Narmada Bachao Andolan v. Union of India: Sardar Sarovar, judicial restraint and the pari passu R&R framework

The petition filed by the Narmada Bachao Andolan in 1994 brought before the Supreme Court a constellation of questions that environmental jurisprudence had not previously had to answer at the same time. The Sardar Sarovar Dam — at the centre of the Narmada Valley Development Project — was the kind of inter-state, decadal, central-government-sponsored infrastructure project that engaged the executive's discretion at the highest level. The displaced populations — drawn predominantly from Scheduled Tribes and other forest-dependent communities in Gujarat, Madhya Pradesh and Maharashtra — engaged the constitutional architecture of Article 21. The environmental consequences engaged Articles 48A and 51A(g), the Environment (Protection) Act, 1986, and the Forest (Conservation) Act, 1980. The inter-state water-sharing dimension engaged the Narmada Water Disputes Tribunal Award, 1979 under the Inter-State River Water Disputes Act, 1956.

The three-judge Bench was constituted by Chief Justice Dr A.S. Anand, Justice S.P. Bharucha and Justice B.N. Kirpal. The judgment of the Bench was delivered on 18 October 2000 and is reported at (2000) 10 SCC 664 / AIR 2000 SC 3751, in W.P.(C) 319/1994. Justice Kirpal wrote the majority judgment for himself and the Chief Justice. Justice Bharucha dissented.

The doctrinal architecture the majority articulated — judicial restraint in major-project PILs, the binding finality of inter-state Water Disputes Tribunal awards, sustainable development as a balancing exercise, and the stage-wise pari passu R&R framework — has been the foundational architecture of the Indian jurisprudence on large infrastructure projects ever since. The Bharucha dissent has been the foundational counter-architecture.

The architectural premise: the NWDT Award and the project

To understand what the 2000 judgment did, the framework it engaged needs to be sketched.

The Narmada Water Disputes Tribunal was constituted under the Inter-State River Water Disputes Act, 1956 to resolve the dispute on the sharing of Narmada waters among the riparian States — Madhya Pradesh, Gujarat, Maharashtra and Rajasthan. The Tribunal's Award, finalised on 7 December 1979 and notified by the Union, allocated waters and provided the framework for the Sardar Sarovar Dam in Gujarat (the terminal dam in the project), together with upstream dams in Madhya Pradesh. The Award contained provisions on Relief and Rehabilitation (R&R) for displaced families, including land-for-land entitlements, encadrement of the Narmada Control Authority, and the institutional architecture for project supervision.

Construction of the Sardar Sarovar Dam began in the 1980s. Environmental clearance under the Environment (Protection) Act, 1986 was granted in 1987, subject to conditions on environmental safeguards. Forest clearance under the Forest (Conservation) Act, 1980 was granted in respect of the submergence area. By the mid-1990s, when the writ petition was filed, the dam had reached a height of approximately 80.3 metres. The proposed final height of 138.68 metres would, on the petitioner's case, submerge a substantial area in Madhya Pradesh and Maharashtra and displace tens of thousands of families — predominantly from Scheduled Tribes.

The petitioner contended that R&R had not been carried out in compliance with the NWDT Award; that displaced families had not received land-for-land in accordance with the Award's terms; that environmental safeguards had not been complied with; and that the project should be halted until compliance was effected. The writ sought, in the alternative, a substantial reduction of the dam height to reduce the submergence area.

The holding

The majority's disposition rested on the architectural premise that the project — having been the subject of the NWDT Award, environmental and forest clearances, and decades of executive consideration — could not be halted in writ. The continuing supervision through the R&R Sub-Group certification at each 5-metre slab was, on the majority's reading, the appropriate constitutional remedy for the petitioner's R&R concerns.

Justice Bharucha, dissenting, held that the project should not proceed beyond the height already attained until full Environmental Impact Assessment was completed and complete R&R safeguards were in place. The dissent's principal proposition was that the majority's stage-wise framework — though architecturally elegant — would in practice prove inadequate, because the certification discipline would not be capable of preventing the displacement of families before R&R was actually effected.

The reasoning

The majority's doctrinal architecture has four connected threads.

The narrow scope of writ review

The first thread is the scope of writ review in major-project PILs. The majority held that where decisions involve technical, economic and policy choices on inter-state projects approved at the highest level of executive deliberation — and where the constitutional architecture has provided a specialised forum (the Inter-State Water Disputes Tribunal) for the resolution of the underlying dispute — judicial review under Article 32 is narrowly confined. The Court does not sit as a super-executive to second-guess the technical assessments of expert bodies, the economic assessments of the planning architecture, or the policy choices of inter-state cooperation.

The doctrinal proposition has had substantial subsequent application. The architecture of judicial restraint in policy-domain matters has been cited in cases concerning telecommunications policy, mineral allocation, infrastructure projects and economic regulation. The doctrinal proposition is that the constitutional architecture of separation of powers requires the Court to defer to the executive in matters where technical and policy expertise is the appropriate analytical posture.

The binding finality of the NWDT Award

The second thread is the finality of the NWDT Award. The majority held that the Award — having been finalised under the Inter-State River Water Disputes Act, 1956 and notified by the Union — was binding on the Court. The Award could not be reopened through a writ petition. The architectural premise of the 1956 Act — that inter-state water disputes are resolved by specialised tribunals whose awards are binding once notified — would be defeated if writ jurisdiction could be invoked to reopen the underlying allocation.

The doctrinal proposition has had application in subsequent inter-state water litigation. The architecture of binding finality has been cited in Cauvery, Krishna, Mahanadi and other inter-state water disputes — the proposition that the Court will not reopen Tribunal awards in writ has been the architectural premise.

Sustainable development as balancing

The third thread is sustainable development. The majority engaged the precautionary principle and the polluter-pays principle articulated in Vellore Citizens' Welfare Forum v. Union of India (1996), but read sustainable development as a balancing exercise — the constitutional architecture requires the costs of a major project (displacement, environmental impact, forest loss) to be weighed against its benefits (irrigation, drinking water, hydroelectric power, food security). Where the executive has performed that balancing through the institutional architecture of project approval, the Court should not displace the executive's balance.

The doctrinal proposition has been a recurring point of doctrinal disagreement in subsequent environmental jurisprudence. The architecture of balancing — and the question of how the Court should engage the precautionary principle when it points against a project that the executive has approved — sits at the centre of cases such as Lafarge Umiam Mining v. Union of India (2011), Hanuman Laxman Aroskar v. Union of India (2019), and the Vanashakti v. Union of India saga (May 2025 / November 2025 review).

The pari passu R&R framework

The fourth thread is the operational architecture. The R&R Sub-Group certification at each 5-metre slab is the operative mechanism the majority devised. The architectural premise is that displacement and construction must proceed pari passu — that each increment in submergence must be preceded by the rehabilitation of the families it displaces. The certification discipline operationalises the architectural premise — the dam cannot be raised until R&R for the corresponding submergence area is certified.

The pari passu framework has been the doctrinal template for subsequent major-project R&R litigation. N.D. Jayal v. Union of India (Tehri Dam) (2004) applied the framework to the Tehri Dam in Uttarakhand. The architecture of stage-wise displacement-conditional construction has been the principal R&R remedy in the years that followed.

The Bharucha dissent

Justice Bharucha's dissent is one of the most cited dissents in Indian environmental jurisprudence. Its principal proposition was that the project should not be permitted to proceed beyond the height already attained until full Environmental Impact Assessment was completed and complete R&R safeguards were in place.

The dissent operates on three connected propositions.

The first was that the EIA conducted before the 1987 environmental clearance was incomplete. The conditions on which the 1987 clearance had been granted — pari passu environmental safeguards, catchment area treatment, command area development, compensatory afforestation — had not been complied with. The architectural premise of conditional clearance was that the project would proceed only if and to the extent that the safeguards were effected. The factual record before the Court did not, on the dissent's reading, establish that the safeguards had been effected.

The second was that the R&R Sub-Group certification framework — though architecturally elegant — was operationally inadequate. The certification discipline would be exercised by the same institutional architecture that had failed to ensure pari passu R&R in the years leading up to the writ. The discipline would not be capable of preventing the displacement of families before R&R was actually effected. The empirical record of R&R since 1987 was, on the dissent's reading, the evidence that the certification framework would not work.

The third was that the constitutional architecture of Articles 21, 48A and 51A(g) — together with the precautionary principle and the polluter-pays principle articulated in Vellore — required a more rigorous engagement than the policy-deference framework the majority had adopted. The constitutional commitment to environment and to the protection of vulnerable populations was not, on the dissent's reading, capable of being honoured by the architecture of deference.

The dissent's subsequent treatment — as a moral compass for environmental PILs — reflects the institutional importance of the disagreement. The architecture of policy-deference vs constitutional rigour that the Bharucha dissent identified has been the recurring doctrinal frame of the cases that followed.

The project should not proceed beyond the height already attained until full environmental impact assessment is completed and complete R&R safeguards are in place. The architecture of certification at each 5-metre slab does not, in my view, provide an adequate constitutional remedy for the displacement and environmental concerns that this petition has raised.

What the judgment did not address

It is worth being precise about the boundary of the 18 October 2000 judgment.

The judgment did not engage a national-level Rehabilitation Policy. The central architecture of an R&R Policy emerged later, in 2003, and was developed further through subsequent policy and statutory instruments. At the time of the 2000 judgment, the R&R framework operated through the NWDT Award's R&R clauses and through state-level R&R packages (Gujarat, Madhya Pradesh, Maharashtra, Rajasthan). The architecture the majority worked with was the NWDT Award framework — not a generic central policy.

The judgment did not address the architecture of the Land Acquisition Act, 1894 in relation to displacement compensation. The compensation-and-resettlement architecture for displaced families operated through the NWDT Award and through state-level R&R packages — not through the 1894 Act's market-value mechanism. The architectural relationship between the dam-displacement framework and the 1894 Act has been developed through subsequent jurisprudence on the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act, 2013.

The judgment did not formulate a substantive doctrine on EIA rigour. The architecture of EIA rigour — and the proposition that EIA cannot be a "tick-box" exercise — was developed in Hanuman Laxman Aroskar v. Union of India (Mopa Airport) in 2019 by a Bench of Justice Dr D.Y. Chandrachud and Justice Hemant Gupta. The doctrinal architecture of cumulative impact assessment, duty of candour on project proponents, and EC suspension as a remedial template was an Aroskar development — building on, and departing from, the Narmada-era policy-deference frame.

The judgment did not engage with the architecture of the Forest Rights Act, 2006 — which was enacted six years later — and did not formulate any doctrine on the rights of Scheduled Tribes whose lands were submerged. That architecture has been developed through the Orissa Mining Corporation v. Ministry of Environment and Forests (Niyamgiri) line and through subsequent FRA jurisprudence.

How the doctrine operates in practice

For practitioners advising in infrastructure, mining, hydroelectric and large-project matters, the Narmada framework has the following operational dimensions.

The architecture of policy-deference

The first operational dimension is the analytical posture of the writ court. The framework signals that in major-project PILs — especially where the executive has performed the relevant balancing through institutional architecture (inter-state agreements, expert bodies, project clearances) — the writ court will engage on procedural and rights-based grounds, not on the merits of the executive's substantive choice. The doctrinal architecture is best engaged through specific challenges to procedural defects, to non-compliance with clearance conditions, and to displacement injustices — not through a wholesale challenge to the project itself.

Stage-wise construction-conditional R&R

The second dimension is the operational template for R&R-conditional infrastructure. The pari passu framework has been applied — with variations — to subsequent major-project litigation. The architectural premise is that construction proceeds in slabs, with R&R certification at each slab; the certification discipline operates as the protective mechanism for displaced families.

The continuing institutional supervision

The third dimension is the institutional architecture of continuing supervision. The R&R Sub-Group of the Narmada Control Authority continues to operate. Successive interlocutory applications since 2000 have engaged compliance with the pari passu framework. The architecture of continuing supervision — through the R&R Sub-Group and through periodic court engagement — has been a persistent feature of the post-2000 litigation, with ongoing R&R compliance proceedings continuing through 2022-24.

The doctrinal arc

The Narmada judgment sits in a substantial line of constitutional engagement with environmental and infrastructure questions. Three lines of subsequent doctrinal development deserve flagging.

The first is the continuing tension between policy-deference and environmental rigour. The Narmada-era policy-deference frame has been challenged by subsequent lines of authority that have engaged the precautionary principle and EIA rigour more searchingly. Hanuman Laxman Aroskar v. Union of India (2019) — the Mopa Airport case — articulated a duty of candour on project proponents and held that the EIA process must have rigour and scientific basis. The Vanashakti v. Union of India saga (May 2025 judgment / November 2025 recall) continues the tension — the May 2025 judgment had taken a position closer to the Bharucha dissent's frame; the November 2025 recall has taken a position closer to the Narmada majority's deference frame.

The second is the subsequent development of the R&R architecture. The pari passu framework has been operationalised through state-level R&R packages, through the Right to Fair Compensation and Transparency in Land Acquisition, Rehabilitation and Resettlement Act, 2013, and through project-specific R&R frameworks. The architecture of R&R-conditional construction has been the foundational template.

The third is the continuing-mandamus model. Like T.N. Godavarman Thirumulpad v. Union of India (1996), the Narmada writ has operated as a continuing engagement — with successive interlocutory applications and orders responding to specific situations. The institutional architecture of continuing supervision has been one of the defining features of Indian environmental and infrastructure jurisprudence.

Related editorial pieces

• T.N. Godavarman Thirumulpad v. Union of India: the dictionary meaning of 'forest' and the continuing mandamus
• Hanuman Laxman Aroskar v. Union of India: EIA rigour and the Mopa Airport
• Vanashakti v. Union of India and the November 2025 review
• Vellore Citizens' Welfare Forum v. Union of India: precautionary principle and polluter pays

---

Verify against the reported judgment. The pari passu R&R framework has been operationalised through hundreds of interlocutory applications since 2000; the broader doctrinal frame of policy-deference vs environmental rigour continues to be engaged in contemporary cases.