THE AUTHOR:
Yash Sinha, Advocate at the Supreme Court of India
The Indian Supreme Court in Gayatri Balasamy v. ISG Novasoft (“Gayatri”) has clarified the courts’ power to modify awards under Sections 34 and 37. In what it terms as ‘limited modification,’ it has found the power to change post-award interest, in addition to correcting clerical, typographical, and other manifest errors. Apart from these instances, the Supreme Court may invoke its Constitutional power under Article 142(1) when the award seemingly falls short of rendering complete justice.
This piece intends to argue that in doing so, the silences in the Arbitration and Conciliation Act, 1996 (“the Act”) have been wrongly read to indicate ‘permissions’, and not ‘prohibitions.’ Further, the decision inserts unpredictability into a system designed for party-driven control.
Premises of the Decision
Gayatri states that silence in the Act amounts to permission for many implied powers of the court. At the outset, it is reasoned that Section 34 is clear in bestowing courts with the power to annul an award. Since modification shares its nature of review but is less drastic in outcome, the court assumed this power lay within annulment-power. To support its reasoning, the court cites the Act’s intent: modification by courts will prevent further arbitration and conclude the dispute with finality and swiftness.
It also states that Section 34 does not bar a court from fashioning a relief while exercising its jurisdiction. Given the primary prohibition Section 34 contains, which is a bar on a merits-based review, the courts may alter an award as long as they do not review it on merits. If the provision wanted to bar modification, the text would have specified it.
It extends this reasoning to Section 33, finding generality in its scope, though its text reserves error-correction solely for the tribunal. The court views Section 33 as an opportunity, not an exclusive power, for the tribunal to make corrections. In case the tribunal fails to do so and an error continues to exist, a court may step in and make those corrections. If not interpreted this way, it states, the error will never cease to exist. This view rests on the absence of any tangible prohibition on the court from doing so. Notably, Gayatri uses the same reason to dispel the argument that modification of an award is solely reserved for the tribunal under Section 34(4).
On its powers to modify post-award interest, the court relies on the phrasing of Section 31(7)(b) of the Act. The court reaches this conclusion by reading two elements in the Act cumulatively:
- There exists a provision for the scenario where a tribunal failed to specify this interest; and
- This is followed by the texts of Sections 34/37, where an award may be challenged.
For Gayatri,these meant that the Act wants courts to scrutinise post-award interest in case the tribunal disregarded it.
Flawed Interpretation
In holding silences as indicators of the court’s implied powers, the court denied the line of reasoning endorsed by McDermott International Inc. v. Burn Standard Co. Ltd. This precedent essentially suggested that the lack of power to modify and the intent to prevent judicial interference are indicative: courts cannot modify awards. Justice K.V. Vishwanathan’s dissent sides with this reasoning.
However, even this logic only recognises one facet of the Act: spelled-out implications create positive zones within which arbitrators and the judiciary can operate. Both majority and minority views then disagree on what the limits of these positive zones are. However, this neglects the law’s other design element: silences are meant to be negative zones where courts and arbitrators are barred from venturing into. Even otherwise, the Act’s overt intent bars court-led modification to preserve arbitral autonomy.
The ‘Negative-Import’ Approach
The Act is a complete and exhaustive code in itself. This is most clearly enunciated by Fuerst Day Lawson v. Jindal Exports Ltd (“Fuerst”). After studying the Act’s nuanced, formulaic tone, it concluded the law to be self-contained. For Fuerst, this translated into two implications: a legal implication not mentioned in the Act must be read as impermissible, and only those mentioned by the text as allowed. Gayatri contradicts Fuerst by not viewing the Act’s silence on court-led award modification as prohibitive.
At the outset, the Act’s text only allows partial annulment or complete annulment. There is no provision for partial or complete modification of awards by the courts. Despite this absence, even if the court wanted to permit modification at the stages of Sections 34 and 37 or beyond, the law’s text suggests only one course of action: remand to the tribunal under Section 34(4). Unlike Justice Vishwanathan, who claimed that if a remand was not used at the stage of Section 34, it would remain shut forever, the Supreme Court itself has suggested and done otherwise. The powers under Section 34 extend equally to a court of appeal, and is available even if the dispute reaches the Supreme Court.
This structure, which expressly allows annulment by courts and modification by arbitrators, has to be read as a ‘bar’ on any contrary legal implication. The court in Associate Builders v. DDA and Delhi Airport Metro Express Pvt. Ltd. v. Delhi Metro Rail Corporation Ltd. (“Delhi Metro”)echoes this when it notes the intent of the various amendments made to Section 34. Namely, the legislature intended to cut down the judiciary’s scope to read new features into its text. The contentious phrase ‘public policy’ in Section 34 is now a set of distilled, specified elements, for this very reason.
The last reason given by the court is that ‘implied powers always exist’ to make legal outcomes more effective. However, this disregards another bar created by the Act’s nature as a self-contained code, as acknowledged by the Supreme Court: any legal event covered by the Act must necessarily follow the procedure under the Act, as opposed to the general legal process. The implied power of any court is drawn from the rules which govern its procedure or Section 151, Code of Civil Procedure, both falling under ‘general legal processes.’
Hence, the power to modify does not exist, and reading it in the Act imports meanings despite prohibitions.
Respecting Arbitral Autonomy
Legislative intent is key in cases of interpretive-difficulties. However, Gayatri’s reliance on ‘speedy resolution of disputes’ as the prime intent of the Act is misplaced.It disregards the law’s larger push to the notion of ‘judicial deference to arbitral autonomy.’
The dispute over the Act’s primary intent has been resolved in Central Organisation for Railway Electrification v. ECI-SPIC-SMO-MCML (JV). It has categorically held that efficiency is not an independent goal in arbitration. Instead, it is subordinate to procedural legitimacy, which in turn prioritises party-autonomy. Honouring this autonomy arguably requires acknowledging that the parties opted out of litigation precisely to avoid judicial adjudication.
In this light, it will be seen that the question formulated in Gayatri was misconceived. It framed the question as ‘whether courts can correct errors in an award?’ However, the correct formulation would have been: ‘Who has the authority to decide which errors are worth correcting?’ This framing recognises that the judiciary may lack any powers of modification due to the exclusivity given to the tribunal in this regard. Phrasing an issue to scrutinise the judiciary’s power in a silo was erroneous, as it failed to respect the importance party-autonomy assigns to the arbitrators.
Regardless of its approach, Gayatri could not have disregarded precedents affirming tribunal’s exclusive control on awards. Decisions such as in Delhi Metro are clear in holding that the tribunal is uniquely positioned to weigh the internal logic and trade-offs within the award. The underlying rationale was to honour the parties’ intent to settle disputes via mutually selected experts. Sticking to the arbitrator’s phraseology and numbers is, then, more aligned with the parties’ contractual intent. Judicial modification, even when limited to seemingly technical aspects like interest rates or computational errors, violates this jurisdictional boundary by reallocating decision-making power to an external body (the court) that lacks access to the full context or commercial rationale underpinning the award.
This deference to arbitral autonomy has become more pronounced following the amendments to Section 34. Judicial decisions in Associate Builders and Delhi Metro have established that when the arbitrator has adhered to the contract and evidence, there exists no reason to review the award. Respect for who decides now extends to how they decide, completing the deference to parties’ choice. Logically, if review is not justified in such a case, modification should not follow either. Gayatri defies this approach, permitting changes even when arbitrators follow contract and evidence.
Lastly, arbitral autonomy is an offshoot of the Act’s larger intent, which was to empower parties to control the process. Any last-minute tweak by courts to an award introduces three variables entirely outside their control:
- Courts will decide whether the ‘errors’ relate to merits or not;
- Supreme Court may determine whether ‘complete’ justice is rendered or not; and
- After the Supreme Court addresses any of these concerns, the only remedy for the parties is judicial: namely, the Supreme Court’s review and curative jurisdiction.
Conclusion
By reading the unspoken as permissive, Gayatri has given rise to several apprehensions. Moving forward, the judiciary may import new meanings into the text of the Act in areas where it is silent. Added to that is the increased uncertainty in the outcome of arbitration: quanta of award-amount, interest rates and other such elements in the award are now all susceptible to change under Sections 34 and 37. The biggest concern is a new incentive for an award-debtor: to prolong litigation so as to drag the dispute till it reaches the Supreme Court. Having said this, these concerns can still be managed if these modifications are indeed ‘limited,’ and if the courts invoke this newfound corrective-jurisdiction minimally.
ABOUT THE AUTHOR
Yash Sinha is presently a lawyer in the courts of New Delhi, and has previously worked as a Judicial-Law-Clerk at the Supreme Court of India. He graduated from the National Law School of India University in 2019, and is interested in legal writing on commercial laws.
*The views and opinions expressed by authors are theirs and do not necessarily reflect those of their organizations, employers, or Daily Jus, Jus Mundi, or Jus Connect.
