models as the flexible end with worked examples of what flexibility produces, and show that each end fails in a characteristic way. Then state the Indian position precisely, because the examiner is testing whether the candidate can say what is non-negotiable in a system that is otherwise flexible. Finish with the two most recent illustrations, the Presidential Reference of November 2025 and the tribunal litigation.

Youngstown Sheet and Tube Co. v. Sawyer, 343 US 579 (1952) is the standard example of rigidity enforced. Faced with a threatened steel strike during the Korean War, President Truman ordered the seizure of the mills. The Court held the order unconstitutional because the power to take private property for public use is a legislative power, Congress had considered and rejected seizure when passing the Taft-Hartley Act, and the President's power must come from a statute or the Constitution itself. Jackson J.'s concurrence gave the three zones now used everywhere: presidential power is at its maximum when Congress has authorised, in a twilight zone when Congress is silent, and at its lowest ebb when the President acts against the expressed will of Congress.

Immigration and Naturalization Service v. Chadha, 462 US 919 (1983) struck down the legislative veto, a device by which one House could annul an executive decision, because any exercise of legislative power must satisfy bicameralism and presentment. The device was convenient, was contained in scores of statutes, and was unconstitutional all the same, which is rigidity at work.

India is flexible by design and the leading statement is Ram Jawaya Kapur. In Ram Jawaya Kapur v. State of Punjab, AIR 1955 SC 549 Mukherjea C.J. held that the Constitution has not recognised the doctrine of separation of powers in its absolute rigidity, but that the functions of the different branches have been sufficiently differentiated, so that one organ may not assume functions that essentially belong to another. The examples of flexibility are everywhere.

The executive legislates. Delegated legislation is the largest single source of Indian law by volume, and the ordinance power under Articles 123 and 213 allows the executive to make law when the House is not in session. Flexibility here has produced its own abuse, addressed in Krishna Kumar Singh v. State of Bihar, (2017) 3 SCC 1, where a bench of seven held that the re-promulgation of ordinances, in that case a chain of Bihar ordinances kept alive for years without ever being placed before the legislature, is a fraud on the Constitution and a subversion of democratic legislative processes, and that the satisfaction of the Governor is subject to judicial review.

The legislature adjudicates, when it punishes for contempt or breach of privilege under Articles 105 and 194, and the judiciary has accepted that it may do so while insisting that the courts decide the existence and extent of a privilege.

Where India is rigid

Flexibility stops at two points, and both are basic structure. The first is that the legislature may not decide a case. Indira Nehru Gandhi v. Raj Narain, 1975 Supp SCC 1 struck down Article 329A(4), which had validated the Prime Minister's election and barred every court from examining it, because Parliament had exercised judicial power rather than legislative power. The second is that judicial review may not be taken away. L. Chandra Kumar v. Union of India, (1997) 3 SCC 261, a bench of seven, held that the jurisdiction of the High Courts under Articles 226 and 227 and of the Supreme Court under Article 32 is part of the basic structure, and struck down the exclusion clauses of Articles 323A(2)(d) and 323B(3)(d) to that extent.

The Presidential Reference shows the boundary being policed in the other direction. In State of Tamil Nadu v. Governor of Tamil Nadu, decided 8 April 2025, the Court had prescribed timelines for the Governor and the President to act on Bills under Articles 200 and 201 and had treated ten Bills as having received deemed assent under Article 142. On a reference under Article 143, a Constitution Bench advised on 20 November 2025 that no judicially prescribed timelines may be imposed, that deemed assent is alien to the Constitution because one constitutional authority cannot substitute itself for another, and that decisions under those Articles are not ordinarily justiciable, while a Governor may not sit on a Bill indefinitely. Whichever way one reads the merits, the reasoning is a separation of powers reasoning applied to a court.

The device that makes Indian flexibility safe

A flexible separation of powers is only safe if something is beyond the reach of a temporary majority, and in India that something is the basic structure. This is the point at which the Indian answer differs from the British one, and it should be given its own place in the answer rather than assumed.

I.R. Coelho v. State of Tamil Nadu, (2007) 2 SCC 1, a bench of nine, completed the structure by holding that laws placed in the Ninth Schedule after 24 April 1973, the date of Kesavananda, are open to challenge on the ground that they damage the basic structure, so the Schedule is no longer an absolute shield.

The relevance to rigidity and flexibility is direct. India can afford to let the executive legislate by rule and by ordinance, to let tribunals adjudicate, and to let the judiciary lay down guidelines in the absence of a statute, precisely because two things cannot be altered even by amendment: the legislature may not decide a case, and judicial review may not be excluded. Britain, having no such device, must rely on political convention for the same protection, and the United States achieves it by rigidity of form rather than by an unamendable core.

For full marks, cover: the classical statement and the modern reformulation in two tight paragraphs, because the question's weight is on the second half; then the current scenario, organised by theme rather than as a list of news items, with the constitutional standard stated for each; and an honest assessment. The themes that matter now are the appointment of independent constitutional authorities, the overhaul of the criminal law in 2024, the retreat of transparency after the data protection rules of November 2025, the treatment of property, the independence of tribunals, the assent controversy, and delay. Every one of these is post-2022, and an answer built only on Kesavananda, ADM Jabalpur and Maneka Gandhi has not answered the question that was set.

In India the concept has three settled constitutional consequences. It is part of the basic structure, so it cannot be amended away (Kesavananda Bharati v. State of Kerala, (1973) 4 SCC 225; Indira Nehru Gandhi v. Raj Narain, 1975 Supp SCC 1). It is enforceable through Article 14 as a prohibition of arbitrariness (E.P. Royappa v. State of Tamil Nadu, (1974) 4 SCC 3; Shayara Bano v. Union of India, (2017) 9 SCC 1). And it is not suspended by an emergency, a proposition established by Khanna J.'s lone dissent in A.D.M. Jabalpur v. Shivkant Shukla, (1976) 2 SCC 521, written into the Constitution by the Forty-fourth Amendment in 1978, and finally vindicated when nine judges expressly overruled the majority in K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1.

The overhaul of the criminal law. On 1 July 2024 the Bharatiya Nyaya Sanhita, 2023, the Bharatiya Nagarik Suraksha Sanhita, 2023 and the Bharatiya Sakshya Adhiniyam, 2023 replaced the Indian Penal Code, 1860, the Code of Criminal Procedure, 1973 and the Indian Evidence Act, 1872. Two rule of law consequences follow. The first is accessibility, which Raz and Bingham both make central: for a period of years the same court applies two codes side by side depending on the date of the offence, and every practitioner, police officer and litigant must know both. The second is that some of the reforms are rule of law improvements on their face, in particular the provision in the Sanhita for the release of a first-time undertaker of a lesser offence who has served one third of the maximum sentence, with the duty of application placed on the jail superintendent, which addresses the largest single rule of law failure in Indian criminal justice. Prison Statistics India for 2023 recorded occupancy at about 120.8 per cent and undertrials at about 73.5 per cent of the prison population, and a system in which nearly three in four prisoners have not been convicted of anything is a rule of law problem before it is a policy problem.

Property and process. Kolkata Municipal Corporation v. Bimal Kumar Shah, decided 16 May 2024, held that Article 300A carries seven procedural sub-rights, of notice, hearing, a reasoned decision, demonstrable public purpose, fair compensation, an efficient and expeditious process and final conclusion. That is the rule of law applied to the most coercive ordinary power of the State, and it is a direct application of Jaisinghani: the power to acquire is not denied, it is confined and made procedural.

Independence of adjudication. In Madras Bar Association v. Union of India, decided 19 November 2025, the Supreme Court struck down the core appointment and tenure provisions of the Tribunals Reforms Act, 2021, holding that Parliament had re-enacted provisions already declared unconstitutional without curing the defects, and directed the constitution of a National Tribunals Commission within four months. Tribunals now decide more disputes than the High Courts, and a tribunal whose members depend on the executive for reappointment cannot deliver the equality before the law that Dicey's second proposition demands.

Assessment

Three honest observations. First, the Indian rule of law has been strongest as a doctrine of procedure: notice, hearing, reasons, publication and review are enforced with real consistency, and Bimal Kumar Shah is the newest instance. Second, it is weakest where the remedy depends on time: a right vindicated after twenty years of litigation, or a bail application decided after the sentence would have expired, satisfies no definition of the rule of law on any list. Third, the pattern of the last three years is that the Court sets a standard and the political branches test it, in the Election Commission appointments, in the tribunals, and in the assent reference. That is not a failure of the rule of law; contestation is what a functioning constitutional order looks like. It becomes a failure only if the contest is settled by the simple repetition of what has already been held unconstitutional, which is exactly what the Supreme Court refused to permit in November 2025.

Two of the current developments worked in full

The current scenario is usually written as a list of dates, and it scores better when two of the episodes are worked out.

The electoral bonds decision of 15 February 2024, Association for Democratic Reforms v. Union of India. The scheme allowed a person to buy bearer bonds from a designated bank and give them to a political party, with the identity of the purchaser known to the bank and to nobody else, supported by amendments removing the disclosure obligations in the Representation of the People Act, 1951, the Companies Act, 2013 and the income tax legislation, and lifting the cap on corporate donations. A Constitution Bench struck down the scheme and those amendments, holding that the voter's right to information about the funding of political parties is protected by Article 19(1)(a), that the State's objective of curbing black money did not justify complete anonymity when less restrictive alternatives existed, and directing the issuing bank to disclose the bonds purchased and encashed. Bingham's formulation is the one that fits: a citizen cannot hold power to account for decisions whose financial origins the law has deliberately concealed, and here the concealment had been created by statute rather than by administrative practice.

For full marks, cover: a precise definition and the forms it takes, because "rules and regulations" is not an answer at this level; the classification by the nature of the power delegated, including the exceptional Henry VIII clause and conditional legislation; the constitutional limit set by In re Delhi Laws Act, because importance must be discussed within a limit; then the importance itself, which is the second and larger limb, argued with Indian examples rather than asserted; and finally the dangers, since the examiner expects a critical treatment and the safeguards follow from the dangers.

What delegated legislation is

Delegated or subordinate legislation is law made by an executive or other authority under a power conferred by the legislature. Three elements matter. The power must come from a statute, so there is no inherent power to legislate. What is produced is legislative in character, that is, a general rule of conduct operating prospectively on a class, as distinct from an order directed at a named person. And it has the force of law, so breach carries the consequence the parent Act attaches.

Classification by the nature of the power is more useful than classification by name. Normal delegation is the ordinary case, where the statute lays down policy and leaves the detail. Conditional legislation is not delegation at all in the strict sense: the legislature makes the law complete and leaves to the executive only the determination of the fact or the date on which it will apply to a place or class, and the distinction is important because a conditional power needs no policy guidance beyond the condition itself. Henry VIII clauses, which allow the executive to modify the parent Act itself, usually to remove difficulties, are the extreme case; Indian courts tolerate them where the power is confined to removing difficulties consistently with the Act, and they are read narrowly. Sub-delegation, where the delegate further delegates, requires express authority: Agricultural Market Committee v. Shalimar Chemical Works Ltd., (1997) 5 SCC 516.

The importance of delegated legislation

First, pressure on legislative time. Parliament sits for a limited number of days and must debate finance, policy and confidence. It cannot draft the hundreds of pages of technical prescription that a single regulatory statute requires. The Goods and Services Tax legislation of 2017 is a working example: the Acts fix the charge, the taxable event, the rates mechanism and the machinery, and the rules and notifications carry the returns, invoices, refunds, e-way bills and procedures, which have been amended continuously since.

Second, technicality. Legislators are not pharmacologists, actuaries, spectrum engineers or accountants. The standards for a drug, the solvency margin of an insurer, the technical conditions of a telecom licence and the accounting norms for a listed company must be written by people who understand them and revised when the science or the market moves.

Third, flexibility and speed of amendment. A rule can be amended in days and a statute cannot. This is not a convenience but often the whole point of the scheme: a price control, a customs duty exemption or an emergency standard that takes a session to change is useless.

Sixth, filling in a framework statute, which is the modern regulatory pattern. The SEBI Act, 1992, the Telecom Regulatory Authority of India Act, 1997, the Electricity Act, 2003, the Insolvency and Bankruptcy Code, 2016 and the Digital Personal Data Protection Act, 2023 all create an institution, state objects and confer powers, leaving the substantive norms to regulations. The DPDP Act is the newest example and a striking one: it was enacted in August 2023 and did not operate until the Rules were notified on 13 November 2025, more than two years later, which shows both the dependence of the modern statute on its subordinate legislation and the accountability gap that dependence creates.

The dangers

The first danger is the transfer of policy under the guise of detail, which is what the essential legislative function doctrine exists to prevent, and which the courts catch only when someone litigates.

The fourth is that subordinate legislation may be used to escape a judicial holding, which is not hypothetical. The rules made under Section 184 of the Finance Act, 2017 for tribunal appointments were struck down in Rojer Mathew v. South Indian Bank Ltd., (2019) 6 SCC 1; their successors were struck down in 2020 and 2021; and when the substance was re-enacted in the Tribunals Reforms Act, 2021, the Supreme Court struck that down too in Madras Bar Association v. Union of India, decided 19 November 2025.

Conditional legislation and the Henry VIII clause

Two forms of delegation sit at the edges of the subject, and an answer that distinguishes them shows a command of it that a general treatment does not.

The Henry VIII clause is the opposite extreme. It permits the executive to modify the parent Act itself, usually for the purpose of removing difficulties in giving effect to the statute, and it takes its nickname from a monarch remembered for governing by proclamation. Indian courts tolerate it, and confine it strictly. Jalan Trading Co. Pvt. Ltd. v. Mill Mazdoor Sabha, AIR 1967 SC 691 struck down Section 37 of the Payment of Bonus Act, 1965, which allowed the Central Government to make such provision as it thought necessary for removing doubts or difficulties and declared its order final, as excessive delegation: a power to alter the Act, coupled with a bar on questioning the alteration, transfers the legislative function itself. Gammon India Ltd. v. Union of India, (1974) 1 SCC 596 upheld the corresponding clause in the Contract Labour (Regulation and Abolition) Act, 1970, because it was confined to removing difficulties in giving effect to the provisions of the Act and carried no finality clause.

Conclusion. Delegated legislation is not a concession to executive convenience but a structural necessity of the modern State, and the honest way to state its importance is that no regulatory statute passed in India in the last thirty years could function without it: the Act supplies the institution and the objects, and the rules supply the law that citizens and firms actually obey. Krishna Iyer J. described it in Avinder Singh v. State of Punjab, (1979) 1 SCC 137 as a necessary evil, and both words earn their place. The necessity is proved by the GST rules, the regulatory codes and the pandemic orders; the evil is proved by the weakness of the controls, a laying requirement that is directory, committees that cannot annul, and no statutory duty to consult before making a rule that will bind millions. The reform that follows from the analysis is therefore not to delegate less, which is impossible, but to legislate the procedure by which delegated power is exercised, which is the one thing India, alone among the systems it is compared with, has never done.

(a) Role of Civil Services

Start from what the service actually decides, because that is what makes it an administrative law subject at all. The permanent executive advises the minister, drafts the subordinate legislation under which most Indian regulation operates, exercises statutory discretion in millions of individual decisions, adjudicates at first instance under most Indian statutes, and delivers the services. For the ordinary citizen the civil service is not a part of the administrative process; it is the whole of it.

The constitutional scheme is Part XIV, Articles 308 to 323, and it is built on a single trade. Article 310 preserves the doctrine of pleasure and Article 311 buys it back with two guarantees: no dismissal or removal by an authority subordinate to the appointing authority, and no dismissal, removal or reduction in rank without an inquiry with notice of the charges and a reasonable opportunity of being heard. Article 309 leaves recruitment and conditions of service to rules made by the executive until the legislature acts, which is why most Indian service law is subordinate legislation; Article 312 creates the All India Services; Articles 315 to 323 secure the Public Service Commissions.

The accountability that actually operates is not dismissal. Financial accountability runs through the Comptroller and Auditor General under Articles 148 to 151 and the Public Accounts Committee. Vigilance accountability was reconstructed judicially in Vineet Narain v. Union of India, (1998) 1 SCC 226, which gave the Central Vigilance Commission statutory status, placed the Delhi Special Police Establishment under its superintendence, and struck down the single directive requiring prior approval before officers of and above the rank of Joint Secretary could be investigated; when that protection was re-enacted as Section 6A of the Delhi Special Police Establishment Act, 1946, it was struck down again in Subramanian Swamy v. Director, CBI, (2014) 8 SCC 682 as violating Article 14, on the reasoning that the rank of a suspect cannot decide whether the law applies to him. Personal liability is possible but narrow: Lucknow Development Authority v. M.K. Gupta, (1994) 1 SCC 243 allowed exemplary damages against a public body and said the amount should be recoverable from the officer personally responsible, while the attempt to fix a minister personally in Common Cause, A Registered Society v. Union of India, (1996) 6 SCC 530 and (1996) 6 SCC 593 was recalled on review on 3 August 1999.

(b) Regulation and De-regulation Policy in India

Take this note as three regimes and one constitutional standard that survives all three.

Regime one, control by prior permission, to 1991. The Industries (Development and Regulation) Act, 1951 required a licence to establish, expand or relocate an industrial undertaking; the Monopolies and Restrictive Trade Practices Act, 1969 required large houses to obtain approval before expanding; the Foreign Exchange Regulation Act, 1973 prohibited every foreign exchange transaction unless permitted. The instrument throughout was an individual discretionary permission, which is the form of State power hardest to review and easiest to sell.

Regime two, regulation by rule, after the New Industrial Policy of 24 July 1991. Industrial licensing went except for a short reserved list, and the State changed technique rather than withdrawing: SEBI became statutory in 1992, the Telecom Regulatory Authority of India followed in 1997, the electricity, insurance, competition, real estate and insolvency regulators after. FERA became the Foreign Exchange Management Act, 1999, whose title records the whole change, and the MRTP Act became the Competition Act, 2002.

The constitutional standard has not changed across any of it, and this is the part that earns the marks. Ramana Dayaram Shetty v. International Airport Authority of India, (1979) 3 SCC 489 held that the State cannot act arbitrarily in distributing largesse and must conform to standards that are not arbitrary, irrational or irrelevant. Centre for Public Interest Litigation v. Union of India, (2012) 3 SCC 1 applied it to cancel 122 telecom licences allotted on a first-come-first-served basis, and the advisory opinion in Natural Resources Allocation, In re Special Reference No. 1 of 2012, (2012) 10 SCC 1 corrected the over-reading by holding that auction is not the only constitutionally permissible method, the test being whether the method serves the common good and is not arbitrary. Against those stands BALCO Employees' Union v. Union of India, (2002) 2 SCC 333, holding the decision to disinvest to be economic policy in which the court will not interfere absent illegality. Read together they state the rule for a deregulating State exactly: the government chooses the policy and the court examines the method.

(c) Role of Tribunals in adjudication

The case for and against should open the note. In favour: specialised members, procedure freed from the Evidence Act and the Civil Procedure Code, lower cost, faster disposal and relief of the courts' docket. Against: a class of disputes is removed from an independent judiciary and given to bodies whose members are selected, paid and considered for reappointment by the executive, which in most of those disputes is the opposing party.

The constitutional foundation is the Forty-second Amendment of 1976, which inserted Article 323A for administrative tribunals in service matters, under which the Administrative Tribunals Act, 1985 was enacted, and Article 323B for the enumerated subjects, including taxation, foreign exchange, industrial and labour disputes, land reform, elections and essential goods.

Union of India v. R. Gandhi, (2010) 11 SCC 1 upheld the transfer of company jurisdiction to the National Company Law Tribunal but struck down the provisions on qualifications and selection, on the principle that a tribunal which takes over the jurisdiction of a court must inherit that court's standards of independence and competence. Madras Bar Association v. Union of India, (2014) 10 SCC 1 struck down the National Tax Tribunal Act, 2005 outright, because substantial questions of law previously decided by High Courts could not be transferred to a body whose independence was not secured. Rojer Mathew v. South Indian Bank Ltd., (2019) 6 SCC 1 struck down the rules made under Section 184 of the Finance Act, 2017 governing appointment, tenure and removal across nineteen tribunals, for excessive executive control.

Assessment. Tribunals now decide more disputes in India than the High Courts, so the question is not whether they may exist but on what terms, and four terms are settled: High Court supervision cannot be ousted; the selection process must be dominated by the judiciary; tenure must be long enough that a member is not seeking the executive's favour for reappointment; and the ministry whose orders are challenged must not administer the tribunal that hears the challenge. What has been missing is an institution to enforce those terms without fresh litigation every few years, which is exactly what the Commission directed in November 2025 is meant to be. The practical criticism is separate and equally serious: vacancies and pendency have in several tribunals reproduced the delay they were created to cure.

For full marks, cover: all three notes are set out although the paper asks for two. Note (a) must distinguish responsibility from accountability and then work through the four channels, internal, parliamentary, judicial and citizen-facing, with authority for each. Note (b) must treat globalisation and liberalisation as a legal event and not an economic one, showing what it did to Indian administrative law. Note (c) must classify the agencies by legal form, because the form decides the accountability, and then say what the proliferation has cost.

Internal accountability comes first because it operates daily. The hierarchy supervises, the Conduct Rules bind, and the disciplinary machinery under Article 311 and the service rules punishes, subject to the safeguards in Union of India v. Tulsiram Patel, (1985) 3 SCC 398. Financial accountability runs through the Comptroller and Auditor General under Articles 148 to 151, whose reports are laid before the House and examined by the Public Accounts Committee. Anti-corruption accountability runs through the Central Vigilance Commission, whose statutory status and power of superintendence over the Delhi Special Police Establishment came from Vineet Narain v. Union of India, (1998) 1 SCC 226, in which the Court also struck down the "single directive" requiring prior approval before investigating senior officers; when the same protection was re-enacted as Section 6A of the Delhi Special Police Establishment Act, 1946, it was struck down again in Subramanian Swamy v. Director, CBI, (2014) 8 SCC 682 as violative of Article 14.

Judicial accountability is the sharpest channel. Writs under Articles 32 and 226 quash the decision; but Indian courts have gone further and made the administration pay. Rudul Sah v. State of Bihar, (1983) 4 SCC 141 awarded compensation in a writ petition to a man kept in jail for fourteen years after his acquittal, and Nilabati Behera v. State of Orissa, (1993) 2 SCC 746 established that compensation in public law for the violation of a fundamental right is a remedy in its own right, distinct from a tort action and unaffected by sovereign immunity. Lucknow Development Authority v. M.K. Gupta, (1994) 1 SCC 243 allowed exemplary damages against a public body for harassment of a citizen and, importantly, said that the amount should be recoverable from the officer responsible where the misfeasance is personal. Common Cause, A Registered Society v. Union of India, (1996) 6 SCC 530 cancelled fifteen retail petrol outlets allotted by a Minister out of his discretionary quota as arbitrary and mala fide, and by a further judgment reported at (1996) 6 SCC 593 directed him personally to pay rupees fifty lakh as exemplary damages. That second direction did not survive. On review, by a judgment of 3 August 1999, the Court recalled both the award of damages and the direction to register a criminal case, holding that the personal liability of a public servant in tort must be established in proceedings brought for that purpose and cannot be fixed by a direction in a writ petition. Cite the 1996 judgment for the cancellation of arbitrary largesse and the 1999 review for the limit on making an individual minister pay in the writ jurisdiction itself.

Assessment. India has more accountability institutions than most comparable systems and less accountability than the number suggests, for three structural reasons. Sanctions attach to procedural default rather than to outcomes, so a scheme may fail entirely without anyone being answerable. The protections designed to keep the honest officer safe, in Article 311 and in the sanction requirement for prosecution formerly in Section 197 of the Code of Criminal Procedure, 1973 and now carried into the Bharatiya Nagarik Suraksha Sanhita, 2023, also shelter the negligent. And the informal sanction that is actually used, transfer, is the one instrument that no accountability framework counts, which is why T.S.R. Subramanian v. Union of India, (2013) 15 SCC 732 directed minimum tenure and a Civil Services Board, and why the recording of oral instructions matters more than it sounds: an instruction that leaves no trace produces a decision with no author, and a decision with no author cannot be accounted for by anyone.

First, the technique of control changed from permission to regulation. The State stopped deciding who might produce and began setting the terms on which everyone may. That transferred power from ministries to specialised statutory regulators, and it created the modern Indian problem of a single body that makes regulations, investigates their breach and adjudicates it, which the courts have addressed by insisting on separation of function within the body and on reasons: Clariant International Ltd. v. SEBI, (2004) 8 SCC 524, Competition Commission of India v. Steel Authority of India Ltd., (2010) 10 SCC 744, and Competition Commission of India v. Bharti Airtel Ltd., (2019) 2 SCC 521 on the priority of the sectoral regulator where jurisdictions overlap.

Second, international obligations became a source of domestic administrative standards. The Agreement on Trade Related Aspects of Intellectual Property Rights required the amendments to the Patents Act, 1970 made in 1999, 2002 and 2005, including the product patent regime and the safeguard in Section 3(d) upheld in Novartis AG v. Union of India, (2013) 6 SCC 1. Sanitary and technical standards obligations shaped the food safety regime under the Food Safety and Standards Act, 2006.

Fourth, the State's own commercial conduct was subjected to public law. Ramana Dayaram Shetty v. International Airport Authority of India, (1979) 3 SCC 489 had already held that the State cannot distribute largesse arbitrarily, and ABL International Ltd. v. Export Credit Guarantee Corporation of India Ltd., (2004) 3 SCC 553 confirmed that a writ petition is maintainable in contractual matters where the State or its instrumentality acts arbitrarily.

Assessment. Liberalisation reduced the most reviewable-resistant form of State power, the individual prior permission, and replaced it with published rules applied by bodies that must give reasons, which is a gain for administrative law. Its unfinished business is threefold: regulators multiplied without a common statute of procedure, appointment and appeal; the withdrawal of the State from direct provision moved large areas of public activity outside Article 12 without moving them outside public concern; and the current deregulation programme, the Jan Vishwas Acts and the Deregulation Commission of 2025, reduces penalties faster than it builds the capacity to enforce what remains.

Government companies, registered under the companies legislation with State shareholding, and registered societies performing governmental functions, are tested by the criteria in Ajay Hasia v. Khalid Mujib Sehravardi, (1981) 1 SCC 722, which made the enquiry independent of legal form, and consolidated by seven judges in Pradeep Kumar Biswas v. Indian Institute of Chemical Biology, (2002) 5 SCC 111 into a single test of financial, functional and administrative domination that is particular to the body and pervasive, mere regulatory control not being enough.

Regulatory commissions, such as the Securities and Exchange Board of India, the Telecom Regulatory Authority of India, the Central and State Electricity Regulatory Commissions, the Insurance Regulatory and Development Authority, the Competition Commission and the Real Estate Regulatory Authorities, combine rule making, investigation and adjudication and are the characteristic institution of the post-1991 State.

Assessment. The proliferation of agencies has given India specialised capacity it did not have in 1991, and it has given it no common law of agencies. Appointment, tenure, procedure, the duty to consult before regulating, the standard of reasons and the route of appeal all differ from statute to statute, so the same administrative question is answered differently depending on which Act created the body. Every other system in the comparison has answered this with a general instrument, the American Administrative Procedure Act of 1946, the German Act of 1976, the French Code of 2016. India has answered it case by case, which is why the same principle has had to be established repeatedly, and why the direction in November 2025 to create a National Tribunals Commission matters beyond tribunals: it is the first serious attempt at a common institution for a class of agencies.