The Inter-State River Water Disputes Act 1956 is the statute that decides who gets the water when two Indian States claim the same river. It does something no other Indian statute attempts on this scale: it takes a class of disputes that would ordinarily belong to the courts, removes them from the courts by name, and hands them to a tribunal that exists for one river and one quarrel at a time. The Act, formally the Inter-State River Water Disputes Act, 1956, Act 33 of 1956, was Parliament’s answer to a design problem the Constitution had deliberately left open. Article 262 of the Constitution empowers Parliament to build exactly this kind of machinery and then to close the courtroom door behind it. The 1956 Act is what Parliament built, and the door it closed has never stayed shut.

Two features of the design determine how the whole system behaves, and a reader who holds them in view can predict most of what has happened under the Act. The first is gatekeeping. A tribunal does not come into existence because a dispute exists, or because a State demands one, or because a river is running dry in one State while another State builds a dam on it. A tribunal comes into existence when the Central Government forms the opinion that the dispute cannot be settled by negotiations, and the clock that the statute sets for constituting the tribunal runs only from a State’s request, while no clock at all runs on the forming of that opinion. The second feature is incomplete finality. The Act declares the tribunal’s decision final and binding, gives it the force of a decree of the Supreme Court, and bars the jurisdiction of every court including the Supreme Court itself. Even so, the Supreme Court has held that an aggrieved State may still come before it by special leave under Article 136, because once the tribunal has decided, what remains is no longer the water dispute the Constitution allowed Parliament to fence off. A statute designed to end river litigation in a tribunal has therefore repeatedly ended river litigation in the Supreme Court, and that outcome is not an accident of drafting. It is written into the structure.

The stakes explain the statute’s survival. The basins that have passed through its machinery, the Krishna, the Godavari, the Cauvery, the Narmada, the Ravi and Beas, the Vansadhara, the Mahadayi and the Mahanadi, carry the irrigation, drinking water and power systems of a substantial part of the Indian landmass, and the cities whose names recur in the proceedings, Bengaluru, Chennai, Hyderabad and the delta districts of the south and the canal commands of the north, are not abstractions in the pleadings. A statute that allocates scarcity among governments will be fought over as long as the scarcity lasts, and the 1956 Act’s longevity is owed less to satisfaction with its working than to the absence of any alternative forum the disputants trust more.

There is a third feature, quieter than the other two and in some ways the most revealing. In the same session of Parliament in 1956, the legislature enacted a companion statute, the River Boards Act, 1956, which empowered the Central Government to establish boards to advise on and plan the regulation and development of inter-State rivers before quarrels hardened into disputes. No river board has ever been established under that Act. Not one, in the entire life of the statute. The National Commission to Review the Working of the Constitution described the River Boards Act as having remained a dead letter. So the Indian Union acquired, in a single year, an instrument for preventing river disputes by planning and an instrument for deciding them by adjudication, and then used only the second. This article is the pillar treatment of the adjudication statute: the constitutional design beneath it, the machinery section by section, the tribunals it has produced, the amendments that tried to discipline it, and the honest account of what in the design works and what has never worked.

The problem the Act was built to solve

A river does not respect a boundary drawn on a map, and a State boundary in India almost always cuts across a drainage basin. The State upstream controls the headworks, the storage and the timing of releases. The State downstream holds the older irrigation systems, the established ayacut and the political expectation that the flows its farmers have used for generations will continue. Each State legislates for water within its own territory under its own constitutional entry, each plans projects in its own five year plans, and each can point to genuine need. The dispute that results is not a contest between a right and a wrong. It is a contest between two public interests of the same kind, fought over a resource whose quantity varies from year to year and whose measurement is itself contested. Any legal system that tries to settle such a contest with the tools of private litigation, where a court identifies a proprietary right and decrees it, will find the tools poorly fitted to the task. That misfit is the starting point of the entire design, under the British period, under the Constituent Assembly and under the 1956 Act alike.

The Government of India Act, 1935 confronted the same problem between provinces and princely-influenced units of the federation it created, and its solution is recognisable as the direct ancestor of the 1956 Act. Sections 130 to 134 of that Act laid down a complaint procedure. A provincial government that believed its interests in water supplies were prejudicially affected by another unit’s action could complain to the Governor-General, who, after consulting the governments concerned, would appoint a commission to investigate and report, and would then decide the matter himself. The Governor-General’s decision was given finality, and the Federal Court’s jurisdiction over such complaints was excluded in favour of the special procedure. Three elements of the 1956 settlement are already visible in 1935: a complaint by a government rather than a suit by a litigant, an expert investigatory body between the complaint and the decision, and the exclusion of the ordinary court from the merits of the allocation.

The complaint machinery of the 1935 Act was used sparingly, and the disputes that mattered most in the late colonial period, over the Indus system and its canals, were handled by a commission of inquiry chaired by Sir Cyril Radcliffe in a different capacity and then overtaken entirely by Partition. What the Constituent Assembly inherited in 1947 was therefore not a working system but a constitutional idea: that inter-unit water disputes needed a special channel, distinct from the courts, in which investigation and political negotiation could do the work that pleadings could not. The Assembly wrote that idea into the Constitution in the spare language of Article 262 and left every operational question, who complains, who decides, in what form, with what finality, to Parliament.

The pre-constitutional record also left the Assembly a specific warning about agreements, which matters because the 1956 Act treats agreements as both a foundation and a failure mode. The Cauvery offers the standard example. The Madras Presidency and the princely State of Mysore settled their sharing in agreements of 1892 and 1924, the second of which was expressed to run for fifty years and to be open to review thereafter. The 1924 agreement regulated Mysore’s construction of the Krishnaraja Sagar dam and Madras’s construction at Mettur, and for half a century the basin’s development proceeded inside its terms. When the fifty years ran out in 1974, the successor States inherited the works but not the settlement: Karnataka treated the agreement as spent and its constraints as lapsed, Tamil Nadu treated the utilisation built under it as an established entitlement that any new arrangement had to protect, and the disagreement between those two characterisations is recognisably the dispute the Cauvery tribunal was constituted to decide sixteen years later. The lesson the 1956 draftsmen drew from such arrangements was not that agreements fail, but that an agreement without a permanent institution behind it decays with the politics that signed it. Section 2(c) therefore counts the interpretation and implementation of water agreements as water disputes in their own right, so that a broken settlement can be carried into adjudication without waiting for a fresh quarrel over the water itself.

Partition sharpened the problem in a way the Assembly could not ignore while it drafted. The new international boundary cut through the Indus basin canal systems, and the standstill arrangements and interim agreements of 1947 and 1948 showed how quickly water could move from an administrative subject to an existential one. The Indus question went to international negotiation and eventually to the Indus Waters Treaty of 1960, which sits outside this article’s statute altogether because the disputants were two sovereign states. Within the Union, though, the lesson generalised. The reorganisation of States in 1956 was about to redraw internal boundaries on linguistic lines, and every redrawing multiplied the number of rivers that flowed from one State into another and the number of governments with a statutory claim on the same water. Parliament passed the disputes Act and the river boards Act in the same year as the States Reorganisation Act for a reason: the map was being redrawn, and the quarrels the new map would generate needed machinery before they arrived.

The constitutional design: Article 262 and the two entries

Article 262 is short, and its two clauses do different work. Clause (1) provides that Parliament may by law provide for the adjudication of any dispute or complaint with respect to the use, distribution or control of the waters of, or in, any inter-State river or river valley. Clause (2) provides that, notwithstanding anything in the Constitution, Parliament may by law provide that neither the Supreme Court nor any other court shall exercise jurisdiction in respect of any such dispute or complaint as is referred to in clause (1). The first clause is an enabling power over adjudication. The second is something rarer in the Constitution: a licence to oust the jurisdiction of the courts, including the constitutional courts, over a defined subject. Parliament used both clauses in the single statute. Section 11 of the 1956 Act is clause (2) at work, and the tribunal machinery is clause (1) at work.

Why does the Constitution keep river water disputes away from the ordinary courts?

The Constitution keeps these disputes away from the ordinary courts because a dispute between two States over a shared river is polycentric: the allocation that suits one State alters cropping, power generation, drinking water and ecology in the other, and the remedy has to be administered season after season rather than pronounced once.

The data are hydrological and contested, and Article 262 therefore lets Parliament substitute a specialist tribunal and bar the courts, so that investigation and continuing adjustment replace pleadings about proprietary rights.

The deeper reasoning deserves a fuller statement, because the design is often misdescribed as a mere distrust of judges. The framers’ difficulty was that a water dispute between States does not present a court with a stable res to distribute. Rainfall varies, utilisation grows, projects are built in stages, and a decree framed on one decade’s flows becomes a grievance in the next. An adjudication that cannot be reopened will age badly; an adjudication that can be reopened at will is no adjudication at all. The tribunal model was Parliament’s attempt to locate the decision in a body that could investigate like a commission and decide like a court, staffed by judges to command confidence and assisted by technical assessors to handle the hydrology. The bar on court jurisdiction was meant to protect that hybrid body from being second-guessed at every interlocutory step by the very litigation it replaced. Whatever its failures in practice, the theory is coherent, and any fair account of the Act has to grant the theory before indicting the execution.

Around Article 262 sit the two legislative entries that create the permanent tension of Indian water law, and the pillar reader needs them stated exactly. Entry 17 of the State List gives the States legislative power over water, that is to say, water supplies, irrigation and canals, drainage and embankments, water storage and water power, but it does so expressly subject to the provisions of Entry 56 of the Union List. Entry 56 of the Union List gives Parliament power over the regulation and development of inter-State rivers and river valleys to the extent to which such regulation and development under the control of the Union is declared by Parliament by law to be expedient in the public interest. The State entry is the default: water is a State subject. The Union entry is the override, but it is an override that sleeps until Parliament wakes it by a declaration of expediency. The distribution of legislative fields here is a question about which legislature may act and how far, and it is kept distinct in this series from the different doctrine that governs what happens when both legislatures have acted and their laws collide, which is the doctrine of repugnancy under Article 254 handled in the article on Article 254 and repugnancy between laws. A full side by side treatment of the Union, State and Concurrent fields sits in the comparison of the three Lists of the Seventh Schedule, at the entries this pillar is built on.

Parliament has used its Entry 56 power with extraordinary restraint. The River Boards Act, 1956 contains, in its Section 2, a declaration that it is expedient in the public interest that the Central Government should take under its control the regulation and development of inter-State rivers and river valleys to the extent provided in that Act. Beyond that self-limiting declaration, Parliament has never enacted a law taking the regulation and development of a particular inter-State river under the control of the Union. The consequence is a constitutional structure with a strong centre that declines to use its strength: the Union may enter the field at will, the States hold the field in practice, and the only Union instrument in regular operation is the adjudication statute that Article 262, a provision outside the Seventh Schedule entirely, makes possible. The 1956 disputes Act does not rest on Entry 56 at all. It rests on Article 262, which is why the Act can coexist with an unrepealed State primacy over water: it adjudicates disputes about the waters without transferring the subject of water itself.

The phrase “subject to the provisions of Entry 56” in Entry 17 carries a consequence worth stating separately, because it is the point on which the competence analysis turns. A State’s legislative power over water is not suspended merely because a river crosses a boundary: the State may legislate on the inter-State river’s waters within its territory, build works, run canals and levy its charges, and it does so every day. What the subordination means is that the field is held on a defeasible title: to the extent Parliament declares Union control expedient and legislates under Entry 56, the Union’s law occupies the ground, and State action inconsistent with it must yield. Since Parliament has made no river-specific declaration in general form, the defeasance has remained theoretical for particular rivers, and State primacy on inter-State rivers is a fact of practice resting on a condition of law. The distinction explains why the disputes Act, which does not transfer the subject at all, was the instrument the Union reached for: adjudication under Article 262 settles the quarrel while leaving every State’s Entry 17 title exactly where the Constitution put it, which is why States that would resist an Entry 56 law with everything they have will appear before a tribunal and argue their title instead.

One more structural point belongs in the design section, because it explains a recurring confusion. Article 131 gives the Supreme Court original jurisdiction over disputes between States involving legal rights, and a reader might expect inter-State water quarrels to arrive there as suits. Two features of the design prevent that for a dispute that falls within the 1956 Act. The first is Section 11, which bars the Supreme Court and every other court from exercising jurisdiction in respect of any water dispute which may be referred to a tribunal under the Act, and the definition of a water dispute in Section 2(c) is wide enough to cover the ordinary allocation quarrel. The second is that Article 262(2) was written precisely to let Parliament enact such a bar notwithstanding the rest of the Constitution, Article 131 included. The boundary of the bar matters, and the finality section of this article returns to it, because the cases in which the Supreme Court has heard water matters are cases that fall on one side or the other of the statutory definition, or that arrive after the tribunal’s work is done.

One neighbouring institution belongs in the design map because it is the road the Constitution itself signposts for inter-State friction and because its non-use for water is part of the same pattern as the river boards. Article 263 empowers the President to establish an Inter-State Council charged with inquiring into and advising upon disputes which may have arisen between States, and with discussing subjects of common interest. The Council was established in 1990, following the Sarkaria Commission’s recommendation, with the Prime Minister as its chair and Chief Ministers among its members. It is a deliberative and recommendatory body: it decides nothing, allocates nothing and binds nobody. Successive governments have preferred, for water, the tribunal route that produces a binding decision over the council route that produces advice, which is a defensible preference in a single case and a revealing one across seven decades. The federation has built three instruments for inter-State water conflict, a planning board system, a council of governments and a tribunal system, and has used, in anger, only the one that behaves most like litigation.

The machinery: from a State’s complaint to a tribunal’s report

The Act’s machinery is a sequence of gates, and at every gate a different actor holds the key. Reading the sections in order, rather than as a list of powers, shows where the system’s tempo is really set. Section 2(c) defines the subject matter. Section 3 opens the door. Section 4 decides whether anyone walks through it. Section 5 runs the hearing and writes the clocks. Nothing in the machinery is self-starting: the Act creates no standing tribunal, no registry of disputes and no obligation on any government to complain, and that absence is itself a design choice with consequences the amendment section takes up.

The definition comes first. A “water dispute” under Section 2(c) is any dispute or difference between two or more State Governments with respect to the use, distribution or control of the waters of, or in, any inter-State river or river valley. The definition reaches three situations: a quarrel over the waters themselves; a quarrel over the interpretation of the terms of an agreement relating to such waters, or over the implementation of such an agreement; and the levy of any water rate in contravention of the prohibition contained in Section 7. Two boundaries follow from the text. The disputants must be State Governments: a farmer, a municipality, a power utility or a private project company has no standing to set the Act in motion, however directly its interests are affected, and its remedy lies in ordinary litigation or in persuading its own government to complain. And the subject must be the waters of an inter-State river or river valley: a river wholly within one State, however fiercely contested inside that State, is outside the Act altogether.

Section 3 states the complaint. If it appears to the Government of any State that a water dispute with the Government of another State has arisen, or is likely to arise, because the interests of the State or of any of the inhabitants thereof in the waters of an inter-State river or river valley have been, or are likely to be, affected prejudicially, the State Government may, in the form and manner prescribed, request the Central Government to refer the water dispute to a tribunal for adjudication. The prejudicial effect may flow from three kinds of conduct by the other State: executive action or legislation taken, passed or proposed; the failure of the other State or any authority in it to exercise its powers over the use, distribution or control of the waters; or the failure of the other State to implement the terms of an agreement relating to those waters. The drafting is deliberately wide. A dispute need not have matured into injury: a proposed project, an announced law, or an omission to act is enough to found a complaint, and the complaint lies where the interests of inhabitants, not only of the State as a government, are prejudicially affected.

What must a State’s complaint under Section 3 actually show?

A complaint under Section 3 must show that a water dispute has arisen or is likely to arise with another State Government, and that the interests of the complaining State or of its inhabitants in the waters of an inter-State river or river valley have been or are likely to be prejudicially affected.

The prejudice must be traced to the other State’s action, its legislation, its inaction or its failure to honour an agreement. The threshold is one of apprehended prejudice, not proof of completed injury.

Between the complaint and the opinion sits the least visible proceeding in the Act’s life, conducted without a statutory script. The complaint is addressed to the Central Government and examined within the Union administration, in practice through the ministry responsible for water resources, which takes technical advice from the Central Water Commission, the Union’s apex technical body for water planning, on the hydrological substance of what is alleged. The negotiation the statute contemplates is then pursued through the instruments of federal diplomacy: correspondence between Chief Ministers, meetings of officials, and, in the harder disputes, Union ministerial mediation between the riparian governments. None of this is in the Act. The Act knows only the request and the opinion, and the space between them is filled by convention, which is why two complaints identical in law can have wholly different biographies: one ripening into a tribunal within a few years because the political configuration favoured reference, another ageing in the negotiation stage because it did not. The rules made under Section 13 regulate the complaint’s form and the tribunal’s procedure once constituted. No rule can regulate the diplomacy, and none pretends to.

The complaint, though, decides almost nothing by itself. Section 4(1) provides that when any request under Section 3 is received from any State Government in respect of any water dispute, and the Central Government is of opinion that the water dispute cannot be settled by negotiations, the Central Government shall, within a period not exceeding one year from the date of receipt of the request, by notification in the Official Gazette, constitute a Water Disputes Tribunal for the adjudication of the water dispute. Every word of the condition matters. The duty to constitute arises only when the Centre forms an opinion about the prospects of negotiation. The statute prescribes no period within which that opinion must be formed, no form in which the negotiations must be attempted, no record that must be kept of them, and no consequence if the opinion is never formed. The one year period, inserted by the 2002 amendment and examined in the amendment section below, disciplines the interval between the opinion and the notification. It does not discipline the interval between the request and the opinion, which is where governments have historically waited, sometimes for years, while the Centre pursued settlements that the complaining State had already concluded were hopeless. A proviso added in 2002 adds a rule of repose: any dispute settled by a tribunal before the commencement of the 2002 amendment shall not be reopened.

Once the tribunal is constituted, Section 5(1) obliges the Central Government to refer the water dispute to it for adjudication, together with any matter appearing to be connected with or relevant to the dispute, subject to the prohibition in Section 8. The reference defines the tribunal’s remit, and the tribunal’s own powers are then set by Section 5(2) and (3). Section 5(2), as substituted in 2002, requires the tribunal to investigate the matters referred to it and to forward to the Central Government a report setting out the facts as found by it and giving its decision on the matters referred to it within a period of three years. The proviso permits the Central Government, if the decision cannot be given for unavoidable reasons within the three years, to extend the period for a further period not exceeding two years. Section 5(3) creates the reference back: if, on considering the tribunal’s decision, the Central Government or any State Government is of opinion that anything in it requires explanation, or that guidance is needed on any point not originally referred, the Centre or that State may, within three months from the date of the decision, refer the matter again to the tribunal for further consideration; the tribunal may then forward a further report within one year from the date of the reference, giving the explanation or guidance it deems fit, and the decision stands modified accordingly; and the one year period for the further report may be extended by the Central Government for such further period as it considers necessary. Section 5(4), inserted in 1968, resolves internal disagreement: where the members differ, the point is decided by the opinion of the majority.

The architecture of periods repays attention because it distributes generosity unevenly. The constitution period is capped and short, one year, but it starts only after the unclocked opinion is formed. The reporting period is capped at three years with a capped extension of two, and the extension is not the tribunal’s to take: it must come from the Central Government, on the ground of unavoidable reasons. The reference back is tightly clocked at the front, three months for the government that wants it, and loosely clocked at the back, one year for the further report extendable by the Centre without any statutory ceiling on the extension. Then the sequence leaves the Act’s clocks behind entirely: Section 6 obliges the Central Government to publish the decision in the Official Gazette, and prescribes no period for publication at all. The step that converts a report into a binding decision is the one step the statute leaves wholly to the publishing government’s sense of timing. How these periods have performed in practice, tribunal by tribunal, is the record owned by the study of why water dispute tribunals take so long, and this pillar states the periods only as the statute prescribes them.

The findable artifact for this pillar follows the same sequence and marks, at each step, who acts, what period the statute prescribes, and whether that period carries any enforcement. The enforceability column is the article’s argument in miniature: the Act’s clocks are real where they bind governments seeking something, and weakest where they bind the governments doing the constituting, extending and publishing.

Step	Who acts	What the statute requires	Statutory period	Is the period enforceable?
Complaint and request, Section 3	A State Government	A request in the prescribed form where the interests of the State or its inhabitants are prejudicially affected	None	No clock runs; the Act obliges no State to complain
Opinion on negotiations, Section 4(1)	The Central Government	Formation of the opinion that the dispute cannot be settled by negotiations	None prescribed	No period and no consequence stated; the Act gives the State no lever of its own
Constitution of the tribunal, Section 4(1)	The Central Government, on nominations by the Chief Justice of India	Notification in the Official Gazette constituting the tribunal	Within one year from the date of receipt of the request, inserted in 2002	A stated period with no consequence for breach written into the Act
Reference of the dispute, Section 5(1)	The Central Government	Reference of the dispute and connected matters to the tribunal	None	No separate clock; the tribunal cannot begin without the reference
Investigation and report, Section 5(2)	The tribunal	A report setting out the facts found and the decision	Three years, extendable by the Central Government by up to two years for unavoidable reasons	The cap binds, but the extension decision rests with the Centre, not the tribunal
Reference back, Section 5(3)	The Central Government or a State Government	Explanation sought, or guidance on a point not originally referred	Within three months from the date of the decision	The window binds the government seeking the reference
Further report, Section 5(3)	The tribunal	A further report giving the explanation or guidance; the decision stands modified accordingly	One year, extendable by the Central Government for such further period as it considers necessary	The extension carries no statutory ceiling
Publication, Section 6(1)	The Central Government	Publication of the decision in the Official Gazette, whereupon it is final and binding	None	No period stated; finality begins only on publication
Scheme for implementation, Section 6A	The Central Government	A scheme, which may establish an authority, laid before Parliament	None	Timing discretionary; Parliament may modify or annul the scheme
Special leave, Article 136	The Supreme Court	Appeals against the published decision, held maintainable in the Cauvery proceedings	Governed by the Constitution, not the Act	Outside the Act entirely; the Act’s bar does not reach it

The tribunal itself: composition, powers and working

A Water Disputes Tribunal under Section 4(2) consists of a Chairman and two other members, nominated by the Chief Justice of India from among persons who, at the time of nomination, are Judges of the Supreme Court or of a High Court. The composition rule does two things at once. It judicialises the tribunal completely at the top: every member is a sitting judge, so the body that hears a water dispute is a bench in everything but name. And it removes the appointment from the political executive’s discretion in form, since the names come from the Chief Justice of India, though the power to decide whether and when the tribunal exists at all remains with the Central Government under Section 4(1). Section 4(3) adds the technical layer: the Central Government may, in consultation with the tribunal, appoint two or more persons as assessors to advise the tribunal in its proceedings. The assessors advise; they do not decide. The hydrology, the engineering and the agricultural economics enter the record through assessors, expert evidence and the tribunal’s own investigation, while the decision remains judicial in form and in signature.

Section 5A, inserted by the amending Act of 1968, keeps the body alive across vacancies: if a vacancy other than a temporary absence occurs in the office of the Chairman or any member, it is filled by a nominee of the Chief Justice of India in accordance with Section 4(2), and the investigation may continue from the stage at which the vacancy occurred. Without such a provision, the death or retirement of a single member in a proceeding measured in years could force a rehearing of a record built over a decade. Section 12 governs the other end of the life cycle: the Central Government shall dissolve the tribunal after it has forwarded its report and as soon as the Centre is satisfied that no further reference to the tribunal in the matter would be necessary. Dissolution, like constitution, is a Central Government act, and its timing follows the Centre’s satisfaction rather than any event in the tribunal’s own calendar.

The tribunal’s powers under Section 9 are the powers of a civil court under the Code of Civil Procedure, 1908 in specified matters: summoning and enforcing the attendance of any person and examining him on oath; requiring the discovery and production of documents and material objects; requisitioning any data the tribunal requires, a clause added in 2002; issuing commissions for the examination of witnesses or for local investigation; and any other matter prescribed. Section 9(2) lets the tribunal require any State Government to carry out, or permit to be carried out, surveys and investigations the tribunal considers necessary. The costs power in Section 9(3) gives a decision an unusual tail: the tribunal may direct which government pays the expenses of the tribunal and the costs incurred by a State Government appearing before it, and the direction, so far as it relates to expenses or costs, may be enforced as if it were an order made by the Supreme Court. Section 9(4) leaves the tribunal to regulate its own practice and procedure, subject to the Act and the rules. Those rules are made by the Central Government under Section 13 after consultation with the State Governments, covering the form of complaints, the matters in which the tribunal holds civil court powers, procedure, remuneration and service terms.

What does a tribunal do besides decide, and what powers does it hold while deciding?

A tribunal under this Act is an investigating body as much as a deciding one. Section 5(2) charges it to investigate the matters referred and to report the facts as found alongside its decision, and Section 9 gives it the powers of a civil court while it works.

Those powers run to summoning witnesses, compelling documents, requisitioning data and commissioning local investigations, and Section 9(2) lets the tribunal direct State Governments to carry out surveys. Its record is built, not merely received.

Two ancillary provisions complete the working picture, and both matter to how the statute behaves. Section 7 polices a temptation of upstream States: no State Government shall, merely because works for the conservation, regulation or utilisation of the waters of an inter-State river have been constructed within its limits, impose any seigniorage or additional rate or fee on the use of that water by another State or its inhabitants, and Section 7(2) deems a dispute over such a levy to be a water dispute within the Act. Section 8 is the Act’s one internal exclusion, and it is a bridge to the companion statute: notwithstanding Sections 3 and 5, no reference shall be made to a tribunal of any dispute regarding any matter which may be referred to arbitration under the River Boards Act, 1956. Parliament imagined a division of labour in which some categories of disagreement would go to arbitration connected with the boards, and fenced those categories out of the tribunal route. The boards were never created and the arbitration route never operated, so Section 8 survives as a fossil of the original design, an exclusion keyed to machinery that has never existed.

How a tribunal proceeding actually runs

A proceeding before a tribunal under this Act looks like a suit and behaves like an inquiry, and the blend explains both its strengths and its tempo. The reference under Section 5(1) sets the agenda: the dispute as complained of, together with connected matters the Central Government includes. The States then plead against one another as parties, filing statements of case that set out their claims to the waters, the projects they have built or propose, the utilisation they say is established and the injury they say a rival’s works will cause. Issues are framed. Evidence follows, and it is here that the proceeding departs furthest from an ordinary trial, because the decisive material is not testimony about events but series: gauge and discharge observations across decades, rainfall records, crop water requirements, evaporation losses, the live and gross storage of reservoirs, and the dependability analysis that converts a variable river into an allocable quantity. Expert witnesses, engineers and hydrologists in the main, carry this material, and the assessors appointed under Section 4(3) exist to help the bench interrogate it.

The tribunal’s investigatory charter in Section 5(2), to investigate the matters referred and to report the facts as found, gives it tools a court would not ordinarily use on its own motion. Under Section 9 it may commission local investigations, require surveys by State Governments, requisition data after 2002 in terms, and summon records from governments that are also its litigants, a combination that makes the proceeding partly adversarial and partly inquisitorial. Interim applications punctuate the main proceeding: a State that fears a rival’s project will present it with a fait accompli seeks interim directions restraining construction or regulating releases while the adjudication proceeds, and the tribunal’s power to make such orders, denied any express mention in the Act, was established through the Cauvery interim proceedings discussed in the finality section below. The interim stage deserves emphasis because in a long adjudication it is frequently the operative law of the river for years at a time. A tribunal that cannot regulate the status quo while it deliberates invites the parties to manufacture facts: storages get built, ayacuts get extended, and the established utilisation that equity will later protect is created by the very party the interim order was meant to restrain. The Cauvery interim order of 1991, with its annual release figure and its restraint on the extension of irrigation, governed the basin’s releases through the whole decade of the main proceeding, and its defiance by ordinance produced the Presidential reference that fixed the tribunal’s interim authority in constitutional law. Later tribunals have worked in that authority’s shelter, regulating releases during drought years and restraining unilateral works while the merits were heard, so that the Act’s unwritten interim jurisdiction is, in practical terms, among its most used provisions. Site inspections by the bench or its commissions, a familiar feature of the long proceedings, belong to the same investigatory conception: the river, its headworks and its command areas are treated as evidence to be viewed, not merely described.

The proceeding ends, when it ends, in the dual instrument Section 5(2) prescribes: a report that sets out the facts as found and gives the decision. The dual character matters. The facts as found, on flows, utilisation and need, are the tribunal’s permanent contribution, because a later tribunal for the same basin, or a court on special leave, starts from the record the first proceeding built. The decision is the operative allocation: the shares, the release schedules, the project clearances or restraints, and the machinery directions that a Section 6A scheme may later embody. States dissatisfied with the report’s reasoning but unable to disturb its operative allocation discover that the Act offers only two continuations: the reference back under Section 5(3), confined to explanation and to points not originally referred, and the constitutional route under Article 136 after publication. There is no appeal on the merits to any body the Act creates. The absence is deliberate, and it is the procedural face of the finality the next section examines.

Finality and its limits: publication, schemes, the ouster and Article 136

A tribunal’s report is not yet the law of the river. Section 6(1) provides that the Central Government shall publish the decision of the tribunal in the Official Gazette, and that the decision shall be final and binding on the parties to the dispute and shall be given effect to by them. Publication is the operative act: until the Gazette notification appears, the parties hold a report, and after it appears, they hold a decision that the statute commands them to implement. Section 6(2), inserted by the 2002 amendment, then raises the decision’s juridical rank: the decision of the tribunal, after publication in the Official Gazette, shall have the same force as an order or decree of the Supreme Court. The pairing is deliberate. Finality language tells the States the matter is closed. Decree language tells everyone what the closed matter is worth in the hierarchy of legal instruments: not a recommendation, not an arbitral award of the ordinary kind, but an instrument carrying the authority of the highest court’s own orders.

Publication alone, however, settles shares on paper. Water is delivered through headworks, regulators, gauges and operating schedules, season by season, and a tribunal’s allocation has to be translated into an administrative apparatus before it binds a single cusec. That translation is the work of Section 6A, inserted by the amending Act 45 of 1980 after the Narmada tribunal’s decision showed that an award of this kind may require an inter-State authority of its own to be real. Section 6A empowers the Central Government, by notification in the Official Gazette, to frame a scheme or schemes making provision for all matters necessary to give effect to the decision of a tribunal. A scheme may establish an authority, by whatever name, for the implementation of the decision, and may settle its composition, jurisdiction, powers, functions, service terms, meeting discipline, staffing, funds and accounts, and may constitute a committee to review specified decisions of the authority. The scheme has effect notwithstanding anything in any other law in force, other than the Act itself, and every scheme, and every regulation made under it, is laid before each House of Parliament, which may modify it or resolve that it should not be made. The Narmada Control Authority is the model instance the 1980 Statement of Objects and Reasons had in view: the implementation machinery that the Narmada decision envisaged needed a statutory route to exist as a legal entity, and Section 6A is that route, generalised for every tribunal that followed.

Against this architecture of finality stands the ouster. Section 11 provides that, notwithstanding anything contained in any other law, neither the Supreme Court nor any other court shall have or exercise jurisdiction in respect of any water dispute which may be referred to a Tribunal under this Act. The words track Article 262(2) and are as complete as statutory language can make them. If the text were the whole story, a water dispute would enter the tribunal and never see a courtroom. The text is not the whole story, for two reasons, one internal to the definition and one constitutional.

The internal reason is that the bar attaches to a “water dispute”, and litigation has a way of presenting itself in other costumes. A suit about the safety of a dam, or about the validity of a State law that rewrites an arrangement a court has already upheld, may be framed, and has been framed, as something other than a dispute over the use, distribution or control of waters. In the Mullaperiyar litigation decided in 2014, the Supreme Court entertained an original suit by Tamil Nadu against Kerala’s dam safety legislation and struck down the State law, in proceedings the Court treated as lying outside the bar because the controversy concerned the safety of the dam and the parties’ rights under an earlier adjudication, not the allocation of the river’s waters. The line the Court polices is the line of Section 2(c): allocation quarrels go to tribunals; quarrels that genuinely concern other legal rights may still be suits, and each side in a hard case will argue the characterisation that opens or closes the courtroom door.

The ouster was tested once more, earlier and in a different posture, through the Presidential reference on the Cauvery interim order. The Cauvery tribunal, in June 1991, passed an interim order directing Karnataka to release 205 thousand million cubic feet of water annually to Tamil Nadu while the adjudication proceeded. Karnataka responded with an ordinance, the Karnataka Cauvery Basin Irrigation Protection Ordinance, 1991, which purported to nullify the interim order and to forbid compliance with it. The President referred questions to the Supreme Court under Article 143, and the Court’s opinion in In re Cauvery Water Disputes Tribunal held the ordinance unconstitutional: a State legislature cannot, by its own enactment, override the decision of a tribunal constituted under a parliamentary statute and seize the adjudicatory function to itself. The opinion also sustained the tribunal’s power to pass the interim order as an incident of its adjudicatory function under the Act. Two consequences of that opinion run through the later law. A State’s remedy against a tribunal order it dislikes is never self-help by legislation. And the tribunal’s authority during the long pendency of a reference, including interim regulation of releases, rests on ground the Court has already defended.

The constitutional reason is Article 136, and it is the article’s second structural finding, so it must be stated narrowly and exactly. Article 136 empowers the Supreme Court, in its discretion, to grant special leave to appeal from any judgment, decree, determination, sentence or order of any court or tribunal in the territory of India. Section 11 does not mention Article 136, and a statute enacted under Article 262(2) speaks to jurisdiction in respect of a water dispute. When the States aggrieved by the Cauvery tribunal’s final order of 2007 carried appeals to the Supreme Court, the Union of India objected at the threshold that the appeals were not maintainable at all: the award, it argued, was final, carried the force of a decree of the Supreme Court itself, and lay beyond appellate correction by the very design of the Act. A three judge bench of Dipak Misra, Amitava Roy and A. M. Khanwilkar, JJ. rejected the objection in December 2016. The reasoning, stated at its narrowest, is that once a water dispute as defined has been adjudicated by the tribunal, it loses the character of a water dispute, and the grievance that survives is a grievance against a determination by a tribunal, which Article 136 reaches in the ordinary way; Sections 6(2) and 11 therefore do not bar the special leave jurisdiction. The Court then heard the appeals on their merits and, on 16 February 2018, delivered judgment modifying the tribunal’s allocation in part, enhancing Karnataka’s share and adjusting Tamil Nadu’s correspondingly, while affirming the tribunal’s work in large measure, and directed the Central Government to frame a scheme under Section 6A within six weeks so that the decision as modified would be smoothly made functional. The Cauvery Water Management Authority followed under that direction, and the tribunal itself was dissolved in July 2018 once the Centre was satisfied no further reference was necessary.

Why did Parliament give the award the force of a Supreme Court decree in 2002 rather than create an enforcement agency?

The 2002 amendment chose juridical rank over administrative machinery. A decree of the Supreme Court carries contempt and execution associations that no new agency could have been given by a short amendment, and it cost Parliament nothing to administer and nothing to staff.

The choice also expressed a theory: that defiance of an award was a problem of legal status, curable by raising the award’s status, rather than a problem of operating works, which a status change cannot reach.

The consequence of the Cauvery rulings for the design can be stated plainly. The finality the Act promises is finality against every forum except one: the tribunal’s decision concludes the dispute as a dispute, bars the High Courts and the subordinate courts, bars a fresh suit under Article 131 on the same allocation, and binds the parties as a decree binds. It does not conclude the matter against the Supreme Court’s discretionary appellate power over determinations of tribunals. Readers who meet the common claim that the Supreme Court has no jurisdiction at all in water disputes are meeting only half the statute. The accurate statement is that the Supreme Court has no original or ordinary jurisdiction over a water dispute referable to a tribunal, and that it has heard and decided the most consequential water appeal of the tribunal era, because Article 136 was held to survive the ouster. Nor is special leave the only door the Court has used in water matters: it has entertained writ proceedings touching drinking water and the right to life, and Presidential references and interlocutory proceedings in pending disputes have kept the Court in continuing contact with rivers whose allocation formally belongs to tribunals. The design removed the dispute from the courts. It did not, and on the Cauvery reasoning could not, remove the tribunal’s output from the constitutional court’s supervision.

The Narmada litigation supplies the necessary counterweight, because it shows the same Court declining to do for the Narmada what it later did for the Cauvery. The Narmada tribunal’s 1979 decision was challenged in proceedings under Article 32 in Narmada Bachao Andolan v Union of India, decided in 2000, where the petitioners attacked the Sardar Sarovar project that the award had authorised. The Supreme Court refused to sit in judgment over the award: the tribunal’s decision, rendered under a statute the Constitution itself licensed, was binding on the parties, and the Court would not reopen the allocation or the height the tribunal had fixed, though it supervised the resettlement obligations that accompanied construction. Read together, the Narmada and Cauvery positions mark the true boundary of the ouster. The Court will not receive a collateral attack on an award dressed as a writ petition, and it will not entertain a suit that re-agitates the allocation as a dispute. What it retains is the special leave jurisdiction over the tribunal’s determination itself, exercised as an appeal, with the discipline that the tribunal’s fact-finding commands respect. Finality under the Act is therefore real but appellate-shaped: one correction, at the apex, on the record the tribunal built.

A word on what enforcement means against a State, since the decree language of Section 6(2) invites the question. A decree is executed against a judgment debtor; a State is not an ordinary judgment debtor, and no bailiff distrains a head regulator. In practice the published decision is enforced by the same instruments that produced it: the Section 6A authority’s operating schedules, the accounting of releases at border gauges, and, where a State resists, the Supreme Court’s continuing supervision in the proceedings before it, exercised through compliance directions of the kind the Cauvery appeals generated after the 2018 judgment. The tribunal’s own costs direction under Section 9(3), enforceable as if it were an order of the Supreme Court, shows the Act reaching for the same device at small scale: where the statute wants a money consequence to bite, it borrows the Court’s authority rather than building an execution machinery of its own. The system’s enforcement is therefore supervisory and political before it is coercive, which is either its realism or its evasion, depending on the reader’s theory of federal compliance; the text supports both descriptions, and this article records the mechanism without pretending the mechanism is a remedy a private decree holder would recognise.

There is a final quiet provision in this part of the Act. Section 6A(7), as enacted in 1980, subjected every scheme to Parliament’s power of modification or annulment, and the Cauvery tribunal itself observed that the insertion of Section 6(2) in 2002, giving the published decision the force of a Supreme Court decree, sits awkwardly with a parliamentary power to modify the scheme that implements that decision. The observation records a genuine tension rather than a resolved doctrine: a decision with the force of a decree is implemented through an instrument Parliament may rewrite. This article states the tension because the tribunal stated it, and takes it no further, since no court has dismantled either provision.

The boundary in action: Ravi and Beas and the Sutlej-Yamuna Link

One basin tests every boundary this article has drawn, and it does so because its quarrel is only partly about allocation. The Ravi and Beas waters were apportioned in the era of the Punjab Settlement of July 1985, the accord between the Union and Punjab’s leadership that sought to close the State’s insurgency-era grievances; paragraph 9.1 of the Settlement provided for the adjudication of the river waters claims of Punjab and Haryana, and paragraph 9.2 concerned the construction of the Sutlej-Yamuna Link canal, the work that would carry Haryana’s share southward through Punjab’s territory. Parliament’s answer inside the Act was Section 14, already described: a special tribunal for the Settlement’s referred matters, with the Central Government alone empowered to make the reference. The Ravi and Beas Waters Tribunal, constituted in April 1986, reported in January 1987 on the allocation, verifying the parties’ claims against the flow series. The Central Government then referred the matter back for explanation and guidance on the points the report had left, and there the tribunal proceeding has rested: the further report under Section 5(3) has never been given, the tribunal has never been dissolved, and the allocation question the tribunal answered in 1987 remains formally open in a way no other award in the Act’s history quite matches.

The canal question travelled a wholly different road, and the road it travelled defines Section 11’s boundary from the other side. The Sutlej-Yamuna Link canal’s Punjab reach was a construction obligation arising from agreements and from the Settlement, and when Punjab declined to complete it, Haryana sued in the Supreme Court under Article 131. The Court decreed the suit in January 2002, directing Punjab to complete the canal within a year, and returned in June 2004 to reject Punjab’s defences and press the decree. These proceedings were not barred by the 1956 Act because their subject was not the use, distribution or control of waters awaiting adjudication: it was the enforcement of obligations the Court found already fixed by agreement, by the Settlement and by its own decree. Punjab’s legislature answered the decrees with the Punjab Termination of Agreements Act, 2004, which purported to terminate the State’s water agreements and discharge its obligations. The President referred the Act’s validity to the Supreme Court under Article 143, and the Court’s opinion of November 2016 held the device impermissible: a State legislature cannot by unilateral enactment nullify the decrees of the Supreme Court or escape obligations adjudicated under them. The canal, as a physical work, has its own long administrative history beyond this article’s scope; the legal architecture is the pillar’s concern, and it is complete. Within one basin, the allocation went to a tribunal under the Act and remains in the Act’s unfinished business, while the construction obligation founded decrees under Article 131 that the State could not legislate away. The boundary of Section 11 is therefore not a wall around a river. It is a line around a subject: the undecided allocation. What has been decided, by tribunal or by court, leaves the shelter of the bar and enters the ordinary world of binding instruments, where decrees are executed and nullifying statutes fall.

The law a tribunal applies: equitable apportionment and the national asset

The Act tells a tribunal how to sit, what to investigate and when to report. It nowhere tells the tribunal by what principles to divide a river. There is no statutory formula, no schedule of factors, no priority of uses enacted in Section 5 or anywhere else in the Act, and that silence is one of its most consequential features. The substantive law of Indian inter-State water allocation is therefore tribunal-made and court-approved: a body of principle assembled from international practice, from the reasoning of successive tribunals, and from the Supreme Court’s treatment of their awards. A pillar account of the statute must state that law, because a reader who knows the machinery but not the principles knows how a dispute is processed without knowing how it is likely to be decided.

The governing principle is equitable apportionment. As the Supreme Court restated it in the Cauvery judgment of 2018, the waters of an inter-State river do not belong to any one State: no State can claim exclusive ownership of, or a proprietary right in, the waters flowing through its territory, because an inter-State river is a national asset held for the people of the basin as a whole, and the riparian States stand toward it in a relation the Court compared to that of co-owners or trustees rather than absolute proprietors. It followed, in that case, that Karnataka’s assertion of an ownership-based claim over the Cauvery waters within its territory was misconceived, and that Tamil Nadu’s correlative assertion, that established downstream use had crystallised into an indefeasible right immune from reallocation, was equally unsustainable. Equitable apportionment distributes the resource according to the equity of the whole basin, weighing each State’s contribution to the flow, its drainage area, its dependent population, its established utilisation, the availability of alternative resources including groundwater, and the comparative efficiency and economy of competing uses. The factors recall those catalogued in the Helsinki Rules on the uses of the waters of international rivers, adopted by the International Law Association in 1966, which Indian tribunals and the Supreme Court have treated as persuasive guidance for the domestic problem, the international materials being adapted rather than applied.

Beneath the principle sits a technical device every reader of awards must understand: dependability. A river’s flow is a series, not a quantity, and an allocation expressed in thousand million cubic feet is meaningful only against the probability with which that flow will occur. Tribunals therefore work with dependable flows, the flow equalled or exceeded in a stated proportion of years, and the choice of benchmark changes the size of the cake. The first Krishna tribunal worked substantially at seventy-five percent dependability, the conservative benchmark that allocates only the flow available in three years out of four; the Cauvery tribunal worked at fifty percent dependability, allocating the median year’s flow and accepting that in drier years the shares must shrink pro rata through distress arrangements. Neither choice is prescribed by the Act. Each is a judgment about how much risk of shortage the allocation should build in, and the difference between the two benchmarks, applied basin by basin, is one reason awards for different rivers are not arithmetically comparable. The distress question, how a shortfall in a bad year is shared, has in practice mattered as much as the normal year shares: the Cauvery arrangements came to include a distress sharing formula precisely because the basin’s hard years, not its average years, generated the political crises.

Two further principles complete the working doctrine as the tribunals have built it. The first is the protection of established utilisation, held in balance rather than in command: uses built in reliance on past flows count heavily in equity, because command areas, towns and industries cannot be unbuilt, but they do not extinguish the upstream State’s right to a share of its own river for its own reasonable development. The second is the priority of drinking water. Across the awards and in the 2018 judgment, domestic and drinking needs are treated as the highest charge on the river, ranking ahead of irrigation and industry, and a State’s claim for its cities’ supply has been received with a sympathy not extended to claims for expanded irrigation. Around these principles the tribunals have developed the operational apparatus of the awards: monthly or ten-daily release schedules measured at inter-State border gauges, carryover rules for storage, obligations to maintain flows for the delta and for environmental purposes that later awards state with increasing explicitness, and review clauses that contemplate recalibration after a term of years. None of this is in the statute. All of it is the law a party actually faces, and the pillar that omitted it would describe a machine without describing what the machine produces.

Awards also operate at the level of the individual project, and for the engineers and governments who must build within them this is their most concrete function. An allocation in thousand million cubic feet authorises nothing by itself; the works that use the water, their storage heights, their canal capacities and their command areas, must be fitted inside the share, and tribunals have repeatedly been asked to rule on the conformity of particular projects with the basin’s equity. The second Krishna tribunal’s treatment of the Almatti dam is the standard modern instance: the height to which Karnataka might raise that dam on the Krishna was the single most contested engineering question in the proceeding, because every metre of height converted into storage that the downstream State read as interception of its share, and the tribunal’s decision permitting the dam to be taken to its full designed height of 524.256 metres, within the State’s allocated share, illustrates the tribunal’s double character in one ruling. It was simultaneously a decree about a structure, measurable in metres, and an application of equitable apportionment, measurable only in the reasoning that connected the structure to the share. Project-level rulings of this kind are why the reports reward close reading: the share is the headline, but the project conditions are the law on the ground.

The tribunals the Act has produced

Nine tribunals have been constituted under the Act across its life, and the census itself teaches the design. The first generation arrived together: the Godavari Water Disputes Tribunal in April 1969, the Krishna Water Disputes Tribunal in April 1969, and the Narmada Water Disputes Tribunal in October 1969, each for a great peninsular system whose sharing the 1956 reorganisation had made urgent. The thirteen year interval between the Act and its first tribunal is itself evidence about the machinery: the Act slept until negotiation, conducted through the Union’s planning and water ministries in the 1960s, had visibly failed on all three rivers at once. The disputes were the reorganisation’s children. Andhra Pradesh’s formation had joined Telugu-speaking districts of the old Hyderabad and Madras units into one claimant on the Krishna and the Godavari; the Bombay State’s division in 1960 had made Maharashtra and Gujarat separate riparians of the Narmada and the Tapi; Mysore, enlarged in 1956 into a State with upper catchment ambitions on both the Krishna and the Cauvery, was pressing to build. In each basin the new States inherited utilisations planned by predecessor units and promptly outgrew them, and the complaints of the later 1960s were the sound of the 1956 map meeting the 1956 rivers. Parliament’s Act had anticipated the quarrels; what it had not supplied was any pressure on governments to use it before the quarrels had hardened into programmes, a defect the census repeats in each generation. The Godavari tribunal reported in 1980, substantially upon agreements the States had reached among themselves in stages from 1975 onwards, which the tribunal examined, completed where the agreements ran out, and gave effect to; the proceeding is the Act’s clearest instance of a tribunal functioning as a scaffold for negotiation rather than as its replacement, and it decided a five State basin with a finality the more famous proceedings never achieved. The first Krishna tribunal, chaired by R. S. Bachawat, reported in 1976: it assessed the dependable yield of the Krishna at seventy-five percent dependability as 2,060 thousand million cubic feet and divided that quantity among Maharashtra, Karnataka and Andhra Pradesh, and its allocation governed the basin until the second Krishna tribunal was constituted in April 2004 to revisit the sharing in the light of grown utilisation and the first tribunal’s own review horizon; the second tribunal reported in 2010 and, on the references back carried to it by the parties, gave its further report in November 2013, leaving the architecture of its 2010 allocation intact while answering the points referred; and the Andhra Pradesh Reorganisation Act, 2014 later extended its term for project-wise allocations between the successor States of Andhra Pradesh and Telangana. The Krishna sequence is the Act’s only instance of a basin returning for a second full adjudication, and it demonstrates the design’s assumption that an award governs a phase of a basin’s development rather than its eternity: the first tribunal itself contemplated review after a term, and the second tribunal’s constitution in 2004 was the machinery working as its makers expected, even if little else about the timetable was. The Narmada tribunal gave its decision in 1979, fixing among other things the height of the Sardar Sarovar dam at a full reservoir level of 455 feet and apportioning the Narmada’s waters among Madhya Pradesh, Gujarat, Maharashtra and Rajasthan, and the implementation authority its award required became, as the finality section records, the occasion for Section 6A itself.

The second generation belongs to the northern and eastern rivers. The Ravi and Beas Waters Tribunal was constituted in April 1986 under the special provision of Section 14, which Parliament inserted to let the Central Government constitute a tribunal for the verification and adjudication of the matters referred to in paragraphs 9.1 and 9.2 of the Punjab Settlement of 1985; uniquely, under Section 14(3), the Central Government alone may refer those matters, on its own motion or at a State’s request. The tribunal reported in 1987, the Central Government referred the matter back for explanation and guidance, and the further report under Section 5(3) has remained the proceeding’s unfinished business across the decades since, with the tribunal never dissolved. The Cauvery Water Disputes Tribunal, constituted in June 1990 for Karnataka, Tamil Nadu, Kerala and the Union territory of Puducherry, passed the interim order of June 1991 that produced the Presidential reference described in the finality section, gave its final order in February 2007 apportioning the basin’s yield at fifty percent dependability and prescribing releases of 192 thousand million cubic feet to Tamil Nadu at the inter-State border, saw that order published in the Gazette in February 2013 after the Supreme Court, in the pending appeals, directed the Central Government to notify it, and then saw it carried through those appeals to the judgment of 16 February 2018, which reduced the Tamil Nadu figure to 177.25 thousand million cubic feet, raised Karnataka’s share correspondingly on the strength of Bengaluru’s drinking water needs and Tamil Nadu’s available groundwater, and directed the Section 6A scheme that produced the Cauvery Water Management Authority.

The third generation is the Act in its amended form. The Vansadhara Water Disputes Tribunal was constituted in February 2010 for Odisha and Andhra Pradesh and reported in 2017. The Mahadayi Water Disputes Tribunal was constituted in November 2010 for Goa, Karnataka and Maharashtra and reported in 2018. The Mahanadi Water Disputes Tribunal was constituted in March 2018 for Odisha and Chhattisgarh. Each of these tribunals was constituted after the 2002 amendment had written the one year constitution period and the three and two year reporting periods into the Act, and each is therefore a test of the amended design on its own terms. What the tests showed belongs to the delay record and is not this pillar’s to argue; the pillar records only that the amended clocks exist, that the tribunals of the third generation were born under them, and that the constitutional questions the Act raises, gatekeeping, finality and the unused boards, survived the amendment untouched.

The shadow system: settlement by agreement

Nine tribunals in the life of the Act is a small census against the number of Indian basins shared between States, and the discrepancy identifies the Act’s shadow system: the inter-State agreement. Most shared basins are governed, where they are governed at all, by instruments the States negotiated themselves, of which three show the range. The Bansagar agreement of 1973, among Madhya Pradesh, Uttar Pradesh and Bihar, apportioned the waters of the Son and created a control board by executive resolution to administer the joint project, an institution negotiated into existence decades before any statute was asked to create one. The Subarnarekha agreement of 1978, a tripartite settlement among Bihar, Odisha and West Bengal, apportioned that river’s waters among the three States and remains the basin’s governing instrument without any tribunal having sat on it. The Yamuna memorandum of 1994, already noticed for the board it created, apportioned the upper Yamuna’s surface flows among five basin units and a Union territory by consent. Each agreement performs the allocation a tribunal would have pronounced, and each lacks what a tribunal’s decision carries: the Gazette publication, the finality of Section 6, and the decree force of Section 6(2). An agreement binds as a compact between governments, honourable and ordinarily honoured, and its enforcement, if a party resiles, runs through the routes the Ravi and Beas history exposed: a complaint under this Act where allocation is reopened, or a suit for the obligation itself.

The relationship between the two systems is written into the Act’s own definition, and it is circular in an instructive way. A dispute over the interpretation or implementation of an agreement is itself a water dispute under Section 2(c), so an agreement that breaks down can be carried into the tribunal system its makers avoided; and a tribunal proceeding, as Godavari showed, can end by adopting agreements the parties reached in its shadow. The two systems are therefore not rivals but phases: negotiation first, in form as in the Act’s own sequence, adjudication when negotiation’s product decays or fails to arrive. The design’s wager, visible in 1956, was that the tribunal’s existence would discipline negotiation by giving every State a credible alternative to a bad settlement. The wager’s weakness is the gatekeeping this article has already examined: a threat to invoke a tribunal one cannot summon at will, against a gatekeeper with reasons to delay, disciplines nobody. The agreements that succeeded did so because the parties’ incentives aligned for reasons the Act did not create; the disputes that reached tribunals are the residue where alignment failed and the gate eventually opened.

What an award leaves behind: Section 6A in practice

An allocation is a schedule of quantities. A river, however, delivers water through works that someone must operate, and the distance between a schedule and an operating regime is where awards succeed or fail after the Gazette. Section 6A exists to cross that distance, and its three principal uses show the range of the power. The first use is the Narmada pattern, for which the section was written: the Narmada Control Authority, envisaged by the tribunal’s 1979 decision and constituted by the Central Government in 1980 as the inter-State administrative authority to secure compliance with the decision, with a Review Committee above it for the matters the decision reserved. The Authority is a permanent body of the party States and the Centre, and the Narmada system’s regulated operation across four States runs through it. The second use is the Cauvery pattern: after the Supreme Court’s 2018 judgment directed a scheme within six weeks, the Central Government framed the scheme establishing the Cauvery Water Management Authority, with a Cauvery Water Regulation Committee beneath it, to supervise the reservoirs’ operation and regulate releases in accordance with the decision as modified by the Court. The Authority’s creation converted a quarter century of ad hoc supervisory committees and court-monitored releases into a standing statutory mechanism. The third pattern is absence: for several awards no comparable standing authority was created under Section 6A at all, and implementation remained with the States’ own irrigation departments under the award’s terms, supervised, where supervision existed, by committees resting on agreement rather than on a scheme.

The contrast among the three patterns is the practical commentary on Section 6A’s drafting. The power is discretionary in form: the Centre “may” frame a scheme. Nothing in the section requires a scheme for every award, sets a period within which a scheme must follow publication, or prescribes the minimum content that would make a scheme an operating regime rather than a paper one. The tribunal may recommend machinery, as the Cauvery tribunal did when it proposed a management board in its final order, but the recommendation enters a scheme only if the Central Government frames one, and the scheme then travels to Parliament, which holds the modification and annulment power Section 6A(7) confers. A reader assessing an award’s real force should therefore ask three questions in sequence: has the decision been published, has a scheme been framed, and does the scheme create a body with the staff, funds and meeting discipline the section authorises. The Act’s history contains awards that clear the first gate and stall at the second, and the distance between a binding decision and a delivered cusec is measured exactly there.

Two statutes, one dead: the River Boards Act 1956

The namable claim of this article can be stated in one sentence. Parliament enacted an adjudication statute and a river boards statute in the same year to give effect to the Union’s role over inter-State rivers, and the boards statute has never been used, so India has spent the decades since resolving river disputes by litigation because the instrument designed to prevent them by planning was never brought into operation. The disputes Act is Act 33 of 1956, dated 28 August 1956. The River Boards Act is Act 49 of 1956, dated 12 September 1956, a fortnight younger than its sibling. Of the two, the younger has the stranger biography: a complete statutory scheme, twenty-eight sections in four chapters, in force for the whole of its life, under which the Central Government has established nothing at all.

The River Boards Act rests on the constitutional foundation the disputes Act does not use. Its Section 2 makes the Entry 56 declaration in terms: it is declared expedient in the public interest that the Central Government should take under its control the regulation and development of inter-State rivers and river valleys to the extent provided in the Act. Section 4 then empowers the Central Government, on a request from a State Government or otherwise, by notification in the Official Gazette, to establish a River Board for advising the Governments interested in relation to matters concerning the regulation or development of an inter-State river or river valley or any specified part of it, and for performing such other functions as the notification specifies; different boards may be established for different rivers. Every board is a body corporate with perpetual succession and a common seal. The proviso to Section 4(1) carries the provision on which the whole history turns: no such notification shall be issued except after consultation with the Governments interested, with respect to the proposal to establish the board, the persons to be appointed as members, and the functions the board may be empowered to perform. Consultation is the statutory condition. Consent is not.

Section 5 composes the board of a Chairman and such other members as the Central Government thinks fit to appoint, and Section 5(2) requires each member, in the Centre’s opinion, to have special knowledge and experience in irrigation, electrical engineering, flood control, navigation, water conservation, soil conservation, administration or finance. Section 13 lists the functions a board may be empowered to perform, and they are the functions of a basin planning authority: advising on the regulation, development and utilisation of the waters, on the conservation of the basin, on schemes for irrigation, water supply, hydroelectric power, navigation and flood control, and on the prevention of disputes themselves. Section 15 provides for the preparation of development schemes. Section 16 arms a board with general powers that include acquiring property and inspecting State waterworks within its area of operation, an inspection power whose existence on paper measures the distance between the statute’s imagination and its history: the works of every basin State were to be open to a board that was never appointed to inspect them. Section 22 completes the scheme by providing for arbitration of disputes between Governments with respect to a board’s advice or its implementation, which is the arbitration to which Section 8 of the disputes Act defers when it bars tribunal reference for matters referable to that arbitration. The remaining chapters deal with the board’s funds, accounts and audit, the rule-making power and the usual miscellaneous provisions, completing a statute that lacks nothing except a single exercise of its Section 4 power. Read together, the two Acts of 1956 describe a complete federal water architecture: boards to plan basins and head off quarrels, arbitration for quarrels about the planning, and tribunals under the disputes Act for the quarrels that survived both. Only the third storey was ever built.

What could a river board have done that a tribunal cannot?

A river board could have acted before a dispute existed. Its statutory functions run to advising the Governments interested on the regulation and development of a basin, preparing development schemes, and coordinating uses across State lines on a continuing basis.

A tribunal, by contrast, is constituted only after a complaint, decides the matters referred to it, and is dissolved. The board was designed as a standing institution for a river; the tribunal is an event in a river’s litigation history.

The explanation for the empty statute is political, and the record states it in those terms. The National Commission to Review the Working of the Constitution, whose report was submitted in 2002, recorded that the River Boards Act had remained a dead letter, and the reason generally given in the official literature is the unwillingness of State Governments to accept a Union-established body exercising planning and advisory functions over rivers they regard as their own constitutional field under Entry 17. The Centre, for its part, has never treated Section 4 as a power to be exercised over State reluctance: the consultation proviso gave reluctant States a procedural forum in which agreement could be withheld indefinitely without ever being formally refused, and no government has chosen to test whether a board could be established over a State’s objections. The obstacle, on the statute’s own text, is not legal capacity. A board requires consultation, not concurrence. The obstacle is that the Union has never been willing to spend the political capital the consultation was designed to price.

The series to which this article belongs has catalogued statutes never notified, statutes never given their machinery, and statutes never operated. The River Boards Act is the purest instance of the last category in the entire catalogue: not a provision awaiting commencement, not a body starved of funds, but an Act in force under which the constitutive act has simply never been performed. It matters to the disputes Act for a practical reason as well as an ironic one. Every function the boards were meant to discharge, basin data, coordinated planning, early warning of conflict, falls by default to the tribunal process or to nobody. The tribunals therefore inherit disputes that arrive fully politicised, with no agreed hydrological record beneath them, which is one reason the 2002 amendment had to build a statutory data bank into the disputes Act itself, a subject the amendment section takes up.

A board does exist for one inter-State basin, and its pedigree proves the same point from the opposite side. The Upper Yamuna River Board, which oversees the sharing of the upper Yamuna among Uttarakhand, Himachal Pradesh, Uttar Pradesh, Haryana, Rajasthan and Delhi, was constituted in 1995 under a memorandum of understanding on the allocation of the river’s surface flows signed by the basin States and the Centre in May 1994. It is a functioning inter-State river institution with a governing body of the participating governments. It is not, and does not purport to be, a River Board under the River Boards Act, 1956. When the Union and the States have wanted a river institution, they have built it by agreement or, as the next passage shows, by a bespoke provision in another statute, and have walked past the one Act that exists for no other purpose. The explanation offered for the Act’s disuse, that its boards would be advisory and toothless, cannot account for the Upper Yamuna board, which is also in substance an allocative and supervisory arrangement resting on consent. The difference is who convenes: institutions negotiated among equals get built; the institution a central statute offered to convene has never been convened.

A contrast inside the wider statute book shows that the failure is specific to the River Boards Act rather than to river boards as an idea. When Parliament reorganised Andhra Pradesh in 2014, it wrote river management boards directly into the reorganisation statute: Section 84 of the Andhra Pradesh Reorganisation Act, 2014 required an Apex Council for the supervision of two boards, Section 85 dealt with the Krishna River Management Board and Section 86 with the Godavari River Management Board, and those boards were in fact constituted and function as part of the successor States’ water administration, as the pillar on the Andhra Pradesh Reorganisation Act records in its own treatment of that Act’s water chapters. River boards exist in India. They exist where a reorganisation Act imposed them as part of a settlement. What has never existed is a board under the one statute Parliament passed for the single purpose of creating them.

Whom the Act binds, and what lies outside it

The Act extends to the whole of India under Section 1(2), and its temporal reach is as important as its territorial reach. It governs disputes whenever they arise, including disputes whose subject matter is utilisation established long before 1956: the age of a use is not a threshold question under the Act but a factor in the equity of the allocation, which is why the Cauvery proceeding could put the agreements of 1892 and 1924, and the works built under them, at the centre of a reference made in 1990. A State cannot defeat the Act by pointing to the antiquity of its works, and a State cannot win under the Act by antiquity alone; the statute processes history as evidence rather than as title. That posture, adjudicating the future distribution of a resource whose past distributions were made by defunct sovereigns and expired agreements, is the deep reason the tribunal model, with its investigatory charter and its freedom from the forms of a suit, was chosen over the courtroom in the first place.

The Act binds governments, and only governments. Its disputants are State Governments, its gatekeeper is the Central Government, its adjudicator is a tribunal of sitting judges, and its sanctions are the obligations of States to give effect to a published decision. A Union territory has stood as a party where a basin touched it, as Puducherry did before the Cauvery tribunal, but the statutory text speaks throughout of State Governments, and no private person can invoke the Act, claim a right to intervene in it, or be directed by a tribunal except through the State Governments whose conduct the tribunal judges. The citizen whose borewell runs dry in a basin under adjudication is a constitutional spectator to this statute: affected by every line of the award, entitled to none of its procedures. That exclusivity is not an oversight. The Act is an instrument of federal diplomacy conducted in judicial form, and its parties are the federation’s units because the resource it divides is held, under Entry 17, by the units themselves.

The Act also binds across time in a way its text leaves the reader to infer: it binds successor States. Reorganisation redraws the parties without dissolving the proceeding or the award, because the reference runs with the river and the successor government steps into the predecessor’s position in the basin. The working examples are built into the census. The second Krishna tribunal, constituted for Maharashtra, Karnataka and Andhra Pradesh, found its Andhra party replaced mid-life by two successor States after 2014, and the reorganisation statute itself extended the tribunal’s term to make the project-wise allocations the new map required. The Mahanadi tribunal of 2018 sits between Odisha and a Chhattisgarh that did not exist when the upper catchment works in question were sanctioned by Madhya Pradesh. A State that inherits a basin inherits its litigation posture, its established utilisation and its grievances together, and the Act treats the succession as a change of parties rather than a change of dispute. Practitioners should note the consequence for agreements as well: an inter-State water agreement signed by a predecessor State is among the instruments whose interpretation Section 2(c) makes a water dispute, and succession disputes are fought, in the first instance, inside the definition.

Three large neighbouring subjects lie outside the Act, and the pillar must mark them so that readers do not import the wrong statute into a water problem. The first is groundwater: the Act speaks of the waters of, or in, any inter-State river or river valley, and its machinery is built for surface flows between States; the extraction of groundwater within a State, even where aquifers cross boundaries, is governed by State law and by the regulatory structures built under environmental legislation, not by tribunal adjudication. The second is water quality. A State that receives its neighbour’s effluent rather than its neighbour’s shortfall has a pollution grievance, not a sharing grievance, and the governing statute is the Water (Prevention and Control of Pollution) Act, 1974, with its boards, consent regime and offences, treated in the pillar on the Water Act 1974 and water quality regulation at the point where this article’s subject ends. Sharing law divides a quantity; pollution law polices a quality; the conflation of the two in public argument is one of the errors this article exists to correct. The third excluded subject is the intrastate river: a watercourse wholly inside one State belongs entirely to that State’s Entry 17 competence, and disputes about it are fought in that State’s politics and courts without any route to a Section 4 tribunal.

The Central Government’s own position under the Act is worth a sentence of precision, because the Centre is everywhere in the machinery and nowhere in the definition. A water dispute is between State Governments; the Centre is not a disputant the Act recognises, even where its own projects, its Union territories or its inter-State works are touched by the quarrel. It appears instead in every other capacity the statute offers: as the recipient of the complaint, the former of the gatekeeping opinion, the constituting authority, the grantor of extensions, the publisher of the decision, the framer of the scheme and the maker of the rules. Where a Union territory’s interests are engaged, as Puducherry’s were in the Cauvery basin, the territory participates in the proceeding while the Union that administers it simultaneously performs the Act’s executive functions, a doubling of roles the design accepts without remark. The Yamuna’s sharing, in which the National Capital Territory of Delhi is a principal claimant, has proceeded by agreement among the basin units rather than by reference, in part because the dramatis personae of that basin fit the Act’s State-to-State template awkwardly. The template’s awkwardness with territories is a small point with a long future, as capital cities’ water security becomes the sharpest edge of basin politics.

For a reader holding a live grievance rather than a syllabus, one limit should be stated once and plainly. This article describes the general law of inter-State river water adjudication as enacted and as read by the courts; it is legal information, and it is no substitute for advice on an individual matter, where the forum, the stage of any pending reference and the Gazette record of the particular basin will decide what can still be done and by whom. Governments act in this field on advice taken at the level of the State, and the statute’s clocks, once missed, are rarely recoverable.

The amendment arc: 1968, 1980 and the discipline of 2002

The Act has been amended three times in substance, and each amendment answers a failure the working of the original text had exposed. The amending Act 35 of 1968 repaired the tribunal’s internal fragility: it substituted the composition provision in Section 4(2) in the form the Act carries, inserted Section 5(4) so that a divided tribunal decides by majority, and inserted Section 5A so that a vacancy does not destroy the proceeding. The amending Act 45 of 1980 answered the Narmada experience by inserting Section 6A, the scheme power discussed in the finality section: the original Act had assumed that a final and binding decision would implement itself, and the 1980 amendment is Parliament’s formal admission that it would not, and that an award might need an authority, a fund, a review committee and a body corporate life of its own before a single structure was operated differently.

The 2002 amendment, the Inter-State Water Disputes (Amendment) Act, 2002, Act 14 of 2002, is the substantial one, and it was built on the recommendations of the Sarkaria Commission on Centre-State Relations, which had examined the Act’s working and urged that the open-endedness of its procedures be closed by statutory periods. The amendment renamed the Act itself, substituting “Inter-State River Water Disputes Act” for the original short title by inserting the word “River”, so that the statute’s name matches the subject its definition had always carried. It wrote the one year constitution period into Section 4(1), with the proviso that disputes settled before its commencement shall not be reopened. It substituted Sections 5(2) and 5(3) wholesale, producing the three year reporting period, the two year capped extension for unavoidable reasons, the three month window for a reference back, the one year further report period with its open-ended extension, and the rule that the decision stands modified by the further report. It inserted Section 6(2), giving the published decision the force of an order or decree of the Supreme Court. It armed the tribunal with the power to requisition data, by inserting clause (ba) in Section 9(1). And it inserted Section 9A, obliging the Central Government to maintain a data bank and information system at the national level for each river basin, covering water resources, land, agriculture and related matters, with State Governments bound to supply data as required and the Centre empowered to verify what is supplied and to summon records for that purpose.

Section 9A deserves a paragraph of its own, because it is the amendment’s most diagnostic provision. Water disputes are, beneath the law, disputes about numbers: dependable flow, utilisation, return flow, evaporation, the yield of a catchment in a fifty percent or seventy-five percent dependable year. Each State historically arrived before a tribunal with its own hydrology, compiled by its own engineers from its own gauges, and a tribunal’s first years were spent constructing a record the parties could not impeach. A standing national data bank, fed compulsorily and verified centrally, was meant to move that construction before the dispute, so that a tribunal could begin from shared measurements. The provision is the River Boards idea returning in miniature: since no board was ever created to hold the basin’s knowledge, the knowledge function was attached to the adjudication statute instead. Whether the data bank has in fact been maintained at the standard Section 9A describes is a question of administrative record; the statutory obligation itself is unqualified, and its presence in the Act is the clearest sign that Parliament understood, by 2002, that the tribunals’ difficulties were evidentiary before they were legal.

What the 2002 amendment deliberately did not do is as instructive as what it did. It did not touch the gatekeeping opinion in Section 4(1): the Centre still decides, without a clock, whether a dispute cannot be settled by negotiations. It did not create the single permanent tribunal that reform proposals had canvassed, leaving the one dispute, one tribunal structure intact. It did not prescribe a period for the publication of the decision under Section 6, the step on which finality depends. It did not cap the extension of the further report period under Section 5(3). And it did not address the Article 136 route, which the Cauvery proceedings would later confirm as surviving the ouster. The amendment is best read as a discipline imposed on the middle of the process, the tribunal’s own working life, while the beginning of the process and the end of it remained where the original Act had left them: with the Central Government’s discretion at the front, and with the Central Government’s Gazette at the back. The pre-2002 position deserves one explicit statement, because accounts of the Act frequently attribute its clocks to the original Parliament. The Act of 1956, as enacted, prescribed no period for the constitution of a tribunal, no period for its report, and no window for a reference back; Section 5 required a report setting out facts and decision and left the timetable to the proceeding itself. Every period a reader finds in the Act arrived in 2002, forty-six years into the statute’s life, which is why the tribunals of the first and second generations cannot be judged against clocks that did not exist when they sat, and why the Sarkaria Commission’s critique, delivered in the 1980s, was aimed at a text genuinely innocent of deadlines.

Where does the Union’s gatekeeping power sit, and why does it matter more than any period in the Act?

The gatekeeping power sits in Section 4(1), in the requirement that the Central Government be of opinion that the water dispute cannot be settled by negotiations before its duty to constitute a tribunal arises. It matters more than every period in the Act because the periods run only after the opinion exists.

The one year clock, the three year clock and the three month window all begin downstream of a decision for which the statute sets no clock, no criteria and no record.

The design defended, and the design indicted

The strongest case for the design is the one its structure makes for itself. Disputes between States over a shared river engage interests that no decree can finally compose, because the subject matter is a variable flow and the parties are permanent neighbours. A tribunal of judges, assisted by assessors, investigating on a record it builds for itself, can weigh dependability, established utilisation, the economics of cropping patterns and the claims of drinking water in a single proceeding of a kind no court entertaining a suit between States could conveniently conduct. The removal of the dispute from ordinary appellate channels protects that proceeding from being dismantled interlocutorily. On this reading, associated with the Sarkaria Commission’s preference for repairing the Act rather than replacing its philosophy, the tribunal is not a failed court but a different and defensible instrument: a standing invitation to treat allocation as a problem of basin management that happens to be decided by judges. The Godavari experience supports the reading. There, the tribunal’s existence concentrated the States’ minds, agreements were reached among them, and the tribunal’s report gave those agreements legal form; the machinery worked as a scaffold for settlement, which is arguably what it was for.

The strongest case against the design is written into the sections themselves, and the Punchhi Commission, reporting in 2010, joined a long line of official reviewers in pressing for the mechanism to be made time-bound in fact and not only in text. The indictment has three counts, one for each feature the brief of this article identified at the outset. The gatekeeping opinion has no clock, so the system’s first delay is invisible to the statute. The reporting clocks run against the tribunal, but their extensions are granted by the Central Government, and the publication on which everything depends has no clock at all, so the system’s discipline binds the adjudicator more tightly than the administrator. And the finality, fortified in 2002 with the force of a Supreme Court decree, was held in the Cauvery proceedings not to exclude the Supreme Court’s own special leave jurisdiction, so the system’s promise of an end to litigation is one the Constitution, as read by the Court, does not permit the statute to keep. A design can survive any one of these. The Act’s history is what the three produce together: proceedings that begin late by discretion, run long by extension, and conclude without concluding.

Between these readings, this article’s verdict is that the design’s theory is sound and its allocation of discretion is not. The tribunal idea answers the real character of the subject matter; the decision to make the tribunal’s existence, timetable extensions and final publication all depend on the judgment of one disputant-adjacent actor, the Union executive, which is itself a political party to federal bargaining, is the flaw from which the others grow. The Sarkaria Commission saw the flaw and prescribed clocks; the 2002 amendment applied the clocks to the wrong segment of the sequence. The Punchhi Commission’s later insistence on time-bound adjudication recognises the same point from the far end of another two decades of experience. Reform that does not move the gatekeeping decision, or clock it, reforms the part of the machine that was never the slowest.

One structural choice remains to be assessed on its merits, because it divides reformers who agree on everything else: the ad hoc tribunal itself. The case for constituting a fresh tribunal for each dispute is the case the 1956 Parliament acted on. A bench assembled for one river can be chosen for the proceeding’s expected character, sits without the institutional interests a permanent body accumulates, dissolves when its work is done, and returns its judges to the courts from which the Chief Justice of India borrowed them; the model keeps water adjudication an event in the life of the higher judiciary rather than a career within a specialised bureaucracy. The case against is the case the 2019 Bill accepted: a fresh tribunal must build its record, its procedure and its technical confidence from nothing in every dispute, its members serve at the intersection of judicial tenure and a proceeding measured in years, and the system’s accumulated learning about hydrology, dependability and drafting an operable award disperses on dissolution instead of compounding. The Sarkaria Commission, characteristically, sought the middle: it did not disturb the ad hoc model but surrounded it with the clocks and the data machinery whose absence made ad hocery expensive. This article’s assessment follows that instinct. The ad hoc tribunal is not the design’s failure; the first Krishna tribunal and the Godavari proceeding show it succeeding at the highest level. The failure is the absence of the standing supports, a permanent data bank in fact, a clocked gate, a publishing duty, that would let an ad hoc bench begin where its predecessors ended. Permanence was needed in the record and the registry, not necessarily in the bench.

The political economy of a reference

The sections distribute legal power; a separate question is who benefits from the distribution as it stands, because that answer predicts which reforms will be resisted and by whom. A reference under the Act rearranges the parties’ incentives asymmetrically, and the asymmetry is structural rather than cynical. The State in possession of the water, typically the upstream State whose works already command the flows, benefits from every year in which nothing is decided, because the status quo is its allocation. The State out of possession benefits from speed. A system whose first gate is an unclocked executive opinion, whose extensions are granted by the executive, and whose concluding act is an executive publication therefore places the tempo of justice in the hands of a third party, the Union government, whose own incentives run toward postponement: a Union government holds office with support drawn from both disputant States in most configurations, and a decided dispute creates one certain loser where an undecided one creates only a continuing grievance. The gatekeeping design does not merely permit delay; it rewards the gatekeeper for it.

The gatekeeping stage’s invisibility can be measured in the most recent constitution. Odisha’s complaint against Chhattisgarh over the Mahanadi was filed under Section 3 in 2016, grounded on the upper catchment works Chhattisgarh was building or sanctioning without, Odisha said, the downstream State’s interests being weighed. What followed was not a tribunal but a season of the negotiation stage: the Centre promoted talks and a committee process, Odisha pressed for the reference its complaint had requested, and the tribunal was constituted only in March 2018. Nothing in that sequence breached the Act, because the Act prescribes nothing for it; the one year period in Section 4(1) speaks to the interval the duty covers once the opinion exists, and the parties to the Mahanadi proceeding spent the better part of two years on the near side of the opinion. A reader who wants to know why a statute with a one year constitution period produces tribunals years after complaints should be directed to this gap between the request and the opinion, which is where the Mahanadi years went and where, in the design as enacted, they lawfully could.

State governments, for their part, face electoral incentives that reward intransigence in public and flexibility only in camera. Water in a riparian State is not an ordinary policy subject: it is tied to the land records, cropping patterns and drinking water security of identifiable districts, and a Chief Minister seen to concede a basin’s water pays a price that no compensating package inside the dispute can repay. The tribunal route has one underappreciated virtue within this politics: it lets a State government accept an outcome it could never have signed, because the outcome arrives as law rather than as concession. The Godavari agreements and the Cauvery judgment both illustrate the absolving function of adjudication, in which the decision’s binding force is precisely what makes compliance politically survivable. Any reform that weakens the tribunal’s authority in the name of negotiation must reckon with the possibility that negotiation, between governments electorally punished for conceding, is the forum least able to produce the settlement it romanticises.

The Union’s position has a further layer, visible in the Cauvery publication history. Publication is a ministerial act with political consequences: to notify an award is to choose the moment at which its costs fall due in the disputant States, and a Union government approaching elections in one of them has reasons of statecraft, as it will see them, to let the Gazette wait. The 2002 amendment’s failure to clock publication is therefore not an oversight that a drafting committee missed; it is the preservation, inside a reforming statute, of the one discretion the reforming government most valued. Students of the Act should read its silences with the same attention as its periods. Where the Act wishes to bind, it binds with numbers. Where it is silent, the silence is the provision.

The recurring proposal: move water itself

No assessment of the Act is complete without the proposal that would abolish its premises. Water, the proposal runs, should be moved from the State List to the Concurrent List, or in the stronger version to the Union List, so that the Union may legislate for inter-State basins directly and the adjudication of remnants may follow administration rather than substitute for it. The proposal has been pressed by Union governments, by experts in water administration, and by commentators who point to Entry 56 as a power Parliament has declined for seventy years to exercise in its general form. Its strongest version is institutional rather than legal: the defects catalogued in this article, the gatekeeping, the absent boards, the contested data, are symptoms of a constitutional settlement that assigns a basin resource to riparian units and then expresses surprise when the units behave as owners. Moving the entry would not merely change a list. It would change the default politics of every project clearance, every allocation and every tribunal reference, by making the Union the legislature of first resort for the rivers that matter most.

The strongest answer is the federalism answer, and it is made by the States and by the commissions that have examined the question and declined to recommend the transfer. The Sarkaria Commission considered the distribution of water competence and left Entry 17 in place, preferring to repair the disputes Act and to strengthen consultation; the Punchhi Commission likewise worked within the existing distribution. The answer has two limbs. The first is that water is not only an inter-State subject: most water use is intrastate, in wells, tanks, canals and municipalities administered by the States, and a transfer of the entry would centralise the whole subject to cure a defect in its inter-State margin. The second is that legislation is not the scarce resource in this field. The Union already holds, through Article 262 and Entry 56, every power needed to adjudicate disputes and to constitute basin bodies; what it has not supplied is the will to use the River Boards Act and the clocks to govern its own gatekeeping. A transferred entry would hand the Union a larger power to leave unexercised. This article’s verdict follows the commissions: the transfer debate, though permanent, is a displacement of the reform question. The instruments exist. The record of their non-use is the finding, and it survives any redistribution of the entries because it is a record about the Union’s choices, not about the Union’s competence.

The comparative mirror: compacts, courts and basin authorities

India’s design is one of three the federal world has produced for the same problem, and the comparisons locate its peculiarities precisely. The United States divides inter-State rivers by two instruments India lacks and one India declined. The first is the interstate compact: States negotiate an apportionment, Congress consents, and the compact becomes federal law administrable by a compact commission, the Delaware and Colorado basin compacts being the standard instances; the Indian equivalent would be an agreement under Section 2(c)’s shadow, and Indian practice shows, from the Cauvery agreements of 1892 and 1924 to the Godavari agreements of the 1970s, that States can conclude such settlements but cannot be brought to the table by any institution the statutes create. The second American instrument is the Supreme Court’s original jurisdiction, exercised as equitable apportionment litigation of the kind begun in Kansas v Colorado: the very forum Article 262 was written to displace for water. The third is the federal basin corporation on the Tennessee Valley model, which the Damodar Valley Corporation imitated in 1948 and which the River Boards Act declined to reproduce in general form. Australia’s Murray-Darling settlement supplies the closest modern parallel to the road India did not take: after a century of State rivalry under a constitutional trade-and-commerce constraint, the basin States referred powers to the Commonwealth, whose Water Act of 2007 created a basin authority and a statutory basin plan with enforceable diversion limits. The Australian sequence required the States’ consent to a transfer, obtained through drought politics; the Indian River Boards Act required only consultation, and the transfer it contemplated was the lesser one of advice and planning. That the smaller step has never been taken, while larger steps were taken elsewhere by harder routes, is the comparison’s sting: the obstacle in India has not been constitutional capacity but the political price of the first move, whoever is asked to make it.

The 2019 Bill and the unfinished reform

The Inter-State River Water Disputes (Amendment) Bill, 2019 is the standing evidence that the diagnosis in the previous section commands wide assent and no enactment. The Bill, as introduced and as passed by the Lok Sabha, proposed to replace the one dispute, one tribunal structure with a single standing tribunal with multiple benches, to which all existing tribunals’ pending disputes would stand transferred. It proposed a Disputes Resolution Committee, constituted by the Central Government with experts from relevant fields, to attempt a negotiated settlement before a dispute reached the tribunal at all, formalising the negotiation stage that Section 4(1) leaves to the Centre’s unrecorded opinion. And it proposed to compress the adjudication into a fixed outer period, two years for the tribunal’s decision with a limited extension, inside a total span stated in the Bill’s comparative materials as a maximum of four and a half years for the adjudication of a dispute. The Bill also carried the data provisions further, building on the Section 9A data bank, and it proposed one change that speaks directly to this article’s thesis: the decision of the tribunal’s bench was to be final and binding on the States concerned without awaiting the Central Government’s publication, abolishing the Gazette step that Section 6 makes the hinge of finality and that the Cauvery history showed could hold an award in suspense for years. A reform Parliament was willing to pass in one House thus conceded, in terms, that the publication gate was a defect worth removing. It was not enacted: it completed its passage in one House and went no further, and the law of inter-State river water disputes remains the Act of 1956 as amended in 2002. This article therefore treats the Bill throughout as what it is, a Bill: evidence of the direction of official reform thinking, and no part of the governing law. The reader who meets accounts describing a single permanent tribunal as the law of India is meeting the Bill misreported as an Act.

The longer reform menu, of which the Bill took only part, shows the same pattern. Proposals to transfer water from the State List to the Concurrent List recur in every generation of federalism debate; they would change who legislates, but they would not by themselves change who adjudicates, and Article 262 would still stand behind whichever legislature held the field. Proposals for permanent basin authorities return, in modern dress, to the River Boards Act that has waited since 1956 for its first notification. Proposals to make awards self-executing without a Gazette step attack the publication gap directly. Each proposal identifies a real defect. The constant in the record is that the defects are identified more readily than they are repaired, because every repair redistributes advantage among States that must consent to the redistribution in the very Parliament where their delegations sit.

Verdict

The Inter-State River Water Disputes Act 1956 endures because the problem it addresses has no better constitutional home, and it disappoints because the discretion it concentrates has never been disciplined at the points where delay and defiance actually enter. Its enduring contributions are three, and they should be weighed before its disappointments are counted, because a statute that has decided the sharing of the Krishna, the Godavari, the Narmada and the Cauvery has not failed in the way an unused statute fails. It located inter-State water adjudication in a constitutionally licensed forum outside the ordinary courts, and that location has held. It produced, across nine tribunals, a body of allocation practice, on dependability, established use and equitable apportionment, that no subsequent statute has replaced. And in the River Boards Act’s empty twinship it preserves, in statutory form, the road not taken: the basin planned before it is litigated. Its standing lesson for the reader of Indian legislation is the one this series returns to across statutes. A law is not only its text. It is also the identity of the actor each step is entrusted to, the clock each step is given, and the consequence each missed clock carries; and by that measure the 1956 Act is strongest exactly where its subject is the tribunal, and weakest exactly where its subject is the governments that summon the tribunal into being and publish what it decides. That distribution of strength, and no want of drafting elegance, is the Act’s whole biography in a single sentence. The working record of the Act is unusually documentary: tribunal reports that run to volumes, Gazette notifications that mark the legal moments, schemes that create the operating authorities, and a parallel file of negotiation that rarely reaches print. A researcher assembling a basin file from those materials, or a journalist covering a live dispute through a dry season, is handling records whose negotiation notes and source material deserve the protection that a private, encrypted workspace for sensitive research files exists to provide, because in water litigation the unpublished paper is frequently the paper that matters.

Frequently Asked Questions

Q: Who decides an inter state river water dispute?

A Water Disputes Tribunal constituted under Section 4 of the Inter-State River Water Disputes Act, 1956 decides it. The tribunal is not a standing body: the Central Government constitutes a separate tribunal for each dispute, staffed by a Chairman and two members who are sitting judges of the Supreme Court or a High Court, nominated by the Chief Justice of India. The tribunal investigates the matters referred to it, reports the facts as found together with its decision, and the decision becomes final and binding when the Central Government publishes it in the Official Gazette. Courts are barred from the dispute itself by Section 11 of the Act, subject to the Supreme Court’s special leave jurisdiction under Article 136 against the tribunal’s determination.

Q: What does Article 262 provide?

Article 262 of the Constitution does two things. Its first clause empowers Parliament to provide by law for the adjudication of any dispute or complaint with respect to the use, distribution or control of the waters of, or in, any inter-State river or river valley. Its second clause empowers Parliament to provide by law that neither the Supreme Court nor any other court shall exercise jurisdiction in respect of any such dispute or complaint, and it does so notwithstanding anything else in the Constitution. The Inter-State River Water Disputes Act, 1956 is the legislation enacted under both clauses: the tribunal machinery rests on the first, and the bar on court jurisdiction in Section 11 rests on the second.

Q: Who sets up a water disputes tribunal and on whose request?

The Central Government sets it up, and it acts on the request of a State Government. Under Section 3, a State Government that considers its interests, or its inhabitants’ interests, in the waters of an inter-State river or river valley prejudicially affected by another State may request the Central Government to refer the dispute to a tribunal. Under Section 4(1), the Central Government must then form the opinion that the dispute cannot be settled by negotiations; once that opinion exists, it must constitute the tribunal by notification in the Official Gazette within a period not exceeding one year from the date of receipt of the request. No State can constitute a tribunal itself, and no tribunal can take up a dispute on its own motion.

Q: What legal force does a water tribunal award carry once published?

Once the Central Government publishes the tribunal’s decision in the Official Gazette under Section 6(1), the decision is final and binding on the parties to the dispute, and the parties are bound to give effect to it. Section 6(2), inserted by the 2002 amendment, adds that the published decision has the same force as an order or decree of the Supreme Court. Before publication the tribunal’s instrument is a report to the Central Government; publication is the act that converts it into a binding decision. The decision may be carried into a scheme under Section 6A for its implementation, and it remains subject to the Supreme Court’s special leave jurisdiction under Article 136, as the Cauvery proceedings established.

Q: Can the Supreme Court hear a water dispute?

Not as a water dispute. Section 11 of the Act bars the Supreme Court and every other court from exercising jurisdiction in respect of any water dispute which may be referred to a tribunal, and Article 262(2) authorises that bar. Even so, the Court has heard water matters at the edges of the bar. In the Cauvery proceedings it held that appeals by special leave under Article 136 against a tribunal’s published decision are maintainable, because once adjudicated the matter is a determination of a tribunal rather than a subsisting water dispute, and it modified the award in 2018. It has also entertained suits and writ proceedings whose subject fell outside the statutory definition of a water dispute.

Q: What is the River Boards Act and has it ever been used?

The River Boards Act, 1956, Act 49 of 1956, is the companion statute Parliament enacted in the same year under Entry 56 of the Union List. It empowers the Central Government, after consultation with the Governments interested, to establish a River Board for an inter-State river or river valley to advise on the river’s regulation and development, prepare development schemes and coordinate uses across State lines. A board would be a body corporate staffed by members with expertise in irrigation, engineering, flood control, navigation, water conservation, administration or finance. It has never been used: no River Board has ever been established under the Act, and the National Commission to Review the Working of the Constitution described the Act as having remained a dead letter.

Q: What time limits apply to a water disputes tribunal?

The periods were inserted by the 2002 amendment and bind different actors differently. The Central Government must constitute the tribunal within one year from the date of receipt of a State’s request, once it has formed the opinion that negotiations cannot settle the dispute; no period governs the forming of that opinion. The tribunal must forward its report within three years, extendable by the Central Government by up to two further years where the decision cannot be given for unavoidable reasons. A government seeking explanation or guidance must refer the matter back within three months of the decision, and the tribunal’s further report is due within one year, extendable by the Central Government without a statutory ceiling. No period is prescribed for publishing the decision.

Q: Who enforces a tribunal award?

No single enforcement agency exists under the Act. The parties, meaning the State Governments, are bound by Section 6(1) to give effect to the published decision, which carries the force of a Supreme Court decree under Section 6(2). Day to day implementation runs through the machinery the award and any scheme under Section 6A create: an implementation authority where one is framed, as with the Narmada Control Authority and the Cauvery Water Management Authority, operating the release schedules and accounting for deliveries at inter-State gauges. Where a State resists, enforcement in practice proceeds through the Supreme Court’s supervision in proceedings before it, rather than through any execution machinery the Act itself provides.

Q: What counts as a water dispute under the 1956 Act?

Section 2(c) defines a water dispute as any dispute or difference between two or more State Governments with respect to the use, distribution or control of the waters of, or in, any inter-State river or river valley. The definition also covers disputes over the interpretation of the terms of an agreement relating to such waters, or over the implementation of such an agreement, and the levy of any water rate in contravention of the prohibition in Section 7 against seigniorage and similar charges. Two limits follow: the parties must be State Governments, so private parties cannot invoke the Act, and the river must be an inter-State river or river valley, so a watercourse lying wholly within one State is outside the definition.

Q: How is the Inter-State River Water Disputes Act different from the River Boards Act?

The two Acts of 1956 do opposite kinds of work. The disputes Act is adjudicatory and retrospective: it is set in motion by a complaint after a quarrel has arisen, and it produces a binding decision through a tribunal of judges. The River Boards Act is planning-oriented and prospective: it empowers the Central Government to establish a board, after consulting the Governments interested, to advise on a basin’s regulation and development and to prepare development schemes before quarrels harden. Their constitutional foundations also differ, the disputes Act resting on Article 262 and the boards Act on Entry 56 of the Union List. In operation they diverged completely: the disputes Act has produced nine tribunals, and no board has ever been established under the boards Act.

Q: Why could Parliament enact a water disputes law when water is a State List subject?

Because the Constitution distributes the subject in layers. Entry 17 of the State List gives States power over water supplies, irrigation, canals, drainage, water storage and water power, but expressly subject to Entry 56 of the Union List, which empowers Parliament over the regulation and development of inter-State rivers and river valleys to the extent Parliament declares Union control expedient in the public interest. The disputes Act, however, does not rest on either entry: Article 262, a provision outside the Seventh Schedule, independently empowers Parliament to provide for the adjudication of inter-State river water disputes and to bar the courts from them. Parliament could therefore build the tribunal machinery without disturbing the States’ primary competence over water itself.

Q: How many tribunals have been constituted under the Act, and for which rivers?

Nine tribunals have been constituted. The first generation, all in 1969, covered the Godavari, the Krishna and the Narmada. The Ravi and Beas Waters Tribunal followed in 1986 under the special provision for the Punjab Settlement, and the Cauvery Water Disputes Tribunal in 1990. A second Krishna tribunal was constituted in 2004 to revisit that basin’s sharing. The third generation comprises the Vansadhara tribunal of 2010, the Mahadayi tribunal of 2010 and the Mahanadi tribunal of 2018. Each tribunal is constituted for one dispute and is dissolved when the Central Government is satisfied no further reference is necessary, so the number sitting at any time is smaller than the number constituted across the Act’s life.

Q: Who sits on a water disputes tribunal, and who nominates its members?

A tribunal consists of a Chairman and two other members. All three must be persons who are, at the time of nomination, Judges of the Supreme Court or of a High Court, and all three are nominated by the Chief Justice of India. The Central Government may additionally appoint two or more assessors, in consultation with the tribunal, to advise it on technical matters; assessors assist and advise but do not decide. If a vacancy occurs in the Chairman’s or a member’s office, Section 5A requires it to be filled by a fresh nominee of the Chief Justice of India, and the investigation continues from the stage it had reached rather than beginning again.

Q: What did the 2002 amendment change in the 1956 Act?

The Inter-State Water Disputes (Amendment) Act, 2002, Act 14 of 2002, rebuilt the Act’s timetable and its finality clause. It required the tribunal to be constituted within one year of a State’s request, barred the reopening of disputes already settled, set the tribunal’s reporting period at three years with a capped extension of two, set a three month window for references back and one year for the further report, and inserted Section 6(2) giving the published decision the force of an order or decree of the Supreme Court. It also gave tribunals the power to requisition data, created the national data bank obligation in Section 9A, and inserted the word “River” into the Act’s short title.

Q: What is a scheme under Section 6A, and why does an award need one?

Section 6A, inserted in 1980, empowers the Central Government to frame a scheme by Gazette notification to make provision for all matters necessary to give effect to a tribunal’s decision. A scheme may establish an implementation authority, settle its composition, powers, functions, funds and meeting discipline, and create a review committee for specified decisions. An award needs such machinery because an allocation is a schedule of quantities, while water is delivered through reservoirs, regulators and gauges operated season by season: someone must run the schedule, measure the deliveries and answer for shortfalls. The Narmada Control Authority and the Cauvery Water Management Authority are the working instances of authorities created to give an award an operating life.

Q: Can a dispute that a tribunal has already settled be referred again?

Not as the same dispute. A proviso inserted into Section 4(1) by the 2002 amendment states that any dispute settled by a tribunal before that amendment’s commencement shall not be reopened, and the Act’s structure treats a published decision as final and binding on the parties. What remains possible is narrower. A government may use the reference back under Section 5(3) to seek explanation or guidance, which may modify the decision in the respects explained. A genuinely new dispute, arising from later utilisation, a later agreement’s interpretation or the review horizon an award itself sets, may found a fresh complaint, as the Krishna basin’s second tribunal shows. Finality attaches to the dispute decided, not to the river for all time.

Q: Why does Section 8 keep some matters away from a tribunal?

Section 8 provides that no reference shall be made to a tribunal of any dispute regarding any matter which may be referred to arbitration under the River Boards Act, 1956. Parliament designed the two Acts of 1956 as one architecture, in which disagreements arising from a river board’s advice or its implementation would go to the boards Act’s arbitration, and only allocation disputes proper would reach a tribunal. Section 8 is the junction between the two routes, preventing parallel proceedings over the same matter. Because no river board has ever been established and the arbitration route has never operated, the exclusion has had almost nothing to exclude, and it survives as evidence of the integrated design Parliament intended in 1956.

Q: What is the reference back for explanation or guidance under Section 5(3)?

It is the Act’s one internal continuation of a tribunal proceeding after the decision. If the Central Government or any State Government considers that anything in the tribunal’s decision requires explanation, or that guidance is needed on a point not originally referred to the tribunal, it may refer the matter back to the tribunal within three months from the date of the decision. The tribunal then forwards a further report, ordinarily within one year, giving the explanation or guidance it deems fit, and the decision is deemed to be modified accordingly. The device lets the tribunal clarify and complete its own decision without a fresh adjudication, and the Ravi and Beas proceeding has rested at this stage since its report of 1987.

Q: Does the Act cover groundwater, or a dispute involving a private party?

Neither. The Act’s definition reaches disputes between State Governments over the waters of, or in, an inter-State river or river valley, which is the language of surface flows in a shared basin. Groundwater extraction within a State is regulated through State law and the environmental framework, not through tribunal adjudication, though the Supreme Court in the Cauvery judgment treated a State’s available groundwater as a relevant consideration in weighing its needs. A private party, a farmer, a utility or a project company, cannot file a complaint, be a party or be directed under the Act; its route is to persuade its own State Government to complain, or to pursue ordinary remedies for its own rights in the courts.

Q: How does a water dispute under this Act differ from a suit between States under Article 131?

Article 131 gives the Supreme Court original jurisdiction over disputes between States involving legal rights, and it is the general federal forum. A water dispute, as the 1956 Act defines it, is carved out of that forum: Section 11 bars the Supreme Court and every other court from exercising jurisdiction in respect of a water dispute which may be referred to a tribunal, and the dispute goes instead to the tribunal the Central Government constitutes. Suits survive where the subject is not the allocation itself, as the Sutlej-Yamuna Link litigation showed, where the Supreme Court decreed a construction obligation arising from agreements rather than apportioning the river. The boundary runs between the undecided allocation, which belongs to the tribunal, and accrued legal obligations, which may still be sued upon.