Federalism is usually argued along a vertical line — how much power the centre holds and how much the states retain. But a federation is also a horizontal arrangement: a set of units that must live beside one another, share rivers that ignore their borders, honour each other’s courts, and let goods and people cross their lines freely.
India built an elaborate apparatus for this second axis and then, almost without exception, declined to switch it on. The disputes that arise between Indian states, the machinery designed to resolve and prevent them, and the reason an instrument that exists on paper and is never used tells us more about Indian federalism than any amount of constitutional text — that is the subject here.
Federal theory concentrates almost entirely on the vertical axis, and the Indian debate has followed it — Article 356, the Governor, the divisible pool, the Concurrent List.
Wheare’s test, which asks whether general and regional governments are each independent within their spheres, is a question about levels, not about neighbours.
Granville Austin, Morris-Jones and every commission from Rajamannar to Punchhi framed the Indian problem as central dominance; the state-versus-state axis appears as a residual category.
The horizontal axis has its own constitutional apparatus, its own statutes, and a failure pattern entirely its own. Three classes of dispute recur, and they are structurally different from one another.
Class of dispute
Subject matter
Designated forum
Actual forum
River water
apportionment of flows, storage, dam operation
tribunal under the ISRWD Act 1956
tribunal plus the Supreme Court under Article 136
Territory and boundary
villages, districts, capital cities, colonial notifications
Parliament under Article 3
Article 131 suits that sit for decades
Shared institutions and resources
assets and liabilities on reorganisation, capitals, employees, trade barriers, domicile
negotiation and reorganisation schemes
ad hoc central mediation and litigation
The design assumption that failed: the drafters expected a dominant centre to absorb inter-state friction before it hardened, and for four decades the party system did that work.
When one party governed at the centre and in almost every state, a dispute between two states was a dispute inside a single party organisation, settled by the high command rather than by an institution.
Cauvery, Ravi-Beas and Belagavi all became intractable in the same period — the late 1970s and 1980s — in which single-party dominance ended and regional parties took office in the disputing states.
The horizontal machinery was never tested while it was easy to use, and by the time it was needed the political conditions for using it had gone.
India built its horizontal institutions during the years it did not need them, and needed them during the years it stopped building.
Comity between the states: full faith and credit
Article 261 is the constitutional guarantee of comity — the obligation of each state to treat the legal acts of every other state as legally real.
Clause (1): full faith and credit shall be given throughout India to the public acts, records and judicial proceedings of the Union and of every state.
Clause (2): Parliament may prescribe the manner and conditions of proof of such acts and records.
Clause (3): final judgments of civil courts in any part of India are executable anywhere in India.
The provision is modelled on the American Full Faith and Credit Clause but does far less work, because India has a single integrated judiciary, uniform procedural codes and single citizenship — the recognition problems that dominate American inter-state law barely arise. Article 261 does not extend to penal laws.
The quiet consequence: the unifying provisions of the Constitution pre-empted the horizontal disputes a looser federation generates, leaving only those unification cannot dissolve — rivers, land and money.
Freedom of trade and commerce: the forgotten horizontal provision
Part XIII (Articles 301–307) guarantees that trade, commerce and intercourse throughout India shall be free. It is the most consistently overlooked piece of horizontal federalism in the Indian scheme.
Article 301 — the freedom itself, binding on the Union and the states alike.
Article 302 — Parliament may restrict it in the public interest; Article 303 forbids either Parliament or a state from preferring one state over another, save on a parliamentary declaration of scarcity.
Article 304(a) — a state may tax imported goods only if similar local goods are taxed alike: the non-discrimination rule. Article 304(b) — reasonable restrictions in the public interest, with the previous sanction of the President.
Article 307 — Parliament may appoint an authority to carry out the purposes of Articles 301 to 304. No such authority has ever been appointed. A constitutional office designed to police the internal common market has stood empty for the life of the Republic.
The judicial history runs in three stages and ends by discarding forty years of doctrine.
Decision
Bench
Holding
Atiabari Tea Co. v. State of Assam (1961)
five judges
taxes that directly and immediately restrict movement fall within Article 301
Automobile Transport (Rajasthan) (1962)
seven judges
added the compensatory tax exception — a levy paying for a facility is regulatory, not restrictive
Jindal Stainless Ltd. v. State of Haryana (11 November 2016)
nine judges
the compensatory tax doctrine is discarded as having no constitutional foundation; taxes fall outside Article 301 unless discriminatory; a levy violating 304(a) cannot be saved by 304(b)
Jindal Stainless matters well beyond entry tax. By reading Part XIII down to a non-discrimination guarantee, the nine-judge bench enlarged the states’ fiscal room and shrank the judicially enforceable content of the internal market. “Free” in Article 301 does not mean free of taxation; it means free of taxation that discriminates by origin.
The practical guardian of the common market is therefore no longer Part XIII but the Goods and Services Tax, which abolished entry taxes, octroi and check-post levies by absorbing them.
Residual horizontal trade barriers persist outside GST and are genuine inter-state disputes even when nobody calls them that: state excise and inter-state liquor movement; APMC regulation and restrictions on foodgrain and cattle movement; inter-state transport permits and motor vehicle taxation; and local-supplier preferences in public procurement, which sit uneasily with Article 303.
Mobility, domicile and the sons-of-the-soil question
The Constitution treats India as one labour market. Article 15(1) forbids discrimination on grounds of place of birth; Article 16(2) forbids it in public employment on grounds of residence; Article 19(1)(d) and (e) guarantee free movement and the right to reside anywhere. Article 16(3) creates the sole exception, and reserves it to Parliament alone.
In Dr. Pradeep Jain v. Union of India (1984) the Supreme Court held that there is no state domicile in India, only domicile of the country, and that wholesale reservation of professional college seats by residence offends the constitutional scheme, though limited institutional preference may survive.
The pressure now takes the form of private-sector local reservation laws. The Haryana State Employment of Local Candidates Act, 2020, reserving 75% of lower-paid private jobs for locals, was struck down by the Punjab and Haryana High Court in November 2023 as violating Articles 14 and 19. Comparable proposals elsewhere have stalled at the same obstacle.
Migration is where horizontal federalism touches ordinary life most directly, and where the absence of inter-state institutions shows. Ration portability was achieved by central scheme design, not inter-state agreement; portability of construction-worker cess benefits, school admission and voter registration across state lines still has no standing mechanism at all.
Water: the largest class of inter-state dispute
Water is where the federal design meets ecology, and ecology recognises no boundary. A river basin is a single hydrological unit that the map has cut into pieces, and every piece has an electorate. That combination — a physically indivisible resource divided among politically autonomous claimants — produces the most persistent, most violent and least tractable class of dispute in Indian federalism.
Why the resource makes the politics
Agriculture absorbs roughly four-fifths of India’s water withdrawals and supports close to half the workforce, so an adverse water order registers as an income shock in an identifiable constituency within a single cropping season.
The per-capita numbers explain the direction of travel.
Annual availability was assessed at about 1,486 cubic metres in 2021, projected to fall to about 1,367 by 2031, on the Central Water Commission’s reassessment using space inputs.
The thresholds are 1,700 cubic metres for stress and 1,000 for scarcity. India sits inside the stressed band and is moving toward the scarcity line.
The national average conceals the real problem: the Sabarmati, the Pennar and the east-flowing rivers between the Mahanadi and the Pennar are already below the scarcity threshold — and those are the basins that generate disputes.
Most of India’s major basins are inter-state. Ganga, Brahmaputra, Godavari, Krishna, Cauvery, Mahanadi, Narmada, the Indus system and the Mahadayi all cross state lines; a purely intra-state major basin is the exception.
The “water wars” prediction deserves scepticism, and the reasons are instructive.
The claim — captured in the line usually attributed to Mark Twain, that whisky is for drinking and water is for fighting over — is rhetorically powerful and empirically weak where states go to war. Comparative work on international rivers finds cooperation far more common than armed conflict, because water cannot be carried away and upstream-downstream relations are repeated games.
The sceptics’ finding does not travel down to the sub-national level. Inside a federation the parties cannot use force but can use legislatures, bandhs and non-compliance; the dispute does not escalate and it also does not resolve. It simply persists.
India’s water disputes do not become wars. They become permanent.
Entry 17 against Entry 56: the division the Union chose not to use
Entry 17, State List — water supplies, irrigation and canals, drainage and embankments, water storage and water power — subject to Entry 56 of the Union List.
Entry 56, Union List — regulation and development of inter-State rivers and river valleys to the extent to which such regulation under Union control is declared by Parliament by law to be expedient in the public interest.
Read together the scheme is unambiguous. Water within a state is a state subject; an inter-state river is available to the Union the moment Parliament declares it expedient to take it. The subordination is written into Entry 17 itself.
The Union has almost entirely declined to make that declaration. Parliament has never enacted a general law bringing India’s inter-state rivers under central regulation, using Entry 56 only for isolated, project-specific statutes. Inter-state rivers are consequently administered as though they were a state subject — precisely what the Constitution says they are not.
The reason is electoral rather than constitutional: water is radioactive. A Union government that apportions a river creates a losing state, and it must contest elections in the losing state.
Analysts of Indian water governance, Nirvikar Singh among them, place the primary responsibility for the persistence of these disputes on the Union’s abdication of Entry 56, and it is this abdication that politicised them. Where no neutral authority regulates, the only actors left are state governments, whose incentive is to maximise their own share and concede nothing.
Article 262: the one dispute the courts were shut out of
Article 262(1) — Parliament may by law provide for the adjudication of any dispute with respect to the use, distribution or control of the waters of any inter-state river or river valley.
Article 262(2) — notwithstanding anything in the Constitution, Parliament may provide that neither the Supreme Court nor any other court shall exercise jurisdiction in respect of such a dispute.
This is the only category of dispute from which the Supreme Court’s jurisdiction is expressly excludable. Nothing else in the Constitution lets Parliament close the doors of the highest court on an entire subject.
The exclusion is permissive, not mandatory — Article 262(2) says may. Parliament exercised it in Section 11 of the Inter-State River Water Disputes Act 1956.
The drafters’ reasoning was that apportioning a river is not justiciable in the ordinary sense: it turns on hydrology, cropping patterns, storage and projections, not on the interpretation of a right.
The consequence is that the dysfunction of this field is a constitutional problem, not merely an administrative one. When a mechanism that replaced the Supreme Court fails, there is no ordinary remedy behind it — which is exactly why the Court found a way back in.
The other routes: Articles 131, 136 and 263
Article 131 — original jurisdiction. The Supreme Court has original and exclusive jurisdiction in disputes between the Union and states, or between two or more states, where the dispute involves a question on which the existence or extent of a legal right depends.
Until 1956 this was the route by which inter-state water disputes were litigated.
The proviso excludes disputes arising out of pre-Constitution treaties and agreements still in operation — a limitation the Cauvery litigation had to negotiate, since the underlying instruments dated to 1892 and 1924.
A space survives wherever the dispute is not a “water dispute” as the 1956 Act defines it. Dam height and dam safety are the classic example, which is why Mullaperiyar has been fought in original jurisdiction and never before a tribunal.
Article 136 — special leave to appeal from any judgment or order of any court or tribunal. Because a water disputes tribunal is a tribunal, the Supreme Court re-entered by the back door a field from which Article 262 had shut it out of the front. Article 262 excluded the Court from the dispute; Article 136 let it back in through the award.
Article 263 — the Inter-State Council, whose first listed duty is inquiring into and advising upon disputes between states. A body whose primary constitutional purpose is inter-state dispute resolution has never been given a water dispute to consider.
The two statutes of 1956
Parliament acted under Article 262 in the same year as the great reorganisation of states, and it enacted two laws, not one. The pairing is the key to the whole failure: one statute was meant to prevent disputes and the other to decide them, and only the second was ever used.
The River Boards Act 1956: the preventive instrument
The Act empowers the central government, on a state’s request or on its own initiative, to establish a River Board for an inter-state river or river valley.
A Board’s function is to advise on the integrated and efficient development of the basin: irrigation, water supply, drainage, flood control, hydro-power, navigation and pollution control, planned across the basin rather than state by state.
The design intention was to make disputes unnecessary rather than to resolve them afterwards, by giving co-riparian states the habit of joint planning, development, execution and monitoring.
Not a single river board has ever been constituted under the Act, in seventy years. The statute is not repealed, not amended, not litigated — simply unused. The Union’s chosen instrument was prevention, and it never switched it on. Everything that follows — the tribunals, the delays, the unimplemented awards, the Court’s improvisations — happens downstream of that decision.
Every water dispute India has adjudicated for seventy years was a dispute a river board existed to prevent.
The Inter-State River Water Disputes Act 1956: the adjudicatory instrument
Section 2(c) defines a water dispute as a difference between states over the use, distribution or control of an inter-state river’s waters, the interpretation of an agreement relating to them, or the levy of a water rate — a definition that leaves dam safety and dam height outside.
Section 3 — an aggrieved state may request the Union in writing to refer the dispute. Section 4 — where the Union is of opinion that the dispute cannot be settled by negotiations, it shall constitute a Tribunal; negotiation is a statutory precondition, not an optional first step.
Section 5 — the tribunal forwards a report and decision; a state may seek clarification within three months, on which a further report issues.
Section 6 — the Union shall publish the decision in the official Gazette, and once published it is final and binding with the force of an order or decree of the Supreme Court. Section 6A — the Union may frame a scheme to give effect to it, including by establishing an implementing authority.
Section 11 — the bar of jurisdiction on the Supreme Court and every other court.
The 2002 amendment, on the Sarkaria Commission’s recommendation, plugged the most visible holes.
Stage
Before 2002
After 2002
Constituting a tribunal
no limit
one year
Delivering the award
no limit
three years, extendable by two
Giving effect to the award
executive discretion
Section 6A implementation scheme
Status of the award
binding on Gazette publication
unchanged
The Inter-State River Water Disputes (Amendment) Bill was the next attempted repair, and it matters both for what it proposed and for what became of it.
A single permanent tribunal — chairperson, vice-chairperson and up to six members — sitting in multiple benches, with all existing tribunals dissolved and their disputes transferred.
A Disputes Resolution Committee of a chairperson, subject experts and one representative of each disputing state, to attempt mediated settlement first within one year extendable by six months.
Tighter limits on adjudication: two years plus one, against three plus two.
A central agency maintaining a national water data bank, so tribunals stop beginning each dispute by arguing about how much water exists.
The award to bind from the date it is given, ending the requirement of Gazette notification — a direct answer to the Cauvery experience, where the Union’s refusal to notify neutralised an award for years.
The Bill passed the Lok Sabha on 31 July 2019, never cleared the Rajya Sabha, and lapsed under Article 107(5) on the dissolution of the 17th Lok Sabha in 2024. It has not been reintroduced. The reform most widely agreed upon in this field is, as of 2026, not law.
The tribunals in practice
The record
Nine tribunals have been constituted since 1956, and a tenth has been ordered into existence by the Supreme Court.
never gazetted; review petition pending since 1987
Cauvery
Tamil Nadu, Karnataka, Kerala, Puducherry
1990
interim 1991, final 2007
notified 2013; modified by the Supreme Court, 2018
Krishna II (Brijesh Kumar)
Maharashtra, Karnataka, Andhra Pradesh, Telangana
2004
2010, further award 2013
not notified; sitting on a 2023 reference
Vansadhara
Odisha, Andhra Pradesh
2010
draft 2017, report 2021
not notified
Mahadayi
Karnataka, Goa, Maharashtra
2010
2018
notified 27 February 2020; under challenge
Mahanadi
Odisha, Chhattisgarh
2018
pending
tenure extended to 13 January 2027
Pennaiyar
Tamil Nadu, Karnataka
ordered 2 February 2026
—
Union given six further months in May 2026
The arithmetic is worth stating plainly. In seventy years: nine tribunals; three awards carrying no Gazette notification at all; two disputes still unadjudicated; and one award rewritten by the court whose jurisdiction the statute excluded.
The first generation — Krishna I, Godavari, Narmada — worked deliberatively and produced consent-based settlements in significant part, which is why the Godavari award of 1979, in substance a set of agreements the tribunal recorded, has generated the least subsequent litigation of any award.
The second generation — Cauvery, Ravi-Beas, Krishna II, Mahadayi, Mahanadi — has behaved like litigation: adversarial pleadings, expert wars, appeals and non-compliance. The change tracks the party system. From the 1980s the disputing states were governed by different parties with distinct regional bases, and a chief minister’s incentive shifted from settling to being seen to fight.
Composition: judges deciding a hydrological question
Section 4 requires a chairman and two members nominated by the Chief Justice of India from serving judges of the Supreme Court or a High Court. The bench is therefore entirely judicial; technical assessors may advise but do not decide.
The Punchhi Commission recommended interdisciplinary tribunals including hydrologists, agronomists, economists and environmental scientists as full members. The recommendation has never been adopted.
The objection is not that judges are unqualified but that the question is not a legal one.
Deciding what a basin yields at 75% dependability is a hydrological estimate, not a finding of right.
Deciding what a state needs requires judgments about cropping patterns, irrigation efficiency, evaporation losses and groundwater substitution.
Deciding what is equitable means weighing command area, population, alternative sources and economic use value — applied economics.
A purely judicial bench must take all of this from partisan expert evidence filed by the states themselves, which is why so much tribunal time is consumed by disputes about data rather than about law. The absent central data bank is not a small omission.
Delay
Before 2002 there was no time limit of any kind, for constituting a tribunal or for delivering an award.
Cauvery is the extreme case: request in 1970, tribunal in 1990, interim award in 1991, final award in 2007 — thirty-seven years from request to award, twenty of them before adjudication even began.
Ravi-Beas has been “in progress” since 1986, its 1987 report untouched by notification and its 1987 review petition unheard for nearly four decades.
Krishna II, constituted in 2004, awarded in 2010, reopened in 2013, overtaken by the 2014 bifurcation of Andhra Pradesh and given a fresh reference in October 2023, is still sitting — twenty-two years and counting.
The post-2002 limits are honoured in the breach, because they are extended by executive notification rather than enforced. Mahadayi has taken seven extensions since its 2018 award; Mahanadi was extended by nine months in April 2026, at the joint request of both states — an extension the litigants themselves sought.
The Pennaiyar episode is the pattern in real time. The Supreme Court directed the Union on 2 February 2026 to notify a tribunal within one month; on 8 May 2026 the Union reported that inter-ministerial consultation was complete but cabinet approval was not, and was granted six further months.
Non-implementation, which is the deeper failure
Section 6 gives an award the force of a decree of the Supreme Court. Nothing in Indian law gives an administrative determination higher status. It does not help.
The failure operates at three points, and each has been used.
The Union does not notify. An unnotified award does not bind, so the losing state’s remedy is to persuade the Union to sit on it — the Ravi-Beas, Krishna II and Vansadhara awards remain unnotified.
The Union frames no Section 6A scheme, so nobody measures flows, allocates in a distress year or records default.
The losing state does not comply, and there is no ordinary process of execution against a state government that is politically unable to obey.
The calculus is the same every time: the political cost of implementing an adverse award inside the losing state exceeds the legal cost of not implementing it. Contempt proceedings against a state are slow, and the sanction that would bite — Article 356 — has been unavailable after S.R. Bommai for anything short of constitutional breakdown. The award has the force of a decree and none of the enforcement of one.
What principle should govern apportionment?
The doctrinal question is unsettled in Indian law and imported almost entirely from international water law, which is itself unsettled.
a state may do as it likes with water in its territory
upper riparian
asserted by US Attorney General Judson Harmon in 1895; abandoned even by the United States
Absolute territorial integrity
the lower riparian is entitled to the natural, undiminished flow
lower riparian
never accepted; would freeze all upstream development
Prior appropriation / historical rights
first and long use creates a protected entitlement
whoever developed first
rejected by the Supreme Court in 2018
Equitable apportionment
each riparian is entitled to a fair share on all relevant factors
neither, by design
the governing principle in tribunal practice
No significant harm
a riparian must not cause significant harm to co-riparians
lower riparian, weakly
complements equitable utilisation
The Helsinki Rules (1966), framed by the International Law Association, established equitable and reasonable utilisation and listed the factors — geography, hydrology, climate, past utilisation, economic and social needs, dependent population, comparative costs of alternatives, avoidance of waste.
The UN Watercourses Convention (1997) codified two obligations that pull against each other — equitable and reasonable utilisation and the obligation not to cause significant harm — with a duty of prior notification of planned measures. It entered into force in 2014; India is not a party. The Berlin Rules (2004) updated the framework to give ecological flows and the human right to water a firmer place.
Indian tribunals apply equitable apportionment in name, but each award invents its own weighting, and there is no statutory or judicial statement of what the factors are or how they rank. The 2018 Cauvery judgment added two things — it rejected historical rights and elevated drinking water to the highest priority — without supplying a general formula. The National Water Policy 2012 ranks drinking water first as well, but a policy document binds nobody.
Case studies
Cauvery: the dispute that produced every failure mode
The colonial inheritance. Agreements between the Madras Presidency and the princely State of Mysore in 1892 and, decisively, in 1924 governed the basin. The 1924 agreement was to run fifty years, expiring in 1974.
Tamil Nadu moved early and was ignored. The state approached the Union in 1970, anticipating the expiry. The Union took twenty years to constitute a tribunal, doing so in June 1990 only after Tamil Nadu obtained a direction from the Supreme Court.
The interim award. The tribunal held it had no power to grant interim relief; Tamil Nadu went to the Supreme Court, which held otherwise, and an interim award of June 1991 directed Karnataka to ensure 205 TMC at Mettur in a water year.
Karnataka legislated against the award, promulgating the Karnataka Cauvery Basin Irrigation Protection Ordinance, 1991 to empower itself to disregard it.
The Presidential Reference under Article 143 that followed produced the field’s single most important constitutional statement. In In Re: Cauvery Water Disputes Tribunal (22 November 1991) the Court held:
the Ordinance was beyond the state’s legislative competence and unconstitutional;
a state cannot legislate to nullify a tribunal’s award, because that is to sit in judgment over an adjudication binding on it;
no state may be a judge in its own cause in a dispute to which it is a party;
the interim award bound Karnataka, and the Union was under a duty to publish it.
The Gazette standoff. Karnataka’s position that nothing bound it until notification was answered only when, after further litigation, the Union published the award. What followed was the ugliest phase of any Indian inter-state dispute: a bandh in Karnataka, a fast by the Chief Minister of Tamil Nadu, rioting against Tamil residents in Bengaluru and against Kannadigas across the border, and farmer suicides in both deltas.
The final award of 5 February 2007, on an assessed availability of 740 TMC, allocated Tamil Nadu 419, Karnataka 270, Kerala 30 and Puducherry 7 TMC, with 10 TMC for environmental protection and 4 TMC for escapages to the sea, requiring Karnataka to deliver 192 TMC at Biligundlu in a normal year.
The award was not notified until 19 February 2013, and then only under Supreme Court direction. A monitoring committee and later an authority headed by the Prime Minister were constituted; neither produced compliance.
The Supreme Court’s judgment of 16 February 2018 was the first occasion on which the Court modified a tribunal award itself.
Holding
Substance
Reallocation
Karnataka’s share raised by 14.75 TMC to 284.75; Tamil Nadu’s reduced to 404.25 TMC; Kerala and Puducherry unchanged
Reason
drinking water for Bengaluru, and the extent of groundwater availability in Tamil Nadu’s delta
Release obligation
177.25 TMC at Biligundlu annually, on a monthly schedule
Groundwater
groundwater availability must be reckoned in assessing need — a first
Ownership
inter-state rivers are national assets; no state may claim exclusive ownership
Priority
drinking water ranks highest among competing uses
Doctrine
the doctrine of historical rights is rejected
Direction
award to hold fifteen years; the Union to frame a scheme within six weeks
The Union missed the six-week deadline, pleading the Karnataka assembly elections, and moved only after the Court’s displeasure. The Cauvery Water Management Authority and the Cauvery Water Regulation Committee were notified in June 2018 — a two-tier structure, the Authority for allocation and supervision, the Committee for ground-level monitoring of storage, inflow, cropping and rainfall.
The machinery has not ended the dispute; it has institutionalised the annual argument.
In the distress year of 2023 Tamil Nadu sought 24,000 cusecs a day; the Authority ordered 5,000 cusecs for fifteen days in September; Karnataka complied under protest and then stopped. A statewide Karnataka bandh on 29 September 2023 was met by rail blockades from Tamil Nadu farmers, with hundreds detained and prohibitory orders in six districts.
The Supreme Court declined to substitute its own figure, asking the Authority to report instead — a notable exercise of restraint after 2018.
The award contains no adequate distress-sharing formula, which is the single most consequential gap in it.
Mekedatu is the next phase. Karnataka’s proposed balancing reservoir — a 99-metre dam of 67.16 TMC, revised cost above ₹14,000 crore — has been resubmitted to the Central Water Commission in a drinking-water-only configuration, a framing designed to invoke the 2018 judgment’s own priority ranking. Tamil Nadu opposes it; clearance is pending over roughly 4,800 acres of forest including part of the Cauvery Wildlife Sanctuary.
The moral is not that adjudication failed. Adjudication succeeded three times — tribunal, Presidential Reference, Supreme Court — and delivery still depended on the Union’s willingness to act against a state it needed.
Ravi-Beas and the Sutlej-Yamuna Link Canal
The Indus Waters Treaty (1960) allocated the Ravi, Beas and Sutlej to India and the western rivers to Pakistan. For Punjab, whose canal command had been built around the whole system, the treaty converted a regional water economy into a national bargaining problem.
Haryana was carved out in 1966 by the Punjab Reorganisation Act and immediately claimed a share of the Ravi-Beas surplus. The two arguments have never met.
Punjab’s objection is riparian: the Sutlej, Ravi and Beas do not flow through Haryana, and a non-basin state has no claim.
Haryana’s answer is from state succession: it is a successor state, and Section 78 of the Punjab Reorganisation Act expressly empowered the Union to allocate the rights and liabilities of the Bhakra-Nangal and Beas projects between successors.
One argument is drawn from hydrology, the other from the law of succession, and no Indian forum has ever chosen between them.
The sequence of settlements, each undone by the next:
1976 — a central executive order under Section 78 allocated the surplus among Punjab, Haryana and Rajasthan.
1981 — a tripartite agreement revised the shares and provided for the Sutlej-Yamuna Link Canal to carry Haryana’s share.
Water became one of the grievances feeding the Punjab agitation; the canal became a symbol.
1985 — the Rajiv-Longowal Accord conceded Chandigarh to Punjab and referred the water claims to a tribunal. Longowal was assassinated within weeks and the accord was never implemented in full.
1986 — the Ravi and Beas Waters Tribunal under Justice V. Balakrishna Eradi; its report of 30 January 1987 reduced Punjab’s share relative to 1981.
The Eradi report has never been published in the Gazette, and Punjab’s review petition of 9 August 1987 has never been decided. Thirty-nine years on, the tribunal’s report is a document with no legal effect.
The 2004 rupture. The Punjab Termination of Agreements Act, 2004 purported to terminate every water-sharing agreement Punjab had entered into — an act without precedent in Indian federalism.
A further Presidential Reference under Article 143 produced the Supreme Court’s advisory opinion of 10 November 2016: the Act was unconstitutional, since a state cannot unilaterally terminate agreements binding on it, and doing so trespasses on the adjudicatory scheme of Article 262.
The aftermath. With elections due in 2017, Punjab de-notified the land acquired for the canal, began returning it to the original owners and returned Haryana’s advance payment. Haryana, whose own section was complete, returned to the Court, which ordered status quo in 2017.
The 2025 escalation moved the dispute from the canal to the dam. In May 2025 the Bhakra Beas Management Board decided to release an additional 4,500 cusecs to Haryana; Punjab refused to operate the gates. The Punjab and Haryana High Court held that Punjab’s police could not interfere with the Board’s operation of the Bhakra dam and directed Punjab to take its grievance to the Union. A basin-level board without operational control is only as strong as the host state’s cooperation.
The position in 2026 is, for the first time in decades, one of movement. The two Chief Ministers met directly on 27 January 2026; before the Supreme Court on 31 January 2026 the Attorney-General reported that the states now spoke in a joint voice and that Punjab was “coming much closer”; the matter was heard again on 20 August 2026, with the Centre mediating and Haryana’s original 1996 suit still the operative proceeding.
“Once the heart opens, the door also opens.” — Chief Justice Surya Kant, on the Sutlej-Yamuna Link Canal
The direction of travel is instructive. After a tribunal that could not report effectively, a statute that was struck down and an advisory opinion that changed nothing on the ground, what has produced movement is negotiation between two chief ministers under judicial encouragement.
Krishna: an award overtaken by a new state
Krishna Water Disputes Tribunal I under Justice R.S. Bachawat, constituted in 1969, allocated the assessed 2,060 TMC at 75% dependability as Maharashtra 560, Karnataka 700, Andhra Pradesh 800, becoming final in 1976 with a review provision after 2000.
Krishna Water Disputes Tribunal II under Justice Brijesh Kumar, constituted in 2004, reassessed the basin and allocated Maharashtra 666, Karnataka 907, Andhra Pradesh 1,005 TMC, with a further award in 2013.
The 2014 bifurcation destroyed the award’s premise. The Andhra Pradesh Reorganisation Act, 2014 created Telangana without dividing the undivided state’s Krishna entitlement between the successors; the Krishna and Godavari River Management Boards were created to administer projects, and Section 89 directed the tribunal to make project-wise allocations. The 2013 award has never been notified, a Supreme Court order having restrained publication.
In October 2023 the Union Cabinet approved a fresh Section 3 reference on Telangana’s request, expanding the terms to reallocate the undivided share between Andhra Pradesh and Telangana and to make project-specific allocations. The tenure has been extended repeatedly, most recently from August 2025.
Krishna makes a general point about Article 3. When Parliament divides a state it divides territory, assets, liabilities and employees — but a river entitlement is not divisible by the same logic, and no reorganisation statute has ever carried a satisfactory formula for it. Every future bifurcation inherits this problem.
Mahadayi: an award that solved nothing
The Mahadayi — the Mandovi in Goa — rises in Karnataka and reaches the sea in Goa. It is a short, monsoon-dependent river with a fragile estuary, which gives the ecological objection real force.
Karnataka’s Kalasa-Banduri Nala scheme would divert water from two tributaries into the Malaprabha, a Krishna sub-basin, for drinking water to Hubballi-Dharwad and Belagavi. Goa’s objection is that a diversion out of the basin is categorically different from a share within it.
The tribunal was constituted on 16 November 2010; its award of 14 August 2018 allocated Karnataka 13.42 TMC (of which 5.5 for drinking water), Maharashtra 1.33 and Goa 24, and was notified on 27 February 2020.
Notification ended nothing. Goa filed a special leave petition in 2023; the diversion works require environmental and forest clearance touching the Mhadei Wildlife Sanctuary and the tiger reserve around it; and the tribunal has taken seven extensions to deal with clarificatory references. The dispute is now a fixed feature of electoral politics in both states — precisely the outcome an adjudicated allocation was meant to foreclose.
Mahanadi, Vamsadhara and the eastern disputes
Mahanadi — Odisha and Chhattisgarh. Odisha’s complaint is that barrages and industrial withdrawals in upper Chhattisgarh have reduced non-monsoon flows into the Hirakud command. The tribunal was constituted in 2018 after Odisha moved the Supreme Court and has not reported; its tenure now runs to 13 January 2027, extended at the joint request of both states — which says something about how urgently either wants a decision.
Vamsadhara — Andhra Pradesh and Odisha. A 1962 agreement provided for 50:50 sharing; Andhra Pradesh’s Gotta barrage was completed in 1977 and its proposed Neradi barrage upstream is the live issue.
Odisha complained in 2006, the tribunal was constituted in 2010 under Justice Mukundakam Sharma, a draft decision came in 2017 and the report in 2021, reaffirming the split and permitting Neradi subject to Odisha assisting with land acquisition.
The report has not been notified and Odisha has not provided the land. The dispute is therefore decided, not notified and not implemented simultaneously.
Mullaperiyar: the dispute that goes to the Supreme Court because it is not about water
The Mullaperiyar dam stands in Kerala, on a Kerala river, but is operated by Tamil Nadu under a 999-year lease of 1886 between Travancore and the Madras Presidency, renewed by agreement in 1970.
Kerala’s case concerns the safety of a masonry dam nearly 130 years old and the permitted water level, not apportionment; Tamil Nadu’s is that the dam irrigates five southern districts with no alternative source.
Because the dispute concerns dam height and safety it is not a “water dispute” under Section 2(c), and so falls outside Article 262’s exclusion. It has been litigated in original jurisdiction under Article 131 — State of Tamil Nadu v. State of Kerala, Original Suit 3 of 2006.
2006 — the Court permitted the level to rise from 136 to 142 feet; Kerala replied with an amendment to its Irrigation and Water Conservation Act capping it at 136.
7 May 2014 — a Constitution Bench held the Kerala amendment unconstitutional, on the same reasoning as the Cauvery Reference: a state cannot legislate to undo a judicial determination binding on it. The level was restored and a three-member Supervisory Committee constituted, strengthened with independent technical members in 2021.
On 6 May 2025 a bench of Justices Surya Kant, Dipankar Datta and N. Kotiswar Singh directed both states to act on the Committee’s recommendations within two weeks, finding no justification for inaction.
Mullaperiyar is the control case for the whole field. Take the same two states, the same water and the same politics, and remove the Article 262 exclusion: what results is faster adjudication and clearer orders, and still no durable settlement, because the obstacle was never the forum.
Yamuna, Delhi and river-linking
The Yamuna is governed not by a tribunal but by an agreement — the memorandum of understanding of 12 May 1994 among Haryana, Uttar Pradesh, Rajasthan, Himachal Pradesh and Delhi, administered by the Upper Yamuna River Board.
Delhi’s recurring summer crises, acute in 2024, are argued as an inter-state dispute about releases at Hathnikund and Wazirabad. The Supreme Court in June 2024 declined to fix quantities and sent the parties back to the Board — telling them, in effect, that a basin institution exists and should be used. The Board’s weakness is that it allocates but does not operate.
The Haryana-Rajasthan memorandum of June 2026, carrying 1.119 billion cubic metres of Rajasthan’s 1994 entitlement by underground pipeline from Hathnikund to Sikar, Churu and Jhunjhunu, is the counter-example: two aligned governments settled by agreement in months what adjudication has not settled in decades.
River interlinking raises a different question — not who gets the existing flow, but whether a basin’s water may be moved out of the basin at all.
The National Perspective Plan of 1980 and the National Water Development Agency identified thirty links, sixteen peninsular and fourteen Himalayan.
In In Re: Networking of Rivers (27 February 2012) the Supreme Court directed time-bound implementation and constituted a Special Committee to oversee it — the Court setting national water policy where the executive had not.
The Ken-Betwa Link Project, the first to be implemented, rests on a 2005 tripartite memorandum among the Union, Madhya Pradesh and Uttar Pradesh, revived by a fresh agreement in 2021 and given its foundation stone in December 2024. At about ₹45,000 crore it involves the 77-metre Daudhan dam, promises 10.62 lakh hectares of irrigation and drinking water for 62 lakh people, and submerges roughly 98 square kilometres of the Panna Tiger Reserve.
The Godavari-Cauvery link, which Andhra Pradesh pressed at the Southern Zonal Council in August 2026 as a means of transferring more than 500 TMC southward, would if built change the arithmetic of the Cauvery dispute itself.
Interlinking converts a horizontal dispute into a vertical one: the donor state resists, the recipient supports, and the Union decides — which is why it attracts the Union and alarms the states with surplus.
The Supreme Court’s re-entry and the comparative argument
The constitutional oddity is stark. Article 262 permits Parliament to exclude the Supreme Court; Parliament did exclude it; and the Supreme Court is now the most consequential actor in the field. The route is Article 136, which reaches any tribunal and which the Court treats as a constitutional power a statute cannot remove.
The case against re-entry has three limbs.
It defeats the constitutional choice. The one field the drafters allowed Parliament to close, Parliament closed; reopening it by discretionary appeal makes the exclusion nugatory.
It wastes the tribunal’s work. Every award is challenged, the record re-argued and years of specialised hearing re-litigated by generalist judges, at large cost in public money and time.
It is done at the expense of the Court’s own docket. Special leave was meant to be exceptional and is now routine, with tens of thousands of special leave petitions pending at any time.
Pratap Bhanu Mehta’s characterisation of the Indian judiciary as a “self-perpetuating institution” — one that does not pass up an opportunity to enlarge its role — is the sharpest available account of why this happened, and it is not answered by pointing to the merits of individual judgments.
The case for re-entry is equally real: the alternative was nothing at all. Every decisive movement in the Cauvery dispute came from a Supreme Court order — the tribunal in 1990, the interim award, its publication, the 2013 notification, the 2018 scheme. Without Articles 136 and 32, an unnotified award would be entirely unreviewable and the Union’s inaction beyond challenge.
The comparison with the United States is where the Indian debate usually lands, and it needs stating precisely.
United States
India
Constitutional route
disputes fall within the Court’s original jurisdiction
excluded from the Court by statute under Article 262
Appellate re-entry
none — original jurisdiction is the whole of it
Article 136 reopens every award
Preferred instrument
inter-state compacts under the Compact Clause, consented to by Congress
tribunal awards imposed on unwilling states
Settled by agreement
dozens of water compacts in force
a handful, most colonial or pre-reorganisation
Effect on behaviour
litigation is slow, costly and final, so states negotiate
litigation is slow, costly and not final, so states litigate
The mechanism behind the American difference is incentive, not virtue. Because the Court will hear the case once and that is the end of it, states prefer to write their own compact. In India, because there is always another forum, there is always a reason not to settle.
Fali S. Nariman’s proposal follows directly: abolish the tribunal system and restore the pre-1956 position, returning these disputes to the Supreme Court’s original jurisdiction.
For Nariman’s proposal
Against
Ends the two-stage process in which every award is re-tried
The Court has no more hydrological expertise than a tribunal and fewer resources to acquire it
A judgment under Article 131 is final and executable as the Court’s own decree
Original suits under Article 131 have themselves taken decades — Assam’s boundary suits date from 1988 and 1989
Removes the Gazette-notification veto the Union holds
Adds a large, fact-heavy, permanently recurring class of case to an overloaded docket
Restores a single accountable forum
Abandons the drafters’ judgment that apportionment is not justiciable
Would push states toward negotiated compacts
The absence of compacts reflects political incentives, not the want of a forum
The Punchhi Commission took the opposite view — keep the tribunal but fix it: a permanent multi-member body, interdisciplinary membership, and basin-level river basin organisations with real authority. The lapsed Amendment Bill was the partial legislative expression of that view.
Boundary and territorial disputes
Boundary disputes differ from water disputes in one decisive respect. Under Article 3, only Parliament can alter a state’s boundaries, and it may do so without a state’s consent, needing only to refer the bill to the legislature for its views. There is therefore a body with unquestioned power to settle every boundary dispute in India, and it does not use it — for exactly the reason the Union does not use Entry 56.
Maharashtra and Karnataka over Belagavi
The States Reorganisation Act 1956 transferred Belgaum district and parts of Bijapur, Dharwad, Kalaburagi and Bidar from Bombay State to Mysore. Maharashtra claims 865 villages with Marathi-speaking majorities were wrongly allotted.
The Mahajan Commission, appointed under former Chief Justice Mehr Chand Mahajan in October 1966 and reporting in August 1967, recommended transferring 264 villages to Maharashtra while leaving Belgaum city and 247 villages with Mysore. Karnataka accepted the report; Maharashtra rejected it, and the Union has never laid it before Parliament — so it binds nobody.
Maharashtra’s Article 131 suit, filed in 2004, has been pending for over two decades. Karnataka’s preliminary objection is that boundary alteration is exclusively Parliament’s function under Article 3 and therefore not a justiciable question of legal right — an objection never finally decided.
The dispute is renewed politically at intervals. Karnataka renamed Belgaum as Belagavi, built the Suvarna Vidhana Soudha there and holds a winter session of its legislature in the city, treating it as a second capital. The 2022 escalation — ministerial visits blocked, vehicles attacked on both sides, resolutions in both assemblies, a truce brokered by the Union Home Minister — showed how quickly a dormant claim can be reactivated when it suits either state.
The North-East
These disputes share an origin: boundaries drawn by colonial notifications for revenue and administrative purposes, inherited as state borders by the North-Eastern Areas (Reorganisation) Act 1971 and its successors, across terrain where land is held communally and mapped poorly.
Dispute
Origin
Recent developments
Assam–Mizoram
Assam relies on the 1933 notification, Mizoram on the 1875 Inner Line
26 July 2021 firing at Vairengte–Lailapur in which six Assam police personnel were killed and about fifty injured — the deadliest inter-state incident since Independence; forces withdrawn in August 2021; about 16.6 sq km claimed as encroached
Assam–Meghalaya
twelve areas of difference dating from Meghalaya’s separation in 1972
memorandum of 29 March 2022 settling six of the twelve, with the Survey of India demarcating; regional committees for the rest; the Mukroh firing of November 2022 interrupted the second phase
Assam–Arunachal
an 804 km boundary and a claim that foothill areas were transferred without consent in 1951
Original Suit 1 of 1989 still at trial; the Namsai Declaration of July 2022 reduced the claim from 123 villages to 86; a memorandum of 20 April 2023 resolved 71
Assam–Nagaland
claims beyond the 1925 notification
Original Suit 2 of 1988 still at trial; violent flashpoints in 1979, 1985 and 2014; no settlement
The Assam-Meghalaya and Assam-Arunachal agreements of 2022–23 are the one genuine success in this field, and the method is worth extracting: regional committees with a fixed remit, five agreed criteria including historical evidence, ethnicity, contiguity, administrative convenience and the wishes of the people concerned, and the easy areas taken first to build the habit of agreement. That is the River Boards Act’s philosophy, arrived at seventy years later by another route.
The remaining disputes and the AP–Telangana residue
Haryana–Himachal Pradesh over the Parwanoo belt; Ladakh–Himachal Pradesh over the Sarchu plateau, dormant until Ladakh became a separate union territory in 2019 and revenue jurisdiction turned live; Andhra Pradesh–Odisha over the Kotia group of about 21 villages, where the Supreme Court ordered status quo in 1968, both states run parallel administrations, and Odisha filed contempt proceedings in 2021 after Andhra Pradesh held panchayat polls there.
The Andhra Pradesh Reorganisation Act, 2014 is the most detailed division ever attempted in Indian federalism, and its residue is a catalogue of what statutory reorganisation cannot do.
Section 5 made Hyderabad the common capital for a period not exceeding ten years. The period expired on 2 June 2024; Hyderabad is now Telangana’s alone. Andhra Pradesh’s own capital at Amaravati was begun, abandoned and revived in the interval, so the state functioned for a decade without a settled seat of government.
Schedule IX listed about 90 state public sector undertakings and Schedule X about 142 institutions for division. Many remain undivided, with the Sheela Bhide Committee’s recommendations only partly implemented and disputes continuing over headquarters, assets and cash balances.
Employee allocation under the Kamalanathan Committee was largely completed but generated years of litigation over seniority and options; pension liabilities, power dues and vacated buildings remain contested.
The lesson: Article 3 lets Parliament create a state in a day and leaves the two states to spend a generation dividing what the old state owned.
Why boundary disputes persist
The Union will not impose a settlement. Every boundary dispute has a losing side, and both sides are states whose parties the Union governs, allies with or hopes to displace. Parliament has the power and no incentive.
There is no standing adjudicatory forum. Water at least has tribunals; boundaries have no equivalent — no permanent commission, no statutory process, no timeline. Each dispute produces an ad hoc commission whose report either side may reject, as Maharashtra rejected Mahajan.
Article 131 suits sit for decades. Assam’s suits of 1988 and 1989 are still at trial; Maharashtra’s of 2004 has not passed the preliminary objection. A remedy that takes forty years is not a remedy.
The disputes are politically useful unresolved. A live claim mobilises a constituency at no cost; a settled one creates a losing constituency at considerable cost. Both governments therefore prefer the dispute to any available settlement of it.
The coordination machinery: a full audit
The Inter-State Council under Article 263
The text. If at any time it appears to the President that the public interests would be served by the establishment of a Council charged with the duty of —
(a) inquiring into and advising upon disputes which may have arisen between States;
(b) investigating and discussing subjects in which some or all of the States, or the Union and one or more States, have a common interest; or
(c) making recommendations upon any such subject, and in particular for the better co-ordination of policy and action,
it shall be lawful for the President by order to establish such a Council, and to define its duties, organisation and procedure.
The ease of creation is the point. No legislation is required — a presidential order suffices, and that order may define the Council’s duties, organisation and procedure. There is no constitutional obstacle of any kind to a strong, active Inter-State Council; there never was.
The Constituent Assembly expected this to be the primary route, and the drafting shows it: dispute resolution is the first duty listed, before consultation and coordination.
The facility went unused for forty years. The Setalvad ARC (1969) recommended its establishment; the Rajamannar Committee (1969) wanted it permanent and statutorily entrenched; the Sarkaria Commission, reporting in January 1988, recommended a permanent Inter-Governmental Council charged with the duties in clauses (b) and (c) — but not with clause (a) dispute resolution. It was finally constituted by presidential order on 28 May 1990 under the V.P. Singh government — a minority government dependent on regional parties, which is not a coincidence.
Composition: the Prime Minister as Chairman; all Chief Ministers; Chief Ministers of union territories with a legislature and Administrators of those without; Governors of states under President’s rule; and six Union Cabinet Ministers nominated by the Prime Minister.
The Standing Committee, created at the Council’s second meeting on 15 October 1996, is chaired by the Union Home Minister with five Union Ministers and nine Chief Ministers, and was last reconstituted on 19 May 2022. The Inter-State Council Secretariat, established in 1991 under the Home Ministry, was merged with the Zonal Council Secretariat in 2018 — an economy that also signalled the Council’s diminished standing.
The meeting record is the criticism.
First meeting
10 October 1990
Meetings in thirty-six years
eleven
Meetings in the first sixteen years
ten
Gap between the tenth and eleventh
12 December 2006 to 16 July 2016 — a full decade
Most recent meeting of the Council
16 July 2016
Standing Committee, most recent sessions
25 November 2017 and 25 May 2018, completing consideration of the Punchhi Commission’s 273 recommendations across seven volumes
Meetings since 2016
none
A body constitutionally charged with resolving inter-state disputes has not met in a decade, and has never in its existence taken up a single inter-state dispute.
The structural critique explains the record rather than merely deploring it.
Advisory and recommendatory only — nothing it decides binds anyone.
Convened entirely at the Union’s discretion; no state or group of states can requisition a meeting.
Chaired by the Prime Minister, so the Union presides over questions to which it is usually a party.
No fixed calendar, no statutory minimum frequency, no independent secretariat with research capacity, no adjudicatory power despite clause (a), and no follow-up on its own recommendations.
A political disincentive rarely stated but decisive: convening the Council seats every opposition Chief Minister in one room with the national press outside. The cost of meeting falls on the Union and the benefit accrues mostly to the states.
Displacement has compounded disuse. The functions the Council might have performed migrated to bodies the Union finds easier — the GST Council for taxation, NITI Aayog’s Governing Council for development, Zonal Councils for regional irritants, and bilateral negotiation for the rest.
Reform proposals converge on a short list: a statutory footing; a mandatory minimum of meetings, three a year being the demand several states have made; a professional independent secretariat; power to take up disputes suo motu, which clause (a) already contemplates; the Punchhi Commission’s proposal that states be consulted through the Council before Concurrent List bills are introduced; and an annual report on implementation of its recommendations.
An institution that can be created by a signature and needs no law to strengthen has been left weak by choice, not by constraint.
Zonal Councils
The idea was Nehru’s, proposed during the 1956 debate on the States Reorganisation Commission’s report: that the reorganised states be grouped into zones with an advisory council, so as, in his phrase, “to develop the habit of cooperative working” among them.
Five Zonal Councils were created under Part III of the States Reorganisation Act 1956. They are statutory, not constitutional — a real difference from the Inter-State Council.
Goa, Gujarat, Maharashtra, Dadra and Nagar Haveli and Daman and Diu
Southern
Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, Telangana, Puducherry
North Eastern Council (separate, under the North Eastern Council Act 1972)
the seven North-Eastern states, with Sikkim added in 2002
Composition. The Union Home Minister is Chairman of each of the five Councils — the horizontal body is chaired by the centre. The Chief Ministers serve as Vice-Chairman by annual rotation. Members are the Chief Minister and two other ministers nominated by the Governor from each state, with two members from each union territory. Advisers include each state’s Chief Secretary and a development commissioner. Standing Committees of Chief Secretaries meet between sessions and are in practice the most productive layer of the structure.
The stated objectives are frank about what the Councils were for: national integration; arresting the growth of acute state consciousness, regionalism, linguism and particularistic tendencies; enabling the centre and states to exchange ideas and experience; and building a climate of cooperation for development projects.
Their competence expressly includes the horizontal disputes — any matter of common interest in economic and social planning, any matter concerning border disputes, linguistic minorities or inter-state transport, and any matter arising out of the reorganisation of states.
The historic indictment. The Councils were created in the same statute and the same year as the reorganisation whose consequences they were to absorb — above all the consequences for rivers, since the Home Minister who inaugurated them, Govind Ballabh Pant, made the point that rivers do not follow linguistic boundaries. For six decades not a single water dispute was ever referred to a Zonal Council, and the Councils met erratically or not at all.
The standard verdict of total irrelevance now needs qualifying, and this is the most significant recent change in India’s horizontal machinery.
2004–2014
2014–2025
Zonal Council meetings
11
28
Standing Committee meetings
14
33
Average frequency
two to three a year
about six a year
Issues discussed
—
1,580
Issues resolved
—
1,287, about 83%
What is being settled in them is bilateral and practical: the Masanjore dam and Taiyabpur barrage between Jharkhand and West Bengal, the Indrapuri reservoir, asset and liability division between Bihar and Jharkhand outstanding since 2000, and long lists of border policing, transport and banking-access issues.
The claim that no water dispute has ever reached a Zonal Council is no longer accurate. At the 31st Southern Zonal Council at Kovalam near Mahabalipuram on 20 August 2026, the agenda included the Cauvery and Krishna disputes, the Mekedatu project, Mullaperiyar dam safety and the proposed Godavari-Cauvery link, alongside delimitation and fiscal devolution. At the 32nd Northern Zonal Council in Faridabad on 17 November 2025, water resource management among Punjab, Haryana, Himachal Pradesh and Rajasthan was on the agenda. The Union Home Minister’s framing of the Councils as “engines of cooperation” rather than discussion forums is a claim about that transformation, and the numbers give it some support.
The qualification to the qualification. The Councils are settling administrative irritants, not allocative disputes. Nothing at a Zonal Council has divided a river, moved a boundary or bound a state, and their agenda is set by the ministry that chairs them — so what is discussed is what the centre is content to discuss. The revival is real; it has not touched the hard cases.
The GST Council: the one that works, and why
Article 279A created the GST Council as a constitutional body, the Union Finance Minister as chairman and every state’s finance minister a member. Four features distinguish it from every other coordination body in India.
Constitutional status — it exists by force of the Constitution, not an executive order.
A defined voting rule — decisions need three-quarters of weighted votes, the Union holding one-third and all states together two-thirds. Neither side can act alone.
A bounded subject matter — rates, exemptions, thresholds, model laws; it is not a general forum for grievances.
A real incentive to agree — the tax does not function unless the Council decides, so the cost of deadlock falls on every member including the Union.
Article 279A(11) requires the Council to establish a mechanism to adjudicate disputes arising out of its recommendations. It has never been operationalised — one more instrument that exists on paper. Disputes have gone to the Supreme Court instead, as in Union of India v. Mohit Minerals (2022), which held the Council’s recommendations persuasive and not binding.
The institutional lesson, stripped of the fiscal argument: coordination bodies work when membership is compulsory, the subject is bounded, the voting rule is defined, and the cost of failure falls on everyone. The Inter-State Council satisfies none of these conditions and the Zonal Councils only the first.
The development channel and the other coordination bodies
The National Development Council, created by executive resolution in 1952 and comprising the Prime Minister, all Chief Ministers and Union Ministers, approved the Five Year Plans. It was never constitutional or statutory, has not met since 2012, and lapsed with the Planning Commission in 2014.
NITI Aayog’s Governing Council — the Prime Minister, all Chief Ministers and Lieutenant Governors — is the successor, meeting roughly annually as a consultative session rather than a decision-making body, since NITI Aayog allocates no funds. Its Regional Councils, contemplated for issues affecting groups of states, have rarely been convened; regional coordination has happened in the Zonal Councils instead. Its Chief Ministers’ sub-groups on centrally sponsored schemes and skill development were a genuine innovation in giving states drafting responsibility, but the device has not been used at scale since.
Other Article 263 bodies have been created more often than is recognised, but almost entirely for sectoral purposes — the Central Council of Health and Family Welfare, the Central Council of Local Government, the Transport Development Council. None was given clause (a) dispute functions.
Conferences of Chief Ministers, Chief Secretaries, Directors-General of Police and Chief Justices are executive convenings without institutional standing — flexible and unaccountable in equal measure. The National Integration Council, constituted in 1961, has not met since 2013.
Basin and technical bodies are the one place where India has something resembling operating machinery, and it is a scatter rather than a system.
The Central Water Commission is technical adviser, data authority and project appraiser — not a regulator and not a basin authority.
The Bhakra Beas Management Board, under Section 79 of the Punjab Reorganisation Act 1966, administers the Bhakra, Nangal, Pong and Pandoh works for six governments. It is the closest India has to an operating basin authority, and the 2025 confrontation showed that its writ depends on the host state’s police not obstructing it.
The Betwa River Board, Tungabhadra Board, Upper Yamuna River Board, Brahmaputra Board, and the Krishna and Godavari River Management Boards are project-specific bodies with no common design — what India has instead of the general river board system the 1956 Act contemplated.
The Mihir Shah Committee (2016) proposed the missing architecture: merging the Central Water Commission and the Central Ground Water Board into a National Water Commission with divisions for irrigation reform, river rejuvenation, aquifer mapping, water security, water quality and water data management and transparency. It has not been implemented.
Article 131 as an inter-state route, and why it fits badly
Article 131 works well for a question of legal right — whether a statute is competent, whether an agreement binds, whether an authority acted within its powers.
It works badly for the disputes in this article, for four reasons.
It is adversarial where the answer is a sharing arrangement. A court declares who is right; a basin needs a schedule of releases both sides can live with in a dry year.
It is polycentric. Changing one state’s allocation changes every other party’s position and every downstream project, and adjudication cannot trace those consequences as a negotiated plan can.
It is slow to the point of futility.
It produces a decree that still needs a willing executive — the same wall the tribunal awards hit.
The proviso to Article 131 excludes disputes arising out of pre-Constitution agreements still in operation — precisely the character of the instruments underlying Cauvery and Mullaperiyar, and a question the Court had to work around in both.
Why every instrument exists and none of them works
The pattern is identical across the whole apparatus, and stating it in one place is the argument of this article.
Instrument
What it was for
What became of it
Entry 56, Union List
central regulation of inter-state rivers
no general law ever enacted
River Boards Act 1956
prevention through joint basin planning
not one board in seventy years
Article 263 Council
inquiring into and advising on inter-state disputes
eleven meetings in thirty-six years; no dispute ever taken up; none since 2016
Zonal Councils
absorbing the consequences of reorganisation, water above all
revived after 2014 for administrative irritants; no allocative dispute settled
Tribunal awards
binding decisions with the force of a decree
three unnotified; the notified ones unimplemented
Section 6A schemes
machinery to give effect to awards
used once, for the Cauvery, and only under court order
Article 279A(11)
GST dispute resolution
never constituted
Article 307
authority to enforce the internal market
never appointed
Article 3
parliamentary power to settle boundaries
never used to resolve a contested boundary
The 2019 Amendment Bill
permanent tribunal, mediation, data bank
lapsed in 2024; not reintroduced
The failure is political, not constitutional. No provision needs amending for any of these instruments to work. Each was designed adequately and left unused.
Two incentive structures explain almost all of it.
For the Union, adjudicating between two states it must live with produces no gain and a certain loss in whichever state the decision goes against. Inaction is costless in a way action is not, and the costs of inaction fall on farmers and city water supplies rather than on the government that chose it.
For a state government, accepting an adverse settlement means paying a domestic price for a national public good that nobody will compensate. A chief minister who signs is punished at home; one who litigates for twenty years is not.
The two incentives are mutually reinforcing. Because the Union will not decide, the states have no reason to compromise; because the states will not compromise, any decision will be resisted, which confirms the Union’s reluctance to make one.
The rare exceptions identify the conditions for success. Assam-Meghalaya and Assam-Arunachal (2022–23) — aligned governments, an active Home Ministry, easy cases first, a defined committee process. Haryana-Rajasthan on the Yamuna (2026) — aligned governments and a technical solution that expanded the pie. The GST Council — compulsory membership, a bounded subject, a voting rule and a shared cost of failure. In every case the ingredient is either political alignment or a rule that makes deadlock expensive for everyone; institutions supplying neither will fail whatever their powers.
Reform: what would actually change the outcome
Institutionalise coordination so that meeting is a duty rather than a favour. A statutory Inter-State Council with a mandatory minimum of meetings, a published agenda, a right of requisition for a specified number of states, an independent secretariat, suo motu jurisdiction over disputes, and mandatory reference of Concurrent List bills before introduction.
Rebuild the water machinery on the lines of the lapsed Bill and beyond it.
A single permanent tribunal with multiple benches, so that expertise, registry and precedent accumulate instead of being disbanded with each dispute.
Interdisciplinary membership — hydrologists, agronomists, economists and environmental scientists sitting as members, not assessors.
A mandatory mediation stage with a fixed timeline, so adjudication is the second step rather than the first.
A national water data bank with a single authoritative assessment of basin yields, removing the largest source of tribunal delay.
Automatic bindingness on the date of the award, ending the Union’s Gazette veto, with a statutory duty to frame an implementation scheme within a fixed period.
Make implementation automatic and justiciable.Standing basin authorities with operational control of the relevant works — measuring, releasing and recording default — rather than committees that request compliance; a distress-sharing formula written into every award, since the disputes always erupt in deficient years; and a defined consequence for default short of the constitutional nuclear options, such as withholding assistance for the project concerned or direct operation of headworks by the basin authority.
Shift the frame from apportioning a fixed quantity to jointly managing a shared basin.
River basin organisations with real authority, as the draft National Water Framework Bill 2016 and the Model River Basin Management Bill propose, with state representation and a professional secretariat.
Demand-side management is where the water actually is. Every one of these disputes is at bottom a demand that exceeds a supply which cannot be increased.
Cropping-pattern change — paddy in the Cauvery delta and in Punjab, sugarcane in the Krishna basin, consume a share of the water wholly disproportionate to their contribution to income.
Irrigation efficiency — flood irrigation loses a large fraction of delivered water; micro-irrigation and canal lining shift the balance more than any reallocation of shares would.
Groundwater regulation, brought into the reckoning by the 2018 Cauvery judgment and still almost entirely absent in practice, and pricing and metering — politically impossible and technically decisive.
A tribunal can divide 740 TMC. It cannot make 740 TMC go further. Only the states can, and no institution currently asks them to.
Revive the preventive instrument. The River Boards Act needs no amendment to be used. Constituting even one board, on one basin, would be the cheapest institutional reform available in Indian federalism.
Conclusion: horizontal weakness as a cause of vertical centralisation
India’s inter-state disputes are not evidence that the Constitution failed to anticipate them. Article 261 provided for comity, Part XIII for the common market, Article 262 for water, Article 263 for a council, Article 3 for boundaries, and the two statutes of 1956 for prevention and adjudication alike. The apparatus is complete. What it has lacked, in every case, is an actor whose interests are served by using it.
The consequence runs back into the vertical axis, and this is where the subject turns.
When two states cannot settle between themselves, the dispute travels upward — to the Union for a tribunal, a notification, a scheme; or to the Supreme Court for a direction or an award.
Every such journey adds to the centre’s stock of authority without anyone deciding to give it any. The Union acquires the power to notify or not, to constitute or not, to release or withhold; the Court acquires the power to rewrite an award.
Centralisation by default is harder to resist than centralisation by design, because there is no amendment to oppose and no bill to defeat — only a vacuum that something has to fill.
The states’ own behaviour completes the circle. A state that refuses to settle with its neighbour is choosing, whether it means to or not, to have the matter settled by the centre or the Court. The rhetoric of state autonomy and the practice of inter-state intransigence point in opposite directions, and the second defeats the first.
A federation is measured not only by how it divides power between levels but by how its units deal with one another. By the first measure India has been argued over for seventy-five years; by the second the record is thinner than the constitutional text promised, and the gap between them is where the work remains.
A federation whose units cannot settle with each other hands the centre a power it never had to ask for.
Previous Year Questions
Discuss the composition and functions of the Inter-State Council. To what extent has this body been successful in achieving its objectives? (2022)
Despite constitutional mandate, the Inter-State Council has not come of age. Discuss. (2020)
Answer in 150 words: Mechanism for settling inter-state disputes. (2020)
Examine the efficacy of available mechanisms for resolving Inter-State disputes in India. (2012)
