On 5 April 1961, Prime Minister Jawaharlal Nehru laid the foundation stone of the Sardar Sarovar dam at Navagam in Gujarat, at a ceremony that presented the structure as a temple of the new republic’s engineering ambition. Twenty-four years later, in 1985, a doctoral researcher from the Tata Institute of Social Sciences in Mumbai named Medha Patkar entered the Narmada valley to study the villages the reservoir would drown, and the encounter produced the Narmada Bachao Andolan, the movement that would contest that foundation stone’s promise for more than three decades. This article defends one claim about the contest. The Andolan operated as citizen-renegotiation conducted through the Sardar Sarovar Project’s own instruments: the tribunal award that fixed the dam’s dimensions, the loan covenants that financed it, the writ petition that challenged it and the statute book that eventually absorbed its demands. Five features mark the operation. A movement was built village by village across three states between 1985 and 1989. Its villagers pledged at Manibeli in 1991 to drown rather than move. Its documentation forced the World Bank’s first independent review and the loan’s cancellation in March 1993. Its lawsuit produced the Supreme Court’s conditional permission of 18 October 2000, construction tied to rehabilitation at every step. Its premises entered national law in the land acquisition statute of 27 September 2013, four years before the dam it fought was completed and dedicated in September 2017.

Two established ways of reading this record dominate the shelf. The first reads the Andolan through displacement: a development project imposed its costs on adivasi and peasant communities who had no share in its benefits, and the movement was their resistance. The second reads it through the dam’s own biography: a project conceived in 1946, stalled by an interstate dispute, awarded by a tribunal in the late 1970s, financed in 1985, litigated through the 1990s and finished in 2017, with the movement as one obstacle in a long engineering chronology. Both readings capture real content, and both are engaged at full strength in the pages that follow. Neither captures the mechanism. Displacement happened, and the dam has a history, but the Andolan’s distinguishing work was renegotiation: it seized the legal and financial instruments through which the Indian state had committed itself, the award’s rehabilitation clauses, the Bank’s safeguard covenants, the Constitution’s Article 21, and held the state to the terms those instruments already contained. The valley’s residents appeared in that process not as victims awaiting rescue and not as an obstruction to a national asset, but as parties demanding that a bargain be rewritten because its original terms had been struck without them.

That claim carries a discipline. Every position in the argument that follows is anchored to a named holder, a dated document or a recorded event. The movement’s case is stated through Medha Patkar’s correspondence and the Andolan’s own documents, preserved at its archive in Baroda, and through Sanjay Sangvai’s movement history, The River and Life, published in 2000. The case for the dam is stated through the Gujarat government’s declared position, the project authority’s benefit figures and the reasoning of the Supreme Court’s majority. The scholarly disagreement is adjudicated on the primary record: the report of the Bradford Morse independent review, submitted in June 1992; the judgment in Narmada Bachao Andolan v Union of India, delivered on 18 October 2000; and the text of the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act of 2013. Where the holders of rival figures disagree, above all on how many people the reservoir displaced, the figures are printed with their holders’ names attached and never averaged into a compromise the record does not support.

The reanalysis also observes a distinction the original controversy often blurred. A movement can be judged by its demands, by its methods or by its effects, and the three judgments need not coincide. Judged by its maximal demand, the Andolan failed: the dam it sought to stop stands complete. Judged by its methods, it was among the most consequential campaigns of the Indian republic, inventing or perfecting a repertoire, enumeration from below, the drowning pledge, the documentary address to financiers, the stage-by-stage judicial audit, that later movements inherited wholesale. Judged by its effects, it sits between the two verdicts, and the effects are the hardest to total because so many of them were registered in other institutions’ names: a bank’s accountability machinery, a court’s conditions, a parliament’s statute. This article’s claim concerns the third judgment. It holds that the effects were not incidental by-products of a failed resistance but the predictable yield of the movement’s chosen strategy, which was to conduct politics as the renegotiation of instruments the state had already signed. The five features are the strategy’s dated instances, and the map that closes the argument is offered so that the claim can be checked, feature by feature, against the record each section assembles.

A Foundation Stone in the River: The Project Before the Movement

The Sardar Sarovar Project was an interstate bargain before it was a dam. The Narmada rises in Madhya Pradesh, flows west through Maharashtra’s Khandesh country and enters Gujarat before reaching the Gulf of Khambhat, which made its waters a shared asset among riparian states with opposed interests. Gujarat, whose Saurashtra and Kutch districts suffered repeated drought, sought a high dam near the river’s end to carry water into its canal system. Madhya Pradesh, which held most of the catchment and most of the future reservoir, resisted a storage large enough to drown its own valley for another state’s irrigation. The dispute outlasted the 1950s and was referred in October 1969 to the Narmada Water Disputes Tribunal, constituted under the Inter-State Water Disputes Act of 1956. The tribunal deliberated for a decade and delivered its final award in 1978; the award was notified in the Gazette of India in December 1979, and its terms structured everything that followed.

The award’s terms deserve statement because the movement later fought inside them. The tribunal fixed the dam’s full reservoir level at 455 feet, 138.68 metres, a height that has never since been revised. It divided 28 million acre feet of utilisable flow at 75 percent dependability: 18.25 million acre feet to Madhya Pradesh, 9 million acre feet to Gujarat, 0.5 million acre feet to Rajasthan and 0.25 million acre feet to Maharashtra. It divided the 1,450 megawatts of power the project was designed to generate in the proportions 57 percent to Madhya Pradesh, 27 percent to Maharashtra and 16 percent to Gujarat. On the human side of the ledger, the award wrote a rehabilitation code into an engineering document. Every family losing its home to the reservoir was to receive a minimum of two hectares of irrigable land, with land made available in the command area for those willing to migrate, house sites and civic amenities at resettlement colonies, and the whole resettlement completed in advance of submergence rather than after it. Rehabilitation, in the tribunal’s drafting, was not charity appended to the project. It was a condition of the project’s legality, sequenced before the water rose.

Why is the Sardar Sarovar dam 138.68 metres high?

The height is the tribunal’s, not any single decade’s engineering preference. The Narmada Water Disputes Tribunal fixed the full reservoir level at 455 feet in its award, balancing Gujarat’s demand for storage against Madhya Pradesh’s resistance to submergence, and every later contest was fought over permission to build up to that figure in stages.

Converted, 455 feet is 138.68 metres, the figure at which the dam was finally completed in 2017. No government, court or project authority has since held power to revise it unilaterally, because the award was an adjudication between states under a constitutional statute, and its terms bound the parties as a decree binds litigants. That fixity shaped the entire politics of the project. The movement could not argue the height down without reopening an interstate adjudication, and the project’s champions could not argue it up; both sides were confined to the sequence, the conditions and the count, which is where the Andolan chose to fight.

Why did Gujarat champion the dam while Madhya Pradesh bore the reservoir?

The award distributed costs and benefits asymmetrically, and the politics followed the distribution. Of the 245 villages in the reservoir’s path, 193 stood in Madhya Pradesh, 33 in Maharashtra and 19 in Gujarat, while the canal command and the drinking water lay overwhelmingly in Gujarat. Each state’s stance toward the dam maps onto that arithmetic almost exactly.

Gujarat therefore organised its politics around completion, with every major party in the state committed to the dam and its governments in the project’s vanguard. Madhya Pradesh governments of both national parties spoke for the project in Delhi, where the state’s water and power shares were at stake, and for their own displaced villagers at home, where the submergence fell. Maharashtra, holding the smallest shares of water and land, produced both the movement’s most famous village, Manibeli, and its most equivocal official record. The asymmetry was the award’s own design, and the movement’s three-state architecture was built to answer it.

The planning that followed the award had a hole at its centre. Project documents of the late 1970s counted 6,147 families in the submergence zone, a figure derived from revenue records that recognised settled landholders and largely missed the valley’s actual population of adivasi cultivators, tenants, encroachers on forest land, fisherfolk and the landless. By the time of the Supreme Court proceedings the official count of project-affected families had risen to about 41,000 families, some 205,000 people, and even that count excluded the canal network’s displaced. The undercount was not a clerical detail. Rehabilitation obligations under the award attached to counted families; a family absent from the list held no entitlement the state was bound to honour. When Patkar’s field teams began their village surveys in 1985, the gap between the revenue record and the inhabited valley was the first fact they documented, and it became the evidentiary basis of the movement’s opening demand: count the affected honestly, then rehabilitate them land for land, before any water rises.

The project’s lineage ran back further than the tribunal. Basin investigations on the Narmada began under the Central Waterways, Irrigation and Navigation Commission in the late 1940s, when the river was surveyed as a system, and the planning tradition that followed contemplated a staircase of major dams, Bargi, Indira Sagar and Omkareshwar among them, with the terminal structure at Navagam last and largest. The dam Nehru blessed in 1961 was a smaller structure than the one the tribunal would award in 1978; the height grew through the interstate bargaining of the 1960s, as Gujarat pressed for storage sufficient to carry water to Kutch and Madhya Pradesh resisted the submergence a high dam entailed. The dispute that consumed the 1960s, and that the tribunal was appointed in 1969 to end, was therefore present before the first concrete: a river whose benefits and burdens fell in different states could only be dammed by adjudication, and the Sardar Sarovar was born as a judgment before it existed as a wall.

Finance arrived in April 1985, when the World Bank approved a package of 450 million United States dollars, an International Bank for Reconstruction and Development loan of 250 million dollars and an International Development Association credit of 200 million dollars, to fund the dam and the first section of the canal system. The agreements carried covenants on resettlement and environmental protection drawn from the Bank’s own safeguard policies, and they made the Bank a party to the rehabilitation question whether it wished to be or not. Construction followed environmental clearance from the Union Ministry of Environment and Forests in June 1987, a clearance granted on the condition that the studies it presupposed, on catchment treatment, flora and fauna, seismicity and health, be completed alongside the works. The Sardar Sarovar Narmada Nigam Limited, incorporated in 1988, executed the project for Gujarat. Upstream, the Bargi dam, the first large dam completed on the river in 1990, supplied the valley with a preview: its displaced villages were still pressing rehabilitation claims years after its reservoir filled, a record the Morse review would later cite as evidence of what Sardar Sarovar’s oustees could expect. When the movement organised, therefore, it did not confront a vague apparatus of progress. It confronted a tribunal award with unfulfilled clauses, a loan with unmonitored covenants and a clearance with unperformed studies, and it organised its politics as an audit of all three.

Feature One: The Narmada Bachao Andolan Takes Shape, 1985 to 1989

Medha Patkar came to the valley as a researcher, not as a campaigner. A student of social work at the Tata Institute of Social Sciences who had worked with Mumbai’s pavement and slum communities, she entered the Narmada’s Maharashtra villages in 1985 to gather material for doctoral research on the dam’s affected population, and found that the families in her interview schedules did not know the terms on which they were to be displaced, or in many cases that displacement was planned at all. The research converted itself into organisation. In Maharashtra the Narmada Dharangrast Samiti, the committee of the dam-affected, carried village-level work; in Madhya Pradesh the Narmada Asargrast Samiti performed the same function across the Nimar districts that held most of the reservoir; in Gujarat, where the dam’s supporters dominated public life, smaller groups of affected families in the nineteen submerging villages organised under the same valley-wide coordination. The names matter less than the method. The samitis were committees of the affected themselves, village by village, and their first programme was informational: to carry the award’s text, the project’s maps and the states’ rehabilitation promises into villages that had received none of them, and to carry the villages’ own counts of families, houses, wells, temples and grazing lands back out.

The coordination had a second register beyond the committee room. In 1986 the valley organisations ran a literacy campaign, the Pratham Saksharta Andolan, across the affected villages, pairing the ability to read with the ability to read a government notice, a land record and a compensation receipt. The sequence was deliberate. A movement that intended to renegotiate a legal bargain needed constituents who could inspect the bargain’s documents, and the Andolan’s formative years were spent manufacturing that competence at scale. The correspondence of those years, Patkar’s exchanges with Baba Amte at Anandwan and with Sundarlal Bahuguna of the Chipko movement, preserved at the Andolan’s archive in Baroda and at the Anandwan archives in Warora, shows the coordination being built letter by letter: questions of tactics, of fasting, of relations with the state governments, argued out in writing between three organisers who agreed on the river and differed on much else. Movements that preserve their own paper create the conditions for their history to be audited, and the instinct is the same one that leads investigative reporters to keep source material in encrypted local custody rather than on exposed servers, a practice set out in the account of how journalists use VaultBook to protect sources. The Baroda correspondence would later matter forensically, because it dated the movement’s demands and refuted the claim, made repeatedly in the 1990s, that the Andolan had shifted its position opportunistically. The letters show the position hardening on a schedule set by the states’ own performance.

Baba Amte’s adherence gave the young organisation a standing no valley committee could have supplied alone. Amte had built Anandwan in Maharashtra’s Chandrapur district into the country’s best-known centre for the treatment and rehabilitation of people affected by leprosy, and his moral credit with governments and publics was of a different order from an activist’s. In 1990 he left Anandwan and settled at Kasrawad, on the Narmada’s bank in Madhya Pradesh’s Khargone district, announcing that he would live among the people the reservoir would displace. The move converted the Andolan’s claim from a regional grievance into a national question, and Amte’s presence at Kasrawad through the following decade, until age and illness withdrew him, supplied the movement with an elder whose fasts and statements the national press could not ignore.

What part did Baba Amte play in the Narmada movement?

Amte supplied the movement with moral standing and a base on the river itself. Having built Anandwan into India’s best-known leprosy rehabilitation centre, he left it in 1990 and settled at Kasrawad on the Narmada’s bank in Madhya Pradesh, living among the families facing submergence and lending the Andolan a lifetime’s accumulated public trust.

His letters, fasts and public statements carried the Andolan’s case to audiences that trusted his decades of service among people affected by leprosy, and his correspondence with Patkar through the 1990s helped settle the movement’s tactics at its hardest junctures, including the sequencing of the Manibeli pledge against the legal campaign. When Amte spoke of the dam, governments that could dismiss an activist had to answer a man whose own life’s work was the rehabilitation of the stigmatised, and who testified that the state’s rehabilitation promises in the valley did not resemble the practice he had spent forty years building.

By September 1989 the valley committees were ready to act as one organisation. At a rally at Harsud in Madhya Pradesh, attended by tens of thousands of valley residents, the state samitis declared themselves a single front under the name Narmada Bachao Andolan, Save the Narmada Movement, and the unified organisation adopted the posture that the record of the following years displays: no displacement without rehabilitation completed in advance, on the award’s own terms, and no faith in promises that the states would not first convert into land, house sites and written entitlements. The number the organisation carried was 245 villages across the three states, each with a committee, a count and a claim. It was this federation, not a charismatic individual, that the World Bank’s reviewers and the Supreme Court’s judges would later have to answer, and its architecture explains the movement’s durability. Arresting Patkar, which the states did repeatedly, did not dissolve the samitis, because the samitis predated her prominence and held their own village mandates.

Why did the Andolan move from demanding rehabilitation to opposing the dam itself?

The hardening was evidentiary, not ideological at origin. Through the late 1980s the Andolan’s stated demand was the award’s own bargain: a full count of the affected and land-based rehabilitation completed before submergence. When Madhya Pradesh stated in writing that it lacked the land to resettle its share, the movement concluded that the dam could not lawfully proceed.

The steps between the two positions are documented in sequence. Village surveys kept showing families missing from official lists. Resettlement sites inspected by the samitis kept failing the award’s standards on land quality, water and amenities. The state administering 193 of the 245 villages acknowledged that it did not possess replacement land of the quantity and quality the award required for oustees who declined migration to Gujarat, and Maharashtra’s record was thinner still. The Morse review, examining the same files four years later, reached the finding the movement had built its case on: resettlement at the promised scale was not being achieved and, on the evidence, could not be achieved within the project’s timetable. Opposition to the project itself followed from the states’ own record, as the second iteration of a minimal demand, honour the award, after the award’s terms had been shown to be beyond the implementing states. Critics of the Andolan, including the Gujarat government, answered that the movement had always intended to kill the project and used rehabilitation as an instrument. The correspondence and petition record predate that charge and answer it in sequence: the no-dam position is dated, in the movement’s own documents, after the land failure was documented, not before.

Feature Two: Manibeli, the Sangharsh Yatra and the Jal Samadhi Pledge

The movement’s second feature was a tactic that converted the reservoir’s arithmetic into a vow. Manibeli, a Bhil village in Maharashtra’s Nandurbar country, stood first in the submergence sequence on the Maharashtra side, and in 1991 its residents, with villagers including Bava Mahalia and Khunchand Wasawe among the oath-takers, adopted the pledge of Jal Samadhi: when the water came, they would not leave. The pledge’s meaning requires care. It was not a threat of self-harm in the manner of a protest fast’s escalation; it was a refusal to vacate, stated in the idiom of the river’s own sacred geography, declaring that the state’s timetable would have to reckon with bodies in houses rather than with empty structures on a survey map. Jal Samadhi re-priced submergence. A family removed by police to a resettlement colony was an administrative act; a family standing in rising water in its own doorway was a political fact that television cameras and visiting reviewers could record, and the Morse team’s itinerary the following year took in the valley’s villages with the pledge already public.

The pledge did not stand alone. In the same season the Andolan organised its Sangharsh Vahini, the struggle corps that carried coordination between villages during confrontations, and in December 1990 it had already tested the geography of the conflict with the Sangharsh Yatra, a march that set out on 25 December 1990 from Rajghat in Madhya Pradesh and walked toward the dam site at Kevadia in Gujarat. The Gujarat government of Chief Minister Chimanbhai Patel sealed the state border, and the march was stopped at Ferkuva, the border village on the Madhya Pradesh side, in the first days of January 1991. For weeks the marchers held position at the border as negotiations and arrests alternated; at the end of January 1991 the Andolan withdrew the column, announcing that the question had been put to the country and that the struggle would continue by other means. The yatra’s significance was diagnostic. It demonstrated that the dam’s politics were interstate in the full sense: one state would deploy its police to keep another state’s displaced citizens from reaching the structure that displaced them, and the Union government, formally the project’s partner and the award’s guarantor, would not intervene. Citizen-renegotiation, if it was to happen at all, would have to proceed against that alignment, through instruments higher than any single state.

The monsoon of 1993 supplied the pledge’s first test. As the reservoir rose behind the partly built dam, water entered Manibeli, and villagers kept their houses as the level climbed, standing in the water in the images that travelled the world that season. Houses in the village went under in the monsoons of 1993 and 1994 while their residents refused the resettlement offered, and the state governments’ response oscillated between negotiation and the arrest of the movement’s organisers. Patkar herself was arrested repeatedly across the decade, and the Andolan’s organisers in Maharashtra and Madhya Pradesh faced charges ranging from obstruction to unlawful assembly. The repression record matters to the analytical reading. A movement negotiating through courts, loan covenants and tribunal clauses simultaneously maintained a bodily presence in the submergence zone that made every administrative deadline a potential confrontation, and the combination, paper above, bodies below, was the Andolan’s operating system rather than a split personality. The paper campaign produced the Morse review. The bodies ensured that the review’s findings could not be treated as an abstraction.

The pledge’s repertoire outlived Manibeli’s own submergence and became the movement’s standard answer to each new height band. At Jalsindhi in Madhya Pradesh and Domkhedi in Maharashtra, villages facing the rising reservoir in the monsoons of 1999 and 2000, residents repeated the Manibeli form, standing in the water as it entered their houses in satyagrahas that the national press covered day by day, and the tactic’s migration up the reservoir’s bands marked the movement’s continuing presence at exactly the points where the ladder of heights met inhabited ground. A generation later the form appeared again at Chikhalda in Madhya Pradesh in September 2017, when families whose rehabilitation remained disputed stood in the rising reservoir as the gates stood closed, in the last of the drowning protests before the dedication. The continuity is the point the feature makes. A tactic invented in one Maharashtra village in 1991 remained serviceable for twenty-six years because the underlying transaction never changed: the state proposed to convert inhabited homes into reservoir by timetable, and the inhabitants answered by making the timetable carry their bodies in public view. Each repetition restated the movement’s founding claim in its most economical form, that the people in the water were parties to the bargain, not debris in its path.

The declarations made at Harsud fixed the movement’s public doctrine for the decade that followed. The rally resolved that the affected would not accept displacement on the basis of cash compensation in place of land, that no village would negotiate separately with the states so long as the enumeration remained disputed, and that the movement’s quarrel was with the terms of the taking rather than with the idea of the river’s use. Those resolutions explain a pattern outsiders often misread: the Andolan’s refusal, repeated across the 1990s, to let individual villages or families settle early on advantageous terms before the collective count was resolved. Solidarity of that kind is expensive to maintain, and it was maintained by the samitis’ village mandates renewed in mass meetings each season, the organisational labour that feature one of this article’s map actually names. The federation also settled its division of labour at Harsud: village committees owned the ground struggle and the counts, a small coordination carried the documentation to Bhopal, Mumbai, Gandhinagar, Delhi and eventually Washington, and Patkar’s role was defined as the carrier between the two levels rather than as a leader above them. The arrest records of the following decades, in which district administrations detained the carrier repeatedly while the committees continued their work, tested that architecture and confirmed its redundancy.

Three States, Three Politics: The Interstate Life of the Dam

The movement’s three-state structure answered a three-state problem, and the differences among the state governments explain the campaign’s uneven terrain better than any account of the dam as a single national project. Gujarat’s politics were the simplest because its interest was the simplest. The state held the dam site, the canal head and the listed beneficiaries, and its submergence share was nineteen villages. Its governments therefore behaved as the project’s owner: they staffed and funded the Nigam, raised the bonds that replaced the Bank, mobilised pro-dam opinion through the 1990s and treated the movement as an external enemy. The state’s consensus crossed every line that divided its politics on other questions. Congress governments laid the foundation and negotiated the loan; Bharatiya Janata Party governments completed the structure; and the state’s representatives in Delhi pressed the same file under administrations of both colours. When the Gujarat government sealed its border against the Sangharsh Yatra in January 1991, it was performing the owner’s right to exclude, and the performance was popular at home in a way that constrained every later Gujarat politician who might have sought a settlement.

Madhya Pradesh was the project’s reluctant giant. It held 18.25 million acre feet of the water and 57 percent of the power, and its capital at Bhopal defended the award as a Madhya Pradesh achievement. It also held 193 of the 245 villages, the reservoir’s whole human cost in administrative form, and its district administrations in the Nimar country faced the movement’s samitis across the table for thirty years. The state’s governments, Congress under Digvijaya Singh through the 1990s and the Bharatiya Janata Party under Shivraj Singh Chouhan thereafter, spoke in both registers at once. They defended the project in Delhi and in the Control Authority, and they pleaded incapacity in the rehabilitation record, stating at intervals from the late 1980s onward that the state lacked the land the award required. Madhya Pradesh also produced the record’s sharpest internal conflict: its oustees were offered migration to Gujarat’s command area as the award contemplated, and large numbers refused, preferring claims in their own state, so that the state’s obligation could not be discharged by export. The movement’s Madhya Pradesh organisation was correspondingly its strongest, rooted in the Nimar districts where the Bargi precedent and the land shortage were matters of direct family knowledge.

Maharashtra held the smallest stake and the most symbolic ground. Its share of the water was 0.25 million acre feet and of the power 27 percent, and its submergence villages, thirty-three of them in the Akrani and Akkalkuwa country of Nandurbar district, were adivasi communities with the valley’s weakest connections to the state’s political centre in Mumbai. Manibeli’s fame made Maharashtra the movement’s postcard and its government’s embarrassment. Successive administrations in Mumbai oscillated between negotiation with the Andolan and enforcement for the project, and Maharashtra’s rehabilitation record, for the fewest families of any party state, was repeatedly found wanting by reviewers who noted that small numbers were no excuse for the defaults the record showed. The state’s Grievance Redressal Authority, headed by a retired judge, became one of the busiest dockets in the project’s machinery precisely because Maharashtra’s affected families were few enough for each default to be individually visible. The three-state pattern also explains the Union government’s posture across the whole period. Delhi was simultaneously the award’s guarantor, the loan’s borrower, the clearance’s grantor and the appellate forum for every failure, and its ministries managed the contradiction by proceduralising it: committees, sub-groups, oversight groups and authorities multiplied, each converting a political impasse into a docket. The movement understood this earlier than most. Its strategy of addressing a different instrument at each stage, award, covenant, writ, statute, was a map of the Indian state’s own divided custody of the dam.

Feature Three: The Morse Review and the World Bank’s Exit, 1991 to 1993

The movement’s third feature was won in Washington, and it began with the loan covenants of April 1985. The Bank’s agreements with India required resettlement and environmental performance to standards the Bank had adopted for itself, and the Andolan’s international allies, organised through networks that included the Environmental Defense Fund in the United States, spent the late 1980s placing the valley’s documentation before the Bank’s executive directors. In June 1991, Bank President Barber Conable announced the first independent review in the World Bank’s history: a team chaired by Bradford Morse, a former administrator of the United Nations Development Programme, with Thomas Berger, a former judge of the Supreme Court of British Columbia, as its legal member. Morse and Berger toured the valley in the winter of 1991 to 1992, took evidence from the movement, the states and the project authority, and submitted their report, Sardar Sarovar: The Report of the Independent Review, on 18 June 1992.

The report’s findings were categorical, and they were findings about paper. Morse and Berger concluded that the project’s resettlement and environmental components had failed to meet the Bank’s own policies and the loan agreements’ conditions; that the basic data on the affected population, on submergence and on environmental impact did not exist in usable form; that the record showed a pattern in which the Bank and the borrower had proceeded as if compliance could be achieved after construction rather than before it; and that, in their central sentence, the wisest course was for the Bank to step back from the projects and consider them afresh. The review also stated that the projects could not be completed except by unacceptable means if the affected people were pressed into resettlement the states could not deliver. William F. Fisher’s 1995 edited volume, Toward Sustainable Development?: Struggling Over India’s Narmada River, assembled the scholarly argument around the review and remains the standard collection on the controversy the report ignited. The episode is the clearest instance in the whole record of the dynamic described in the foreign-policy-as-domestic-politics framework: an instrument of external finance, negotiated between a government and an international institution, was converted by a domestic movement into a lever that moved both.

The Government of India’s official answer to Morse is part of the record and is stated here in its holders’ terms, because the renegotiation reading does not require the report to have been unanswerable. Delhi’s response held that the review had exceeded its function, had relied excessively on the movement’s materials, had understated the rehabilitation performed and the project’s benefits, and had judged an interstate award’s implementation by standards the Bank’s policies had never been applied to retrospectively in any other borrowing country. Indian officials also objected, with some force, that the review’s remedy would punish the oustees it pitied, since a Bank withdrawal stripped the project of the one financier whose covenants obliged the states to perform rehabilitation at all. The objection identified a real cost, and the post-1993 record, in which rehabilitation supervision passed to domestic bodies of uneven rigour, gave it weight. What the objection could not answer was the report’s evidentiary core, the village-level findings on enumeration and land that the movement’s files had supplied and the review’s own fieldwork had confirmed. The argument of 1992 was therefore not between truth and falsehood but between two true statements: the review’s, that the project as documented could not lawfully proceed, and Delhi’s, that the remedy would dissolve the only external discipline the project had. The cancellation of March 1993 enacted both statements at once, ending the discipline and conceding the documentation.

The Bank’s management initially resisted its own reviewers. On the day the report was released, the Bank’s presidency issued a statement maintaining that continued support for the Narmada projects was justified, and management proposed remedial action plans with new benchmarks rather than suspension. Donor governments and legislators in the United States and Europe pressed the other way, and the Bank’s board set performance benchmarks for the following months. The benchmarks were not met to the board’s satisfaction, and the end came in the form the diplomatic record preserves. On 30 March 1993, the Government of India requested cancellation of the undisbursed portion of the 1985 loan package, about 170 million dollars of the 450 million committed, and the Bank announced the cancellation the following day, stating that India wished to proceed with the project on its own, without the financial assistance under the Bank loan. Formally, the borrower withdrew. Substantively, the withdrawal executed the Morse recommendation by the only route that preserved both institutions’ dignity: the loan died, the safeguards regime attached to it died with it, and the project passed wholly into domestic finance and domestic law. Gujarat responded by raising money from the Indian public, floating Sardar Sarovar bonds through the Nigam from 1993, and the dam’s financing thereafter answered to Indian bondholders rather than to Washington covenants.

The renegotiation did not end with the exit; it changed venue and left institutions behind. The Bank’s humiliation over Narmada was a founding trauma of its subsequent safeguard architecture: the Inspection Panel, created in 1993 as the Bank’s first accountability mechanism for people affected by its projects, was a direct institutional consequence of the review model Morse had pioneered, and the World Commission on Dams, launched in 1998 under the chairmanship of Kader Asmal with Medha Patkar among its commissioners, carried the valley’s questions into a global audit of large dams whose report, Dams and Development, appeared in November 2000. The Andolan’s own reading of the episode, set out in Sangvai’s The River and Life, treats the Bank’s departure as proof that documentation defeats concrete when the documentation is addressed to an institution with rules of its own. The record supports the narrower and stronger version of that claim. The movement had not persuaded the Indian state to stop the dam. It had identified the one party to the project whose charter obliged it to answer evidence, and it had made that party answer.

The interval between the report and the cancellation deserves its own record, because it was the renegotiation’s most compressed season. Through the second half of 1992 the Bank’s board, pressed by the United States Congress, where hearings on the Bank’s Narmada lending had put the review’s findings into the legislative record, and by European executive directors answerable to parliaments with their own valley constituencies of concern, required management to demonstrate that benchmarks on resettlement and environmental work could be met by the spring of 1993. India’s government faced the symmetrical pressure in reverse: to certify compliance was to accept external supervision over an interstate project in documentary detail, and to refuse was to lose the loan. The states, whose performance was the benchmark’s object, supplied returns that Delhi could not certify. The cancellation request of 30 March 1993 resolved the triangle by dissolving it. The movement had not lobbied for that precise outcome; it had demanded the Bank enforce its covenants, and enforcement, pursued honestly, produced the outcome on its own. The distinction matters to the reading defended here. Citizen-renegotiation works by obliging institutions to apply their own rules. The rules, applied, sometimes end the relationship they were meant to govern, and the Andolan’s archive treats March 1993 not as an expulsion it engineered but as a verdict the project’s paperwork delivered on itself.

Feature Four: The Supreme Court Verdict of 18 October 2000

With the Bank gone, the contest moved to the forum the Constitution supplied. In 1994 the Andolan filed Writ Petition (Civil) Number 319 in the Supreme Court of India, challenging the project on the award’s own terms and on Article 21: the environmental clearance of 1987, it argued, had been granted without the studies the clearance itself required, and the rehabilitation the award made a precondition was not occurring. The Court’s interim orders froze construction in January 1995 at the height then reached, about 80.3 metres, and the dam stood still for four years while the case was argued, the longest judicial pause in the project’s history and itself a renegotiated fact: the state’s timetable thereafter ran through a courtroom. Interim orders in 1999 permitted the height to rise to 85 metres, and the final hearing produced judgment on 18 October 2000 in Narmada Bachao Andolan v Union of India, reported at (2000) 10 SCC 664. The bench comprised Chief Justice A. S. Anand and Justices B. N. Kirpal and S. P. Bharucha. The majority, in judgments delivered by Anand and Kirpal, dismissed the petition and permitted construction to proceed under conditions. Bharucha dissented. The holdings deserve separate statement, because the verdict’s political life was lived in its conditions.

The majority held, first, that the tribunal’s award was final and binding on the party states and that the project’s height and parameters, settled by that adjudication, were not open to revision by the Court under the guise of a writ petition. Second, displacement by a public project did not of itself violate Article 21: the oustees, on the majority’s reasoning, would on rehabilitation have access to amenities superior to those of their valley villages, and their gradual assimilation into the mainstream of settled agricultural life would work their betterment. Third, the environmental clearance of 1987 was not vitiated merely because some studies remained incomplete; the clearance had contemplated studies proceeding alongside construction, and the Court declined to hold the project hostage to their completion. Fourth, and most consequentially for the following two decades, construction was tied to rehabilitation as a running condition: the dam could be raised to 90 metres forthwith, and thereafter in stages of five metres, each stage requiring the clearance of the Narmada Control Authority’s environment and rehabilitation sub-groups and the completion of rehabilitation for the families affected at that stage in advance of submergence, in accordance with the award. Fifth, the Court endorsed the machinery of the Grievance Redressal Authorities that the states had constituted under retired judges, to which oustees could carry complaints on enumeration, land allotment and compensation, and it recorded with approval the liberalised packages by which Gujarat and Madhya Pradesh had extended entitlements beyond the award to major sons, encroachers and landless families. Sixth, the Court warned against the conversion of public interest litigation into an instrument for obstructing projects of national importance, a passage of the majority judgment that entered the general jurisprudence of Indian development litigation far beyond the Narmada.

What does pari passu mean in the Sardar Sarovar rehabilitation orders?

Pari passu means step for step. In the tribunal’s award and again in the 2000 judgment, rehabilitation was required to proceed pari passu with construction and to be completed for each height stage before that stage’s submergence: as the dam rose, the families in the corresponding band had first to be enumerated, allotted land and resettled.

The phrase converted rehabilitation from a promise about the project’s end into a condition on its next increment. A government could no longer plead the project’s overall progress as an answer to an individual family’s default; the obligation attached to the five metres in front of it. The movement’s post-2000 litigation consisted largely of proving, stage by stage and family by family, that the step had not been taken, and the Grievance Redressal dockets filled with exactly that proof. When later benches and committees asked whether an increment could proceed, pari passu was the test they were obliged to apply, which is why two Latin words carry so much of this history.

Bharucha’s dissent held the opposite position on the decisive point of law. On his reading of the record, the environmental clearance of 1987 had been granted without the basic studies, on flora and fauna, on seismicity, on health effects, on the extent of submergence itself, that the law required to precede clearance, and a clearance granted in ignorance was no clearance at all. He would have directed that all further construction stop until a fresh environmental assessment, conducted on data actually collected, supported a lawful clearance. The dissent is the verdict’s second life. The majority decided the case; the dissent supplied the terms in which the verdict would be criticised for twenty years, and its insistence that the state may not plead its own failure to study as a reason to proceed became a standard citation in Indian environmental argument. The judgment also stands in the line of cases through which the Indian judiciary defined the limits of its supervision over national infrastructure, and it is read most profitably against the most-consequential-constitutional-moment framework: a court that possesses the power to stop a project of this scale, and declines to use it while multiplying the conditions attached to its continuation, is allocating constitutional authority as surely as a court that intervenes.

Patkar’s response to the verdict, issued the day it was delivered, called it illogical, dangerous and anti-people, and the Andolan’s campaign continued in the arenas the judgment had left open: the Grievance Redressal Authorities, the Narmada Control Authority’s sub-group clearances, and periodic returns to the Court. The returns defined the ladder’s jurisprudence. When permission to raise the dam toward 100 metres was challenged in 2002 on the ground that families in the corresponding band stood unrehabilitated on the government’s own records, the matter was first adjourned amid differences on the bench and then dismissed by a bench headed by Kirpal, by then Chief Justice and the author of the 2000 majority’s principal judgment, on the ground that the Andolan could not maintain the complaint in the oustees’ place: affected families were to carry their own grievances first to the Grievance Redressal Authorities constituted for them. The movement read the ruling as the closing of the public interest door through which it had entered in 1994, and civil liberties commentary of the period agreed, noting that the judgment’s logic required the least literate litigants in the country to exhaust a bureaucratic remedy before the Court would hear that the remedy had failed them. The episode belongs in the verdict’s balance sheet. The 2000 conditions had made the authorities the machinery of first resort; the 2002 dismissal made them, for the oustees’ collective claims, close to the machinery of only resort, and the movement’s post-2002 energy flowed accordingly into the dockets family by family. The post-2000 height ladder was correspondingly slow and contested. Clearance to 95 metres and then to 100 metres followed in the years after the verdict, each preceded by litigation over whether rehabilitation for the corresponding band was complete. The authority permitted 110.64 metres in 2004. In March 2006 the Control Authority cleared the raise to 121.92 metres, and Patkar undertook an indefinite fast in Delhi that ended only when Prime Minister Manmohan Singh constituted an Oversight Group under the former Comptroller and Auditor General V. K. Shunglu to verify rehabilitation in Madhya Pradesh; the group’s report questioned the state’s figures for resettled families, and construction to 121.92 metres proceeded later that year under the Court’s supervision, with the judges recording that rehabilitation must keep pace. The verdict had not ended the renegotiation. It had proceduralised it, giving the movement a docket, a ladder of heights and a standing invitation to prove default, stage by stage, which it used for seventeen more years.

The five frozen years between the petition and the judgment had their own politics, which the verdict’s text tends to eclipse. With construction halted at about 80.3 metres from January 1995, the dam existed in suspension: a wall across the river that stored nothing to its design, gates ungrown, the reservoir’s fringe villages living under a submergence notice that neither executed nor expired. Governments changed in Delhi and in all three states during the freeze, and each inherited the suspension as a fact to be managed rather than a decision to be owned. The litigation years also produced the controversy’s ugliest legal episode. In 1999, as interim orders permitted the height to rise first to 85 metres, Gujarat brought the Andolan’s public statements to the Court’s attention in contempt proceedings, and the Court’s orders of that year noticed press interviews by Patkar and passages of Roy’s The Greater Common Good in terms that the movement and civil liberties organisations read as a warning against campaigning speech about a pending case. The contempt track decided nothing about the dam, but it marked the temperature of the endgame: by 1999 the argument was no longer only about water and land, but about who might say what, in print, about a project the Court was seized of. The final judgment’s sixth holding, against the obstructionist use of public interest litigation, was written in that atmosphere, and its afterlife in Indian development litigation cannot be separated from it.

The Case for the Dam, Stated at Its Strongest

An adjudication that understates the winning side’s argument proves nothing, and the argument for Sardar Sarovar was formidable. Its foundation was thirst. The drought years of 1985 to 1987, which coincided with the movement’s founding, burned the case into Gujarat’s politics: Saurashtra and Kutch watched wells fail and villages supplied by tanker, and the state’s political class, Congress and Bharatiya Janata Party alike, committed itself to Narmada water as the permanent answer. The project authority’s stated benefits, repeated in the World Bank’s own announcement of the loan’s cancellation in March 1993 even as the Bank stepped away, were irrigation for some 1.8 million hectares across Gujarat and Rajasthan, drinking water for the domestic needs of some 30 million people, and food production sufficient for about 20 million people, from a dam generating 1,450 megawatts of peaking power shared among three states. Gujarat’s governments added the political arithmetic: 9,633 villages and 131 urban centres in the state’s water-scarce regions were listed in project documents as the drinking water beneficiaries, a constituency of tens of millions set against a displaced population the official count placed at about 41,000 families. On that arithmetic, the dam’s supporters argued, opposition was a veto exercised by the few, organised by activists headquartered outside the valley, over the drinking water of the many.

The Gujarat government’s stated position went further and indicted the movement rather than the arithmetic. Successive chief ministers, from Chimanbhai Patel, whose government sealed the border against the Sangharsh Yatra, to Narendra Modi, whose government completed the dam, maintained that the Andolan was responsible for delaying a project whose benefits were never honestly in dispute, and that the delay itself had victims: the drought districts that waited an extra generation for water. Gujarat’s political consensus on the dam was total, extending across parties and to the state’s press, and it produced a counter-mobilisation of its own, pro-dam rallies and a public culture in which the Narmada figured as the state’s lifeline and its critics as enemies of Gujarat. Madhya Pradesh’s position, by contrast, was structurally divided and both halves were genuine: the state held the largest share of the water, 18.25 million acre feet, and 57 percent of the power, and its governments defended the project in Delhi, while its valley districts supplied the movement’s mass base and its administrations conceded, in the rehabilitation record, that they could not find the land their signature on the award required. The beneficiary farmers of the command area in Gujarat’s central districts, finally, made the quietest and strongest version of the case: canal water reached their fields in the project’s early sections from the 2000s, and the prosperity of the Narmada command became a standing exhibit in the argument that the valley’s sacrifice, properly compensated, purchased a transformation visible in the state’s farm incomes.

The champions’ historical argument deserves its place beside their arithmetic. In Gujarat’s public case the dam stood in the line of Bhakra, the earlier river valley project that Nehru had consecrated as a temple of modern India, and the Sardar Sarovar’s champions presented the Narmada structure as that lineage’s completion on the western rivers: a single work that would do for Saurashtra and Kutch what Bhakra had done for Punjab. The analogy carried the project’s naming politics. The dam bears the name of Sardar Vallabhbhai Patel, Gujarat’s own architect of the Union, and its champions framed opposition to it as an affront to his memory as well as to the state’s development. The power figures reinforced the interstate version of the case: the 1,450 megawatts were divided by the award with the largest share, 57 percent, to Madhya Pradesh and 27 percent to Maharashtra, so that the project’s electricity was never Gujarat’s alone, and its defenders in Delhi could present the dam as a national asset in the strict sense of an award performed among states. The movement never denied the analogy’s emotional force. Its answer was to note what the Bhakra precedent itself contained: a displacement settled in an era before enumeration, consent and judicially supervised rehabilitation had entered Indian law, and to argue that the republic’s standards had risen since, in part because valleys like the Narmada had made them rise.

Arundhati Roy’s essay The Greater Common Good, published in May 1999, was written against exactly this arithmetic and remains the movement’s most consequential public document. Roy attacked the majoritarian calculus at its root, arguing that the counting of beneficiaries in tens of millions against the displaced in lakhs concealed the distribution of sacrifice: the costs fell on adivasi communities with the least access to the state’s protections, while the benefits flowed through canals and power grids to those already connected to them. She widened the ledger beyond the reservoir, counting the canal system’s displaced, the downstream fisheries and the sanctuary lands, and placed the full number of the affected at several hundred thousand people. The essay’s ferocity made it a target; the Supreme Court’s proceedings of 1999 noticed it in the contempt litigation that surrounded the case, and Roy was eventually convicted of contempt in a separate proceeding arising from a later affidavit, in March 2002. As an artefact of the movement, though, the essay performed a renegotiating function of its own: it transferred the Sardar Sarovar argument from the specialist literature into the country’s common reading, and no subsequent defence of the dam has been able to proceed without answering it.

The delivery record behind that case was not imaginary, and it accumulated during the years when the dam’s height was still litigated. The Narmada main canal, 458 kilometres from Kevadia toward the Rajasthan border and among the largest lined canals in the world, was completed along its length in 2008, and Narmada water first crossed into Rajasthan that year, reaching the Barmer country that project documents had listed among the driest beneficiary districts. Branch canals and the drinking water grid extended through the following decade into Saurashtra and Kutch towns whose tanker dependence had been the project’s founding argument, and Gujarat’s governments cited each extension in the completion campaign. The movement’s answer to the delivery record was distributional rather than dismissive: it pointed to the gap between the canal’s head reaches, where command area prosperity was visible, and the tail reaches and the submergence zone, where the promises ran the other way, and it treated the sequence, water delivered to the beneficiaries as rehabilitation lagged for the displaced, as the award’s pari passu principle inverted. Both exhibits were real. The prosperity in the command and the arrears in the valley coexisted in the same project accounts, and the argument between them is the argument this article adjudicates in its closing sections, not an argument either side’s exhibits can settle alone.

The Environmental Ledger: Forests, Catchment and the Studies That Lagged

The rehabilitation argument has a twin that ran beside it for forty years, and the verdict turned on it as surely as on the land question. The Sardar Sarovar reservoir at full level submerges roughly 37,690 hectares, the greater part of it in Madhya Pradesh, including forest land in the Shoolpaneshwar tract of Gujarat and the riverine forests of the Nimar, together with the fields, grazing commons and village sites of the 245 villages. The environmental clearance of June 1987 was granted on conditions that read, in retrospect, like the movement’s own charge sheet: the catchment area treatment, compensatory afforestation, flora and fauna studies, seismic surveillance and health studies that the clearance required were to be completed pari passu with construction, the same step-for-step formula the award applied to rehabilitation. The Morse review examined the environmental file in 1992 and found the same pattern it found in resettlement: the studies were years behind the works, baseline data on the river’s ecology had not been assembled, and the institutions charged with the conditions lacked the staff and the mandate to enforce them against the construction schedule. The Narmada Control Authority’s Environment Sub-group, chaired by the Union Environment Secretary, existed to clear each stage, and its clearances through the 1990s were granted on the strength of action plans whose completion the next stage’s papers would again record as pending.

The substantive losses were argued in specifics rather than in generalities about nature. The Shoolpaneshwar sanctuary, notified in 1982 in the Bharuch country and later expanded and renamed for the sloth bear, stands downstream of the dam in the project’s impact zone, and its management through the filling years was a standing item of environmental contention. The river’s downstream ecology, the fisheries of the estuary and the ingress of salinity in the Gulf of Khambhat reach, entered the argument through Roy’s essay and the movement’s documentation of fishing communities whose catch the altered flows diminished, a population the official enumeration never counted as affected because no land of theirs was taken. Seismicity supplied the legal argument’s sharpest edge: the dam stands in a seismic zone whose assessment Bharucha’s dissent found inadequately studied, one of the specific gaps, alongside flora and fauna inventories and public health baselines, that in his judgment voided the 1987 clearance. The majority answered from the record of the clearance process itself, holding that the sanction had contemplated studies in progress rather than studies completed, and that expert bodies seized of the conditions could be trusted to enforce them as construction advanced. The disagreement was therefore not between environment and development as slogans. It was between two accounts of what a conditional clearance means: a licence that lapses when its conditions are unperformed, as the dissent read it, or a continuing supervision that tolerates arrears, as the majority practised it. Indian environmental law has lived inside that disagreement ever since, and the Narmada file, with its four decades of studies trailing works, remains its fullest documentary exhibit. The movement’s contribution to this ledger mirrors its contribution to the rehabilitation one. It did not invent the conditions. It kept the only continuous public record of their non-performance, monsoon by monsoon, until the record itself became a document no clearance could ignore.

Feature Five: The 2013 Land Acquisition Act as Institutional Residue

The movement’s fifth feature is the one in which its premises outlived its campaign. On 27 September 2013, the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act received the President’s assent, enacted by the United Progressive Alliance government in its second term, with Jairam Ramesh as the Union Minister for Rural Development who had piloted the bill, and it entered into force on 1 January 2014, replacing the Land Acquisition Act of 1894 under which the Sardar Sarovar acquisitions had proceeded. The statute’s architecture reads as a codification of the demands the valley had pressed for twenty-five years. Acquisition for private companies required the consent of 80 percent of affected families, and for public-private partnership projects 70 percent. A Social Impact Assessment, conducted in public and appraised by an independent expert group, was made a precondition of acquisition, with the affected families’ enumeration built into the assessment rather than left to revenue records. Compensation was set by formula to reach up to four times the market value of land in rural areas and twice the market value in urban areas, through a distance-based multiplier compounded by a solatium of 100 percent. Rehabilitation and resettlement ceased to be a project’s discretionary welfare and became statutory entitlements scheduled in the Act itself, applying to the landless and to livelihood losers as well as to titleholders.

The genealogy of these provisions is traceable rather than rhetorical. The consent thresholds answered the valley’s founding grievance, that a bargain had been struck over inhabited land without the inhabitants. The Social Impact Assessment answered the Morse review’s central finding, that the state had acquired first and counted later, and it institutionalised the enumeration the samitis had performed by hand in 1985. The inclusion of livelihood losers among the entitled answered the award-era exclusion of encroachers, tenants and fisherfolk that the movement had documented village by village. The Act was not the Andolan’s work alone, and this article does not claim it: the land wars of the 2000s, at Singur and Nandigram in West Bengal, at Bhatta Parsaul in Uttar Pradesh and across the special economic zone conflicts, had made acquisition the most combustible question in Indian politics, and the statute answered a national insurgency of which Narmada was the longest-running theatre. The political line runs through the post-2004 UPA mandate, renewed in 2009 on a platform of inclusive growth that made a new acquisition law a governing necessity. But the drafting history matters for the analytical claim. When the state finally rewrote the rules of taking land, it wrote them in the grammar the Narmada movement had spent three decades forcing into official documents: consent, prior assessment, counted families, land-linked compensation and rehabilitation as entitlement. That is institutional residue in the strict sense, the deposit a movement leaves in the statute book after its own immediate object has been decided against it.

The residue was contested immediately. The government elected in 2014 sought through an ordinance of December 2014, and then a bill that passed the Lok Sabha in March 2015, to exempt five categories of project, including infrastructure and industrial corridors, from the consent and Social Impact Assessment provisions; the bill failed in the Rajya Sabha against opposition that mobilised the statute’s own language of consent, and the ordinance was allowed to lapse in August 2015. Several states thereafter legislated their own acquisition frameworks under the Act’s provisions for state amendment. The episode completes the feature’s meaning. A renegotiation is not finished when its terms are written into law; it is finished when attempts to write them out again must argue on the terrain the renegotiation created. By 2015, defenders of acquisition without consent were obliged to argue that consent was a delay and an obstacle, which is to say they were obliged to concede that consent was the rule.

The statute’s distance from the Sardar Sarovar acquisitions can be measured in the 1894 Act’s mechanics, under which the valley’s land was actually taken. The colonial statute required a notification of intent, a hearing of objections that the acquiring authority itself judged, an award of compensation at market value as the collector assessed it, and possession that could be taken under urgency clauses with the objections stage curtailed. It knew nothing of consent, of impact assessment, of rehabilitation as an entitlement, or of the landless as persons the acquisition concerned. Every element of the 2013 Act is a negation of a specific 1894 mechanic, and most of the negations had been formulated as demands in the Narmada petitions of the 1980s, two decades before the national land conflicts of the 2000s made them a governing party’s programme. The afterlife confirmed the statute’s contested weight. Its clause governing acquisitions that lapse when possession or compensation stalls became the subject of sustained litigation, resolved by a Constitution Bench of the Supreme Court in Indore Development Authority v Manoharlal in March 2020, which read the lapse provision against landowners in the contested cases. The 2020 reading narrowed one remedy without touching the consent, assessment and entitlement architecture, and the architecture is the residue this article claims: a generation of officials acquiring land in India works inside procedures whose defaults were set by an argument that began in 245 villages on the Narmada.

The dilution attempt’s content shows what the movement’s residue had made normal. The ordinance’s five exempted categories, defence and national security projects, rural infrastructure including electrification, affordable housing, industrial corridors, and infrastructure and social infrastructure projects including public-private partnerships, were framed by the government of the day as the minimum surgery required to make acquisition workable, which is to say the consent and assessment provisions were conceded to be the operative obstacles. The farming and movement opposition that defeated the bill in 2015 argued in the statute’s own vocabulary, treating consent as a right acquired rather than a concession granted, and state governments that later legislated their own relaxations, Telangana’s amendment of 2016 among them, sought Presidential assent under the Constitution’s repugnancy provisions, an admission that the national statute’s terms were the ones to be displaced. In this series’ terms, the episode is the renegotiation’s ratchet: terms once extracted could be contested, narrowed and litigated, but the contest had to be conducted against them, on ground the movement had staked.

The Endgame: Gates, a Full Reservoir and an Unfinished Ledger, 2000 to 2019

The dam’s completion, when it came, came through the procedural channel the 2000 verdict had dug. Each height increment after 121.92 metres required the Control Authority’s satisfaction on rehabilitation, and the final clearance, to the full 138.68 metres with the installation of the radial gates, was granted on 12 June 2014, within weeks of the change of government in Delhi, a sequence the completed dam’s champions presented as the removal of the last political obstruction and its critics as the relaxation of the last judicial one. The completion belongs to the arc traced by the integrated three-tensions framework: a Gujarati project of Gujarati consensus, carried to term by a prime minister who had governed Gujarat for thirteen years and who treated the dam’s completion as a pledge of his tenure. Construction of the remaining height proceeded, the seventeen radial gates were installed, and in June 2017 the gates were closed, permitting the reservoir to rise toward its full level. On 17 September 2017, his sixty-seventh birthday, Prime Minister Narendra Modi dedicated the Sardar Sarovar dam to the nation at Kevadia, describing it in his address as an engineering marvel built over the corpses of campaigns against it. The reservoir first filled to the full reservoir level of 138.68 metres in September 2019, fifty-eight years after the foundation stone and forty years after the tribunal’s award had specified the figure to the centimetre.

The Andolan’s terminus was quieter. Patkar’s fasts continued into the completion years, including a protest in 2017 against the gates’ closure while families in the submergence zone, on the movement’s count, remained unrehabilitated, and a further fast in 2019 as the reservoir filled. The movement’s final position, stated in its documents of those years, was that the gates had been closed over an unfinished rehabilitation ledger, that Madhya Pradesh families in particular had been pressed into cash compensation in place of the award’s land, and that the submergence of 2017 to 2019, which entered villages in the three states as the level rose, repeated the founding pattern of water first and entitlements later. The Gujarat government and the Nigam answered that rehabilitation stood complete on the official enumeration and that the remaining claims were of families ineligible under the award’s categories or unwilling to accept the packages on offer. Both statements are printed here as their holders made them. What the official record establishes without contradiction is narrower: the Grievance Redressal Authorities continued to receive and adjudicate oustee complaints after the dedication, and the enumeration disputes that began in 1985 had not been exhausted when the water reached the awarded line. The renegotiation ended, as it began, with the question of who counts.

The site of the dedication acquired a second monument in the completion years, and its presence states the endgame’s balance as plainly as any document. On 31 October 2018, the Statue of Unity, the 182-metre figure of Sardar Vallabhbhai Patel for whom the dam is named, was inaugurated on Sadhu Bet island facing the dam at Kevadia, and the colony built for the project’s engineers became a national tourism destination organised around the reservoir the movement had fought. The valley’s transformation into a landscape of pilgrimage and leisure proceeded in the same years in which the last submergence-band villages were negotiating their files before the Grievance Redressal Authorities. Both facts belong to the record. The dam’s completion was not merely an engineering event but a settlement of memory, in which the state that built the structure also curated the terms on which it would be remembered, naming, statuary and spectacle included. The movement’s counter-memory survives in its archive, its histories and the 2013 statute, and the two memories coexist in the same valley, one cast in bronze above the water and one filed in dossiers below it. A reanalysis that reported only one of them would fail the record it claims to adjudicate.

The filling itself produced the interstate argument’s last public round. In the monsoon of 2019, as Gujarat raised the reservoir toward the full level for a state celebration of the river at Kevadia on 17 September, Madhya Pradesh’s Chief Minister Kamal Nath protested in writing to the Union government that his state’s villages were facing submergence ahead of the agreed schedule while rehabilitation and compensation questions remained open, and demanded that the filling be regulated to the protocol the Control Authority had set. Gujarat’s government answered that the level was within the clearances granted and that Madhya Pradesh’s affected families had been provided for under the packages. The exchange, conducted between two state governments through Delhi in the project’s fifty-eighth year, reproduced the award-era geometry exactly: Gujarat storing, Madhya Pradesh submerging, the Union arbitrating by procedure. It also closed the endgame’s argument. The renegotiation the movement began in 1985 did not end the dam’s interstate politics. It changed their content, from whether the river would be dammed to whether the terms on which it was dammed would be performed, and on that question the files were still open when the water reached the line the tribunal had drawn in 1979.

The Archive and the Argument: How the Movement Manufactured Its Evidence

Every feature of the renegotiation ran on evidence the movement had made itself, and the evidentiary method deserves treatment as a feature of the politics rather than as background colour. The samitis’ village surveys of the late 1980s were the foundation: house-to-house counts in 245 villages, recording family size, landholding, tenancy, wells, livestock, temples and burial grounds, made because the official enumeration had been shown to undercount by an order of magnitude. When the Morse team arrived in 1991, it did not have to commission a census; it had to arbitrate between two censuses, the state’s and the movement’s, and its report’s devastating quality came from finding the movement’s numbers, on village after village, closer to the ground. The same pattern repeated at every stage. The rehabilitation default cases the Andolan carried to the Grievance Redressal Authorities after 2000 were individual dossiers, family by family, allotment by allotment. The contempt-era controversies over official claims were fought with affidavits answered by counter-affidavits built from the same village files. Sangvai’s The River and Life, published in 2000 as the movement’s own history, is best read as the public edition of that archive, and Fisher’s 1995 collection as its first scholarly processing.

The method had a politics of its own, which the framework developed for the information movement illuminates: as the anti-corruption-political-architectural-substantive-content framework describes for the right to information campaign, a movement that can compel the state to answer documents changes the relation between citizen and administration even when it loses the substantive argument. The Andolan’s version of that politics predated the information statute by two decades and ran through treaty texts and loan covenants instead of file inspections, but the mechanism was identical: take an instrument in which the state has described its own obligations, measure performance against the description, and publish the difference. Patkar’s correspondence, preserved at Baroda, shows the method being taught inside the movement, organisers instructed to record dates, officers’ names and the precise wording of oral promises, on the stated ground that the state’s memory is short and the file must be longer. The under-cited record of the movement, the Patkar, Amte and Bahuguna correspondence of 1991 and after, documents the moment this evidentiary discipline was joined to the Manibeli pledge, when the three organisers debated in writing how documentation and bodily resistance should be sequenced so that neither discredited the other. Their conclusion, visible in the practice of the following monsoon, was that the pledge gave the documents an audience and the documents gave the pledge a defence.

The state’s counter-method is also part of the record and should be stated in its holders’ terms. Project authorities and Gujarat officials maintained through the 1990s that the movement’s figures were inflated for fundraising and international consumption, that families counted by the samitis included the ineligible and the speculative, and that official enumeration, conducted under the award’s categories by revenue officers, was the only count with legal standing. The dispute was never resolved by a neutral census, and this article has printed the rival figures with their holders attached throughout. What can be said on the documents is that every independent examination that tested the two counts against the ground, the Morse review in 1992 and the Shunglu Oversight Group’s verification in 2006 among them, found material error on the official side of the ledger, including families omitted and rehabilitation recorded as complete where the group’s site visits found it was not. The renegotiation’s evidentiary asymmetry, a state with sovereign authority and a movement with better files, is the Sardar Sarovar record’s most portable lesson, and it is the ground on which the citizen-renegotiation reading principally rests.

The archive had a visual register as well, and it carried the valley into rooms the dossiers could not enter. Anand Patwardhan’s documentary A Narmada Diary, filmed through the early 1990s and released in 1995, recorded the movement from inside its own marches and monsoon standings, and its circulation through festivals and solidarity screenings built the international audience that the Bank’s directors later answered to. The film’s evidentiary value differs from the samitis’ counts and should be marked as such: it testifies to presence, sequence and conduct, to who stood where as the water rose and what the police did at the border, in a form that affidavits could not give a distant reviewer. The Andolan’s organisers understood the division of labour among their records precisely. The counts persuaded committees. The correspondence disciplined the movement itself. The images made the counts impossible to file away. When the World Commission on Dams took evidence at its hearings at the end of the decade, the Sardar Sarovar record arrived in all three registers at once, which is why a project of one Indian river became the test case through which the commission’s global argument about dams and development was conducted. The movement had spent fifteen years ensuring that, wherever the dam was discussed, its own documentation would be on the table beside the state’s, and by 2000 that condition was simply a fact of the case.

Three Ways of Reading the Andolan: An Adjudication

The development-displacement reading, whose strongest statements are Roy’s essay and the Morse report’s own framing, holds that Sardar Sarovar is a case of development’s costs being assigned to those least able to refuse them, and that the movement is intelligible as the resistance of the assigned. It explains the valley’s anger, the adivasi composition of the submergence zone and the moral charge of the Jal Samadhi pledge. What it does not explain is the movement’s behaviour in victory. A resistance movement whose object was the distribution of development’s costs should have settled, repeatedly, when the distribution improved, and the Andolan was offered improved distributions at every stage: liberalised packages beyond the award, cash enhancements, judicially supervised ladders. It refused settlements that left the underlying instruments untouched and fought hardest over enumeration, sequence and consent, the terms of the bargain rather than its price. The displacement reading also struggles with the movement’s own constituency, which included families who stood to gain from rehabilitation on the award’s terms and who organised not to escape the project but to be counted inside it correctly.

The dam-history reading, whose strongest statement is the project literature itself and the engineering record, holds that the Andolan is an episode in a sixty-year infrastructure chronology, significant chiefly for the delay it imposed. It explains the state’s behaviour: the award, the loan, the litigation and the completion form a single administrative line, and from within that line the movement does appear as friction. What it cannot explain is the residue. An episode of delay does not rewrite the national law of land acquisition, does not found the World Bank’s accountability machinery, does not put a movement leader on the World Commission on Dams and does not leave a dissenting judgment that reorganises environmental argument. The chronology reading also cannot account for the form of the state’s own conduct after 2000, when the project’s managers organised their work around clearances, dossiers and grievance dockets that the movement had forced into existence. Delay is a measure of time. The Andolan’s effects are measures of institutions.

The citizen-renegotiation reading defended in this article, which operationalises the movements-as-citizen-renegotiation framework at the Sardar Sarovar case, holds that the movement’s object from 1985 forward was the terms on which citizens displaced by a public project stand in relation to the state: whether they are counted, whether their consent or its statutory substitutes are obtained, whether the obligations written for them are performed before the water rises, and whether the law that governs the next taking carries the memory of this one. It explains the refusal of price-only settlements, the investment in enumeration, the choice of the Bank’s covenants and the award’s clauses as the battleground, and the shape of the residue in 2013. It also locates the Andolan in the series of Indian movements that worked the same seam: it shares its riverbank idiom and its elder network with the environmental-political citizen-renegotiation framework developed for Chipko, from which Bahuguna came to the Narmada correspondence, and its suspicion of the post-1991 growth settlement’s bookkeeping aligns it with the questions posed by the economic-policy-as-politics framework, since the Sardar Sarovar of the clearance era was the Nehruvian project par excellence surviving into the liberalisation decades on borrowed finance and inherited law.

The complication must be faced in its strongest form. The renegotiation reading risks aestheticising a defeat. The dam stands at 138.68 metres; the reservoir filled in 2019; the movement’s founding demand in its maximal form, that this dam at this height should not be built, was rejected by the tribunal’s award before the movement existed, by the Supreme Court in 2000 and by the electorate of Gujarat at every intervening election. Families were displaced, on any holder’s count in the tens of thousands, and the rehabilitation many received fell short of the award’s promise, a shortfall the dissent predicted and the Grievance Redressal dockets document. A reading that dwells on instruments and residue can underweight that brute content, the drowned fields and the colony plots that did not match them. The answer the record supports is that the reading does not deny the content; it specifies the terrain on which the content was decided. Development and displacement were the stakes. Renegotiation was the game actually played over them, by a movement that understood earlier than its opponents that in a constitutional republic the stakes are distributed through instruments, and that citizens who master the instruments can change the distribution even when they cannot change the outcome. The dam was built. The law of building dams on inhabited land was rewritten in the process, by the people the water displaced.

One scholarly challenge to the reading defended here comes from inside the valley rather than from the project’s defenders, and it sharpens rather than weakens the adjudication. Amita Baviskar’s ethnography of the Bhilala adivasi movement in Madhya Pradesh’s Jhabua and Alirajpur country, In the Belly of the River, published in 1995, documented a constituency whose politics were never simply the Andolan’s. The adivasi organisations of the western Madhya Pradesh hills pursued their own programme, centred on land, forest and liquor contractor questions, and their alliance with the valley campaign was negotiated, partial and at times strained by the difference between a movement against displacement and a movement for adivasi self-rule. Baviskar’s account complicates any narration in which 245 villages speak with one voice, and this article accepts the complication. The renegotiation reading does not require a unitary citizenry. It requires parties capable of holding the state to its instruments, and the valley contained several such parties, whose coordination under the Andolan’s banner was an achievement of organisation rather than a fact of nature. Rajni Kothari, the political theorist who chronicled India’s movements of the period, placed the Andolan in the line of struggles that had shifted the country’s democratic argument from the distribution of offices to the terms of the state’s presence in citizens’ lives. The ethnographic caution and the theoretical placement agree on the essential point: the movement’s significance lies not in the purity of its voice but in the machinery it built for making claims, machinery that outlasted the occasions, and the constituencies, that built it.

The Five-Feature Map

The argument of this article is summarised in the five-feature map of citizen-renegotiation through Sardar Sarovar, set out in the table below, which names each feature, its decisive dates, the institutional content it carried and the term of the original bargain that the feature renegotiated. Read with the sections above, the map is the article’s citable residue: it converts a thirty-year campaign into five dated, evidenced operations.

Feature	Decisive dates	Institutional content	Term renegotiated
One: the movement’s formation	1985 to 1989	Village samitis across 245 villages in Madhya Pradesh, Maharashtra and Gujarat; enumeration by the affected; unified front declared at Harsud in September 1989	Who counts as affected: the official list of 6,147 families against the inhabited valley
Two: the Jal Samadhi pledge	1990 to 1994	The Sangharsh Yatra halted at Ferkuva in January 1991; the Manibeli villagers’ drowning pledge of 1991, kept in the monsoons of 1993 and 1994	The timetable of submergence: removal as administration against presence as politics
Three: the Morse review and the Bank’s exit	1991 to 1993	The independent review commissioned in June 1991, report of 18 June 1992; India’s request of 30 March 1993 cancelling some 170 million dollars undisbursed of the 450 million dollar package	The loan itself: safeguard covenants enforced against the borrower and the lender
Four: the Supreme Court verdict	1994 to 2000	Writ Petition (Civil) 319 of 1994; construction frozen from 1995; judgment of 18 October 2000 permitting staged construction tied to rehabilitation, Bharucha dissenting	The sequence of the project: height increments converted into judicially supervised conditions
Five: the 2013 land statute	2013 to 2015	The Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act, assent 27 September 2013, in force 1 January 2014; dilution defeated by 2015	The law of acquisition itself: consent thresholds, prior impact assessment and rehabilitation as statutory entitlement

The map also marks the reading’s limits honestly. Every feature after the second is a renegotiation conducted inside institutions the movement did not control, a bank board, a court, a parliament, and each closed with the dam higher than it had been when the feature opened. The features are therefore not victories in sequence. They are the record of a citizenry extracting, from a project it could not stop, a succession of enforceable acknowledgements about how such projects may lawfully proceed, until the acknowledgements were thick enough to be enacted as a general law. Whether that exchange, a dam for a doctrine, was worth its price is the question the valley’s residents answered differently among themselves, and the disagreement among the affected, between those who held out at Manibeli and those who took the Gujarat packages early and built new lives in the command area, is part of the record this article preserves rather than resolves.

Frequently Asked Questions

Q: What is the Narmada Bachao Andolan?

The Narmada Bachao Andolan, the Save the Narmada Movement, is the federation of village organisations formed in 1989 by the families facing displacement from the Sardar Sarovar dam on the Narmada river. It united the state-level committees of the dam-affected in Madhya Pradesh, Maharashtra and Gujarat, which had organised since 1985 across the 245 villages in the reservoir’s path. Led publicly by Medha Patkar, the movement demanded first that the tribunal award’s rehabilitation terms be honoured in full, and then that the dam be stopped when the states conceded they could not perform those terms. Its methods combined village enumeration, litigation, fasting and bodily resistance in the submergence zone with documentation addressed to the project’s financiers and judges.

Q: Who is Medha Patkar?

Medha Patkar is the social worker and researcher who founded and led the Narmada Bachao Andolan. Trained at the Tata Institute of Social Sciences in Mumbai, she entered the Narmada valley in 1985 to conduct doctoral research on the Sardar Sarovar dam’s affected villages and remained to organise them, building the village samitis whose federation became the Andolan in 1989. Over three decades she led its marches, fasts and litigation, was arrested repeatedly, and carried its case to the World Bank’s reviewers and the Supreme Court. She served as a commissioner on the World Commission on Dams, whose report appeared in 2000, and her correspondence with Baba Amte and Sundarlal Bahuguna, preserved at the movement’s Baroda archive, documents the movement’s tactics from inside.

Q: What is the Sardar Sarovar Project?

The Sardar Sarovar Project is the terminal dam of the Narmada valley development plan, built across the river at Kevadia in Gujarat to a full reservoir level of 138.68 metres, the height fixed by the Narmada Water Disputes Tribunal’s award notified in December 1979. The project was designed to irrigate some 1.8 million hectares in Gujarat and Rajasthan, supply drinking water to 9,633 villages and 131 urban centres on the project authority’s listing, and generate 1,450 megawatts of power shared among Madhya Pradesh, Maharashtra and Gujarat. Nehru laid its foundation stone on 5 April 1961; it was financed in part by a 450 million dollar World Bank package approved in 1985, litigated for a generation, and completed and dedicated in September 2017.

Q: When did the NBA begin?

The movement’s organising began in 1985, when Medha Patkar entered the valley and the first committees of the dam-affected formed in Maharashtra and Madhya Pradesh to demand a full count of the families facing displacement and rehabilitation on the tribunal award’s terms. Through the late 1980s these state-level samitis coordinated across the 245 affected villages, ran a literacy campaign in 1986 and built the village surveys that became the movement’s evidence base. The organisations declared themselves a single front under the name Narmada Bachao Andolan at the Harsud rally in Madhya Pradesh in September 1989, attended by tens of thousands of valley residents. Both dates are real: 1985 marks the organising, 1989 the unified movement under its enduring name.

Q: What was the 2000 Supreme Court verdict on Sardar Sarovar?

On 18 October 2000, in Narmada Bachao Andolan v Union of India, a three-judge bench dismissed the movement’s writ petition and permitted the Sardar Sarovar dam to be built, by a majority of two to one. Chief Justice A. S. Anand and Justice B. N. Kirpal held that the tribunal award was final, that displacement did not of itself violate the right to life, and that construction could rise to 90 metres at once and thereafter in five-metre stages, each stage conditioned on rehabilitation for the affected band being completed in advance and cleared by the Narmada Control Authority’s sub-groups. Justice S. P. Bharucha dissented, holding that the 1987 environmental clearance was invalid because the required studies had never been done, and that construction should stop until a lawful clearance existed.

Q: What is the post-1992 World Bank withdrawal architectural moment?

The phrase names the sequence in which the World Bank’s first independent review, chaired by Bradford Morse and submitted on 18 June 1992, found that the Sardar Sarovar projects had violated the Bank’s own resettlement and environmental policies and advised the Bank to step back and consider them afresh. Bank management first resisted the finding, donor pressure mounted, and on 30 March 1993 the Government of India requested cancellation of the undisbursed balance, about 170 million dollars of the 450 million dollar package of 1985, which the Bank announced the next day. The moment is treated in this series as an architectural one because it re-founded the politics of the dam: finance passed to Indian bondholders, safeguards passed into domestic law and litigation, and the Bank built its Inspection Panel in the review’s aftermath.

Q: How did NBA shape Indian rehabilitation policy?

The Andolan’s influence runs through three channels. Inside the project, its litigation and agitation extracted rehabilitation packages beyond the tribunal award’s floor, extending entitlements to major sons, encroachers and landless families, and tied every height increase after 2000 to rehabilitation completed in advance under judicial supervision. In administrative practice, its enumeration campaigns and the Morse review’s findings made the counting of the affected, and the sequencing of resettlement before submergence, tests that later projects were expected to pass. In national law, the demands it had pressed since 1985, consent, prior assessment, counted families and rehabilitation as entitlement, were codified for all acquisitions in the land acquisition statute of 2013, which made Social Impact Assessment and rehabilitation statutory rights rather than project benevolence.

Q: What did the Narmada Water Disputes Tribunal decide in its award?

The tribunal, constituted in 1969 under the Inter-State Water Disputes Act after a decade of interstate disagreement, delivered its final award in 1978, notified in December 1979. It fixed the dam’s full reservoir level at 455 feet, 138.68 metres. It divided 28 million acre feet of utilisable flow among Madhya Pradesh with 18.25, Gujarat with 9, Rajasthan with 0.5 and Maharashtra with 0.25 million acre feet, and divided the 1,450 megawatts of power 57 percent to Madhya Pradesh, 27 percent to Maharashtra and 16 percent to Gujarat. For the displaced, it prescribed land-based rehabilitation: a minimum of two hectares of irrigable land per oustee family, house sites and amenities, with resettlement completed a year in advance of submergence at each stage.

Q: What did the Morse Commission find in its 1992 report?

The independent review chaired by Bradford Morse, with Thomas Berger as legal member, was commissioned by World Bank President Barber Conable in 1991 and reported on 18 June 1992. It found that the Sardar Sarovar projects had failed to comply with the Bank’s own resettlement and environmental policies and with the conditions of the 1985 loan agreements; that the basic data on affected people, submergence and environmental impact did not exist in usable form; that resettlement on the scale promised was not being achieved and could not be achieved on the project’s timetable; and that the Bank shared responsibility for the failure. Its central recommendation was that the Bank step back from the projects and consider them afresh, the first such verdict an independent review had delivered on the institution.

Q: Why did the Government of India ask the World Bank to cancel the loan in March 1993?

The formal act was India’s: on 30 March 1993 the government requested cancellation of the undisbursed portion of the 1985 package, about 170 million dollars, and the Bank announced that India wished to proceed with the project on its own. The substance was compulsion by benchmark. After the Morse report advised the Bank to step back, the Bank’s board set performance conditions on resettlement and environment for the following months, and the Indian and state authorities could not or would not certify compliance on that timetable. Cancellation by the borrower ended the loan before the lender could suspend it, preserved the project’s freedom from safeguard supervision, and transferred its financing to domestic sources, including the Sardar Sarovar bonds Gujarat floated from 1993 through the project Nigam.

Q: What was the Jal Samadhi pledge at Manibeli?

Jal Samadhi, immersion in the water, was the pledge adopted in 1991 by the residents of Manibeli, a Bhil village in Maharashtra that stood first in the submergence sequence on that state’s side of the reservoir. Villagers including Bava Mahalia and Khunchand Wasawe vowed that when the reservoir rose they would not vacate their homes, but would remain as the water entered, accepting drowning rather than removal to resettlement colonies they judged worthless. The pledge converted submergence from an administrative act into a witnessed event, and it was kept in the monsoons of 1993 and 1994, when water entered Manibeli and residents stood in their houses as the level climbed, in images that carried the valley’s case to the world.

Q: What did Justice Bharucha’s dissent hold in the 2000 case?

Justice S. P. Bharucha dissented from the majority that permitted construction to continue. On his reading of the record, the environmental clearance granted to the project in 1987 was invalid in law, because it had been issued without the studies the clearance process required: the data on flora and fauna, seismicity, health effects and the true extent of submergence had not been collected, and a clearance granted in ignorance of the facts it existed to assess was no clearance at all. He held that all further construction should stop until a fresh assessment, made on data actually gathered, supported a lawful clearance. The dissent decided nothing in 2000, but it supplied the terms in which the verdict was criticised for two decades and became a standard citation in Indian environmental litigation.

Q: What do the Grievance Redressal Authorities do for Sardar Sarovar oustees?

The Grievance Redressal Authorities are the complaint bodies constituted by each party state under the rehabilitation framework, each headed by a retired judge, which the Supreme Court’s 2000 judgment endorsed as the forum of first resort for oustee grievances. An affected family can carry to its state authority complaints that it has been omitted from the enumeration, denied the land allotment the award prescribes, offered inferior land, or short-paid on compensation and amenities. The authorities investigate, direct the project administration to remedy established defaults, and oversee corrective machinery such as the expert committees that supervised negotiated land purchase for resettlement. The Andolan used the dockets intensively after 2000, filing family-by-family dossiers that kept individual defaults on official record stage by stage as the dam rose.

Q: How many people did the Sardar Sarovar dam displace?

The figures are disputed and are held by named parties. Project documents of the late 1970s counted 6,147 families in the submergence zone. The official count used in the Supreme Court proceedings had risen to about 41,000 project-affected families, roughly 205,000 people, across 245 villages, 193 in Madhya Pradesh, 33 in Maharashtra and 19 in Gujarat, and it excluded those displaced by the canal network. The movement’s counts were higher: its surveys treated the revenue-record enumeration as systematically missing tenants, encroachers, fisherfolk and the landless, and Arundhati Roy’s 1999 essay placed the full number of the affected, counting canals, downstream fisheries and ancillary works, at several hundred thousand. No neutral census ever reconciled the counts, which is itself one of the movement’s central exhibits.

Q: What did Arundhati Roy’s The Greater Common Good argue?

Published in May 1999, Arundhati Roy’s essay carried the Andolan’s case to a mass readership by attacking the arithmetic used to justify the dam. Roy argued that counting beneficiaries in tens of millions against the displaced in tens of thousands concealed who paid and who gained: the costs fell on adivasi and peasant communities with the least protection, while irrigation, power and drinking water flowed to those already connected to the state’s networks. She widened the count of the affected beyond the reservoir to the canal system, the downstream and the sanctuary lands, reaching several hundred thousand people, and she indicted the development model that treated such transfers as progress. The essay made Roy a party to the controversy’s legal life as well: the Supreme Court noticed it in the contempt proceedings surrounding the case.

Q: What was the Gujarat government’s position on the dam and the movement?

Every Gujarat government of the period, across parties, treated Sardar Sarovar as the state’s lifeline project and the Andolan as an obstruction to be defeated. The stated case rested on drought: Saurashtra, Kutch and north Gujarat faced recurring water scarcity, and project documents listed 9,633 villages and 131 urban centres as drinking water beneficiaries, against an official displaced count of about 41,000 families concentrated in other states. Chief Minister Chimanbhai Patel’s government sealed the border against the movement’s Sangharsh Yatra in January 1991; later governments floated bonds to replace the World Bank loan, defended the project in the Supreme Court, and on completion presented the dam as Gujarat’s vindication. Officials also maintained that the movement’s figures were inflated and that its leadership prolonged the valley’s uncertainty for international attention.

Q: What happened at Ferkuva during the Sangharsh Yatra of 1990 to 1991?

The Sangharsh Yatra, the struggle march, set out from Rajghat in Madhya Pradesh on 25 December 1990, a column of valley residents and supporters walking toward the dam site at Kevadia in Gujarat to demand a review of the project. The Gujarat government of Chief Minister Chimanbhai Patel sealed the state border, and the march was halted at Ferkuva, the border village on the Madhya Pradesh side, in the first days of January 1991. For weeks the marchers held their position as negotiations, arrests and releases alternated and national attention fixed on the standoff between one state’s displaced citizens and another state’s police. At the end of January 1991 the Andolan withdrew the column, declaring that the question had been put to the country and would be pursued by other means.

Q: When was the Sardar Sarovar dam completed and dedicated?

The Narmada Control Authority granted the final clearance, to raise the dam to its full height of 138.68 metres with gates installed, on 12 June 2014. The seventeen radial gates were thereafter fitted, and in June 2017 they were closed, allowing the reservoir to rise toward the full level the tribunal had fixed in its award. On 17 September 2017 Prime Minister Narendra Modi dedicated the dam to the nation at Kevadia, fifty-six years after Nehru laid the foundation stone on 5 April 1961. The reservoir first filled to the full reservoir level in September 2019. Completion did not close the rehabilitation account: the Grievance Redressal Authorities continued to hear oustee complaints after the dedication, on the enumeration disputes that had run since 1985.

Q: What did the 2013 Land Acquisition Act change for families losing land?

The Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act, which received assent on 27 September 2013 and entered into force on 1 January 2014, replaced the acquisition law of 1894. It required the consent of 80 percent of affected families for acquisition for private companies and 70 percent for public-private partnership projects. It made a Social Impact Assessment, conducted publicly and appraised independently, a precondition of acquisition, so that affected families were counted before land was taken rather than after. It set compensation by formula to reach up to four times market value in rural areas and twice in urban areas, and it made rehabilitation and resettlement statutory entitlements extending to the landless and to livelihood losers, not only to titleholders.

Q: How were dam height increases controlled after the 2000 verdict?

The 2000 judgment converted the dam’s remaining height into a supervised ladder. The majority permitted construction to 90 metres immediately and thereafter only in stages of five metres, with each stage requiring clearance from the Narmada Control Authority’s environment and rehabilitation sub-groups and the completion of rehabilitation for the families in the corresponding submergence band in advance, in accordance with the tribunal award’s sequencing. The Andolan litigated the ladder stage by stage: clearances to 95 and 100 metres followed the verdict, 110.64 metres came in 2004, and the raise to 121.92 metres in 2006 proceeded only after Patkar’s fast in Delhi prompted a prime-ministerial Oversight Group under V. K. Shunglu to verify rehabilitation in Madhya Pradesh. The final clearance to 138.68 metres was granted on 12 June 2014.