The Indus Waters Treaty of 1960 was negotiated over nearly a decade in a spirit of goodwill, friendship, and post-partition reconstruction, with the World Bank mediating an arrangement that allocated roughly 80 per cent of the Indus system’s waters to Pakistan while India accepted severe design constraints and paid substantial compensation.
Sixty-five years later, following decades of Pakistani objections to Indian hydro projects and a series of cross-border terrorist attacks culminating in the April 2025 Pahalgam incident, India declared that the IWT would be held “in abeyance” until Pakistan credibly ends support for terrorism.
On 31 August 2026, a Court of Arbitration (CoA) at The Hague constituted under the IWT issued an award holding that the treaty “remains fully in force”; that India has no unilateral right to suspend or terminate it; and that none of India’s cited grounds—material breach, fundamental change of circumstances, countermeasures, or necessity—justified “abeyance.” India rejected the award, calling the CoA “illegally constituted” and lacking jurisdiction, and reiterated that the treaty stays in abeyance.
This article examines India’s possible legal position through six analytical pillars:(1) forum-allocation under Article IX and the claim of illegal constitution; (2) the collapse of the goodwill premise underlying the IWT;(3)exceptional doctrines permitting temporary non-performance; (4) the legal effect of non-participation and ex parte proceedings; (5) the limits of competence-de-la-compétence in inter-state disputes; and (6) the distinction between suspension or abeyance and outright termination.
Drawing on treaty text, arbitral awards, and state practice, the article argues that India’s stance, while contested, is not unreasonable and can be framed as consistent with a law-respecting approach to an extraordinary situation, as it rests on plausible interpretations of the IWT’s architecture and recognised doctrines of the law of treaties.
Pakistan’s case: text, continuity, and institutional integrity
Pakistan’s case before the CoA—and in public commentary—rests on familiar, textually grounded propositions.
First, the IWT contains no clause permitting one party to suspend or terminate the treaty. Article XII(4) provides that the treaty “shall continue in force until terminated by a duly ratified treaty.”
Second, even assuming allegations of cross-border terrorism are true, the IWT governs water uses, not security conduct; thus terrorism cannot amount to a “material breach” of the IWT itself.
Third, demographic growth, climate stress, and technological change were foreseeable and do not meet the strict rebus sic stantibus (a legal principle implying that a treaty remains binding only as long as the essential conditions and circumstances existing at the time of its creation remain unchanged) threshold under the Vienna Convention on the Law of Treaties (VCLT).
Fourth, the CoA held that the requirements for lawful countermeasures or necessity were not satisfied, noting that India continued building hydro projects despite claiming security impediments.
Finally, Pakistan argues that unilateral “abeyance” undermines the treaty’s dispute-settlement system and the credibility of international adjudication; States must use agreed mechanisms, not fiat. In this framing, the CoA’s award is undoubtedly a “decisive affirmation of international law”: Pacta Sunt Servanda (agreements must be kept), and water treaties—being lifelines—must withstand political shocks.
Pillar one: forum-allocation under Article IX and the claim of illegal constitution
India’s first pillar rests on the IWT’s dispute architecture, which is deliberately tiered. The treaty channels “questions” to the Permanent Indus Commission, technical “differences” to a Neutral Expert (NE) under Annexure F, and “disputes” to a Court of Arbitration (CoA) under Article IX(3)–(5) and Annexure G. India argues that Pakistan’s challenges to Kishenganga/Ratle design features are technical differences properly referable to the NE, not a full arbitration.
By pursuing—and the World Bank enabling—parallel proceedings (NE and CoA) on substantially overlapping technical questions, the process created a structural conflict the treaty did not contemplate. Even if the World Bank followed the letter of Annexure G’s appointment mechanics for constituting CoA, the substantive mismatch between the nature of the controversy (design/operation differences) and the forum selected (CoA) can be characterised as a breach of the treaty’s functional design. India therefore treats the CoA as illegally constituted for this subject matter and its awards as null and void.
It is pertinent to note that India does not reject all arbitration under the IWT; it only rejects this CoA’s assertion of jurisdiction over technical design questions while an NE process was pending—contrary to Article IX’s allocation of competence. At this stage it must be acknowledged that the CoA rejected India’s competence objections in 2023 and 2025, holding it was validly constituted and remained competent despite India’s abeyance declaration.
India presumably treats these awards as contested, not settled law, and advances its forum-allocation argument as a serious, good-faith interpretation of the treaty’s own dispute-settlement scheme.
Pillar Two: collapse of goodwill premise underlying the IWT
India’s second pillar emphasises the historical and normative context in which the IWT was concluded. The treaty’s preamble and negotiating history emphasise “the most complete and satisfactory utilisation” of the Indus system “in a spirit of goodwill and friendship.” India’s concessions were extraordinary: an asymmetric allocation that gave India roughly 20 per cent of flows (Eastern rivers) while relinquishing claims to the far larger Western system (~80 per cent to Pakistan); a financial concession of ~£62 million (≈$2.5 billion today) to help Pakistan build replacement works; and one-sided design/operational constraints on India’s use of Western rivers, with no equivalent constraints on Pakistan. India’s compliance has been scrupulous for 65 years—through multiple wars and sustained cross-border terrorism (2001 Parliament, 2008 Mumbai, 2016 Uri, 2019 Pulwama, 2025 Pahalgam).
However, today there is a growing perception that Pakistan has used the treaty to obstruct development in India: virtually every major Indian hydro project on Western rivers (Baglihar, Kishenganga, Pakal Dul, Tulbul) has faced prolonged Pakistani objections and referrals, delaying development in Jammu & Kashmir; simultaneously and even in the past, Pakistan has lost no opportunity to portray India as a potential “water aggressor” internationally despite its unbroken compliance. Treaties derive legitimacy not only from text but from good-faith implementation and the basic premises on which consent was given.
Additionally, there is enough evidence emerging from Pakistan itself, be it from the speeches of UN-designated terrorists or interviews by senior functionaries of the establishment, all openly admitting to their past as well as present and continued support for ‘Mujahideen’ in Kashmir. Where one party persistently breaches foundational norms of inter-state conduct (here terrorism) and exploits the treaty to obstruct the other’s development, the cooperative premise of the agreement is undermined. This is not a claim that any tension allows suspension. It is a claim that persistent, state-sponsored terrorism combined with developmental obstruction strikes at the object and purpose of a treaty premised on goodwill and friendship.
Pillar three: exceptional doctrines permitting temporary non-performance or suspension or abeyance of a treaty
India’s third pillar invokes recognised doctrines that excuse temporary suspension or abeyance of a treaty without terminating it. Under the VCLT and the law of state responsibility, suspension is permitted in narrow situations, including material breach (VCLT Art. 60), supervening impossibility of performance (Art. 61), fundamental change of circumstances (Art. 62), and countermeasures/necessity. All of these are temporary, conditional, and reversible by design. They do not end the treaty; they pause obligations until the breach is cured, the peril passes, or circumstances stabilise.
India argues that Pakistan’s support for terrorism, coupled with systematic obstruction of Indian projects permitted under the IWT, defeats the cooperative object of the treaty (material breach); that the security environment and mutual trust underlying the IWT have changed so fundamentally that continued performance under the same terms is untenable (fundamental change); and that temporary non-performance may be justified as a proportionate countermeasure to induce compliance with international obligations, or under necessity to safeguard essential interests.
Notably, the CoA rejected these grounds, holding that terrorism is outside the IWT’s scope and that the strict thresholds were not met. But India’s position remains arguable: the IWT is not an isolated technical compact; it is a bilateral confidence-building measure whose viability depends on a minimum baseline of security and good faith.
The ICJ’s Gabčíkovo-Nagymaros precedent sets a high bar for these doctrines, and that is precisely why India presents “abeyance” as a provisional, conditional posture—not a termination—pending restoration of a security environment compatible with the treaty’s cooperative premise.
India’s declaration is thus carefully framed: the treaty is held “in abeyance” until Pakistan credibly and irrevocably ends support for cross-border terrorism; it is tied to a specific, verifiable change in Pakistan’s conduct; and it is not a termination, as India continues to engage (or has engaged) through the NE track and diplomatic channels, showing it has not abandoned the IWT’s entire framework—only the branch it deems misapplied. This distinguishes India’s stance from unilateralism for convenience. It is a security-linked, reversible suspension aimed at inducing compliance with foundational norms of inter-state behaviour.
Pillar four: non-participation and the legal effect of ex parte proceedings
A separate but critical question is the legal force of an international arbitral award rendered when one party refuses to participate. India has consistently stated that it “never recognised the legal existence” of the CoA, did not take part in its proceedings, and therefore is not bound by its decisions. Pakistan, for its part, proceeded alone—and even paid India’s share of arbitration costs—to keep the case moving.
Under most major arbitration rules—including those commonly used in inter-state disputes—a tribunal may continue proceedings and render an award even if one party does not appear, provided the absent party was given proper notice and an opportunity to present its case. This is sometimes described as “default” or “ex parte” proceedings, but it is not the same as a domestic-court default judgment. Non-participation does not stop the case; arbitration rules and the UNCITRAL Model Law expressly allow tribunals to proceed where a party, duly notified, fails to appear or submit evidence.
Unlike in many national courts, the tribunal cannot treat non-appearance as an admission of liability. It must still independently assess whether the claims are grounded in fact and law on the evidence before it. Best practice—and often a legal requirement—is that the tribunal document all steps taken to notify the absent party, copy it on all communications, and afford equal time to respond at each stage.
In the IWT arbitration, the CoA recorded that India did not participate in the constitution of the court, did not attend meetings, and did not advance funds toward costs; Pakistan bore both shares. The tribunal nonetheless proceeded, invited India to participate, and issued its award on the evidence and submissions presented by Pakistan. An award rendered in a party’s absence can be binding and enforceable if the tribunal had valid jurisdiction under the parties’ arbitration agreement (here, the IWT’s Annexure G framework) and the absent party received proper notice and a fair opportunity to be heard.
However, the absent party may still challenge the award on limited grounds, most notably lack of jurisdiction, procedural unfairness, or public policy/fundamental defects. Crucially, non-participation does not waive jurisdictional objections. A state that believes a tribunal is illegally constituted may refuse to appear and later argue—before enforcing courts or in diplomatic fora—that the award has no legal effect because the tribunal lacked authority from the outset.
India’s stance fits within this framework. India says it was not merely absent; it never accepted the tribunal’s existence because, in its view, the CoA was constituted in breach of the IWT’s own dispute-settlement design. If a tribunal is ultra vires (beyond its powers) under the treaty, its award—whether contested or default—may be treated as null and void by the objecting state.
Even if the CoA proceeded correctly under its own rules, India’s non-participation means the award’s practical enforceability depends on political and diplomatic pressure, not on a global enforcement mechanism akin to domestic judgments. International tribunals can issue awards in a party’s absence, and such awards can be valid if jurisdiction and procedural fairness are satisfied. But a state that contests jurisdiction from the outset—on the ground that the tribunal was illegally constituted—can plausibly maintain that the award has no binding force on it, even if the tribunal proceeded ex parte.
Pillar five: competence-de-la-compétence and its limits in inter-state disputes
That leads to the next question: what if the tribunal itself has ruled that it is validly constituted? The CoA has done exactly that, issuing an Award on Competence in 2023 and reaffirming it in 2025, holding that it was properly constituted and remained competent despite India’s “abeyance” declaration. Does that self-validation settle the matter? International law recognises the doctrine of competence-de-la-compétence (or Kompetenz-Kompetenz): an arbitral tribunal has the authority to determine whether it has jurisdiction over a dispute. This is built into most arbitration rules and treaty frameworks, including the IWT’s Annexure G scheme.
The tribunal can hear jurisdictional objections (like India’s forum-allocation argument) and issue an Award on Competence. If it rejects those objections, it can proceed to the merits and issue a final award. In many commercial and investment arbitrations, national courts will give considerable deference to the tribunal’s own jurisdictional findings when asked to enforce the award.
In the IWT case, the CoA exercised this power: it considered India’s objections, ruled it was validly constituted, and continued. But a tribunal’s self-validation is not the last word in inter-state disputes. In commercial/investment cases, enforcement often runs through national courts under the New York Convention or ICSID regimes, which tend to respect the tribunal’s competence rulings unless there is a clear defect. In inter-state disputes, there is no global sheriff. Enforcement depends on diplomatic pressure, reciprocal compliance, and political cost, not on a centralised enforcement court.
That means a tribunal’s own ruling that it is competent carries weight, but it does not conclusively settle the jurisdictional question for the absent state. The absent state can still maintain, in good faith, that the tribunal was ultra vires (beyond its powers) under the treaty, and treat the award as non-binding on that basis.
The dispute then becomes a genuine legal disagreement about the treaty’s interpretation—precisely the situation we see with India and the CoA. India’s position is not “we ignore the tribunal because we dislike the result.” It is: the CoA misread the IWT’s own architecture, and its self-validation does not cure that defect. Specifically, India says the IWT allocates technical “differences” to the Neutral Expert, not the CoA.
If that is correct, then the CoA’s assertion of jurisdiction over design questions was contrary to Article IX, regardless of what the CoA said about its own competence. India contends that allowing NE and CoA to run in parallel on overlapping issues created a structural conflict the treaty did not contemplate.
The CoA’s self-validation does not resolve that structural concern. Because this is a bilateral treaty between sovereigns, not a commercial contract, India can plausibly argue that its own interpretation of the treaty’s dispute-settlement scheme remains open for diplomatic and legal contestation, even after the CoA’s competence ruling. The CoA’s ruling on its own jurisdiction is legally significant, but not dispositive in an inter-state dispute where the other party rejects the tribunal’s constitution ab initio.
The practical upshot is that the CoA’s competence rulings and final award provide a strong legal and diplomatic platform to argue that India is in breach of its IWT obligations. But the CoA’s self-validation does not foreclose India’s jurisdictional objection; it simply means the two sides now have competing, good-faith interpretations of the same treaty text. For third parties, the situation is a classic inter-state legal dispute—not a clear-cut case of lawlessness. India can argue it is upholding the treaty’s own design by rejecting a tribunal it views as misallocated, while Pakistan can argue it is upholding textual continuity and institutional integrity.
Pillar six: repudiation and termination—could India walk away?
So far, the analysis has focused on suspension (“abeyance”)—a temporary pause in performance. But what if India went further and sought to terminate or repudiate the treaty altogether? International law treats this more strictly, but it is not impossible. Article XII(4) of the IWT provides: “This Treaty shall continue in force until terminated by a duly ratified treaty concluded for the purpose between the two Governments.”
This is a mutual-consent termination clause. On its face, it means neither party can unilaterally terminate the treaty by notice; termination requires a subsequent, duly ratified treaty agreed by both India and Pakistan. The CoA leaned heavily on this text, holding that the IWT “does not provide either party with the right to unilaterally suspend or terminate its operation.” We have already noted that suspension or abeyance is not something which is impermissible as per the VCLT.
Even if a treaty lacks a unilateral exit clause or explicitly requires mutual consent Art. 56(1) allows withdrawal if it is established that the parties intended to admit the possibility of withdrawal, or a right of withdrawal may be implied by the nature of the treaty. Article 56(2) requires 12 months’ notice. Some states have withdrawn from treaties without express exit clauses, arguing implied rights or changed intentions. But this is contested and often criticised.
The VCLT and customary international law recognise limited grounds on which a party may terminate or withdraw: material breach (VCLT Art. 60), fundamental change of circumstances (Art. 62), impossibility of performance (Art. 61), emergence of a new jus cogens norm (Art. 64), and denunciation/withdrawal under Art. 56 where the parties intended to admit it or the nature of the treaty implies such a right. These are exceptional, narrowly construed grounds, but they exist precisely because international law recognises that some changes or breaches can make continued performance untenable.
Article 60(1) of the VCLT provides that a material breach of a bilateral treaty by one party entitles the other to terminate or suspend the treaty. Article 60(3) defines material breach as a repudiation of the treaty not sanctioned by the VCLT, or violation of a provision essential to the accomplishment of the treaty’s object and purpose. In the Ethiopia–Eritrea context, scholars have argued that Eritrea’s persistent breaches of the Algiers Peace Agreement could entitle Ethiopia to invoke termination under Art. 60. Similarly, in Gabčíkovo-Nagymaros, the ICJ confirmed that Arts. 60–62 reflect customary law and can apply even to older treaties.
India could argue that Pakistan’s systematic obstruction of permitted Indian projects, combined with state-sponsored terrorism that undermines the cooperative premise, amounts to a material breach of the treaty’s object and purpose. The CoA rejected this, holding terrorism outside the IWT’s scope and the thresholds unmet. But as a legal theory, material breach as a ground for termination is well-established. Article 62 permits termination or withdrawal where the circumstances were an essential basis of the parties’ consent, the change was not foreseen, and the change radically transforms the extent of obligations still to be performed.
The Gabčíkovo-Nagymaros Treaty itself had no unilateral exit clause; Hungary’s suspension and attempted exit on grounds of necessity and fundamental change illustrates how states invoke VCLT-type doctrines to justify abeyance or termination even of ‘permanent’ treaties. Although the ICJ did not uphold Hungary’s contention, the judgment confirmed that grounds for temporary suspension or abeyance do exist in international law, even if they are narrowly construed. In the Gabčíkovo-Nagymaros case political tensions, demographic shifts, and even environmental changes were held insufficient by the ICJ to invoke termination and the CoA applied the same strict standard to India’s arguments (terrorism, climate, demographics), finding them foreseeable and not radical enough.
Still, in theory, if India could show that the security environment and mutual trust underlying the IWT had changed so fundamentally that continued performance was untenable, it could invoke Art. 62 as a ground for termination—not just suspension. Given Art. XII(4)’s explicit mutual-consent requirement, India’s decision to hold the IWT “in abeyance” rather than terminate it outright is strategically and legally significant. Suspension requires a less demanding showing than termination; it is easier to argue that temporary non-performance is justified than that the treaty is dead. Suspension preserves the option to resume full compliance if security conditions improve, without the legal finality of termination. By keeping the treaty in a state of suspended animation, India retains negotiating leverage—it can offer to restore full operation in exchange for concrete changes in Pakistan’s conduct.
State practice: non-compliance with international awards
State practice confirms that unilateral termination or withdrawal from treaties, while exceptional, does occur.
The United States withdrew from the 1972 Anti-Ballistic Missile Treaty in 2002 invoking its “supreme national interests” withdrawal clause; Russia suspended and later moved to terminate its obligations under the Conventional Armed Forces in Europe Treaty citing fundamental changes in the security environment and other parties’ non-compliance; and North Korea invoked the Nuclear Non-Proliferation Treaty’s withdrawal clause in light of heightened security threats. These instances illustrate that states do walk away from major treaties, often framing their action in terms of security changes, breach, or supreme national interest, even if they rely on express withdrawal clauses rather than VCLT Arts. 60–62 directly.
International law has no global enforcement mechanism comparable to domestic courts. Compliance often depends on political will, diplomatic pressure, and reputational costs. As a result, there is a modest but significant body of cases where states have refused to accept or implement adverse decisions. In the 2016 South China Sea Arbitration (Philippines v. China) case, China, a major power and Pakistan’s closest ally has refused to recognise the tribunal’s authority, contesting jurisdiction, and treating the award as non-binding despite the tribunal’s own competence ruling.
In Nicaragua v. United States (ICJ, 1986), the U.S. participated in the jurisdictional phase but withdrew from the merits phase after the ICJ upheld jurisdiction; the ICJ ruled against the U.S. on the merits and ordered reparations; the U.S. did not comply and blocked enforcement in the UN Security Council. In Russia and the Arctic Sunrise Case (ITLOS/UNCLOS, 2013–2015), Russia declined to participate, contesting jurisdiction; ITLOS and the subsequent UNCLOS tribunal proceeded; Russia did not comply with the initial ITLOS provisional measures order. In Pakistan and the Jadhav Case (ICJ, 2019), the ICJ ordered Pakistan to provide “effective review and reconsideration” of Jadhav’s conviction; Pakistan’s compliance has been partial and contested; India has repeatedly argued that Pakistan has not fully implemented the ICJ’s order.
These cases support several points: non-compliance is not unique to India; major powers (China, U.S., Russia) have all, at times, refused to accept or implement adverse international decisions when they contested jurisdiction or deemed the ruling contrary to core interests.
Jurisdictional objections matter; in each case, the refusing state advanced a legal rationale (lack of consent, ultra vires tribunal, political question), not merely a political refusal. Enforcement is political, not judicial; even when a tribunal rules it is competent and issues a binding award, practical enforcement depends on diplomatic pressure, reputational costs, and the balance of power—not a global sheriff. Pakistan is not a stranger to this dynamic; in Jadhav, Pakistan’s compliance with the ICJ’s order has been contested by India, showing that Pakistan itself has faced accusations of incomplete compliance with an international court’s directive.
Also read: Inside the second partition of India-Pakistan—the full story of the Indus Waters Treaty
Conclusion
The CoA’s award is a powerful statement of textual continuity and pacta sunt servanda. But India’s position rests on plausible treaty-interpretation arguments (forum allocation under Article IX), historical context (the IWT’s goodwill premise and India’s asymmetric concessions), and recognised exceptional doctrines (material breach, fundamental change, countermeasures/necessity).
India is not claiming a sovereign right to ignore treaties at will. It is asserting that where the cooperative premise of a bilateral treaty collapses under sustained bad faith and security threats, a temporary, conditional suspension of performance—coupled with rejection of a tribunal viewed as defectively constituted—is a defensible, law-respecting response to an extraordinary situation. That position will remain contested. But it is not unreasonable—and it can be advanced without denying the CoA’s reasoning, simply by treating it as one side of a genuine legal dispute about forum, premise, and exception in the law of treaties.
Dr Amit Krishankant Paul is an independent researcher and author of the book ‘Meghdoot: The Beginning of the Coldest War’. Views are personal.
(Edited by Ratan Priya)
Whatever may be the legal points, the important thing is that IWT is the only treaty in the world which divides rivers and not divide the water volume. That is simply impractical. Why did Nehru sign it? It is the liberal mentality of Mahatma Gandhi who said “whatever Pakistan demands, give it”. Nehru and the Congress are following that policy till today even if Pakistan attacked India four times resulting in war and when it realized it cannot win by conventional war, it started proxy war. This is as clear as day. But Congress is still pursuing the policy of appeasement. India must keep IWT in abeyance until Pakistan stops cross-border terrorism with tangible results. When Pakistan comes to negotiation table, the IWT must be modified based on water volume percentages.