[By Ryan Bang, Final Year B.A. LL.B. (Hons.) student at NLIU, Bhopal]
At the end of the Constitutional Convention in 1787, Benjamin Franklin was asked what kind of government had been created. His answer was “a republic, if you can keep it.”
The Arbitration and Conciliation Act, 1996, might be understood as offering parties something similar, an agreement, if you can keep it. In Gayatri Balasamy v. ISG Novasoft Technologies Ltd.,[i] the Supreme Court appears to have added a couple of qualifications to this understanding: an agreement, even if we cannot quite keep it.
It was ruled in that case that the court may exercise a limited power of post-award modification in cases involving severable portions of an award, correction of clerical, computational or typographical errors, modification of post-award interest, and to do complete justice.
The judgment was not without some criticism. A recurring concern was that allowing this form of post-award modification could open a back door to merits review and invite courts to substitute their own views for those of the tribunal (here, here and here). In my view this is not the strongest criticism. It focuses on the mere possibility (however high that possibility might be) of misuse. The majority opinion sought to guard against that possibility by identifying the circumstances in which modification is permitted and cautioning against rewriting awards on the merits. A critic who says “this may be misused” can, at least in principle, be met with the answer that “it may not be.”
But that response does not answer a more fundamental question. Even if the limits are scrupulously observed, why does the power exist in the first place? To answer that, we must return to a more basic question: what does it mean to enter into an arbitration agreement?
I. When you sign, you sign the consequences
An arbitration agreement is a contract. Section 7 of the Arbitration and Conciliation Act, 1996, (hereinafter “the Act”) recognises an agreement by the parties to submit disputes to arbitration. Like other contracts, it creates an obligation for the parties to submit their disputes to arbitration.
The law may recognise grounds on which a contract can be avoided. But the mere fact that the bargain has turned out badly is not sufficient by itself. Consider a straightforward example, a person agrees to sell an asset for ₹10 lakh when its market value is ₹20 lakh. The seller later discovers the disparity and regrets the transaction. Under Section 25 of the Indian Contract Act, 1872, inadequacy of consideration does not automatically invalidate the agreement. There is no legal requirement that every agreement should produce a commercially desirable result. Instead, the is a focus on the commitments that the parties have undertaken.
The same principle should inform our understanding of an arbitration agreement. Parties choose arbitration because they agree to have their disputes resolved through a particular process. They cannot know in advance whether the tribunal will accept their interpretation of the contract, agree with their evidence, or assess their losses as they expect. Yet they agree to submit their dispute to that process. Whether arbitration is beneficial or cumbersome becomes non est the moment contract with arbitration agreement clause is signed.
To understand this better Randy Barnett’s A Consent Theory of Contract is helpful. Barnett treats contract law as the part of a system of entitlements that identifies the circumstances in which rights can be transferred through consent. The significance of this account is that a contract does not merely record a party’s preferences at the time of agreement. It establishes commitments that continue to bind even when those preferences change.[ii]
Patrick Atiyah’s The Rise and Fall of Freedom of Contract deepens this point. Atiyah identifies risk allocation as one of the more important functions of contracts. A transaction allocates risks between the parties; that allocation loses its purpose if the arrangement is not binding from the outset.[iii] If the consequences of that risk can be shifted merely because another outcome seems preferable, the original allocation loses its meaning. Put plainly, when you sign, you sign the consequences.
II. Arbitration is more than the first stage of litigation
This contractual understanding of arbitration must, however, be stated carefully. An arbitration agreement is a contract, but that does not fully explain the legal nature of arbitration.
In Legal Theory of International Arbitration, Emmanuel Gaillard examines competing ways of understanding international arbitration. One treats the arbitrator as part of the legal order of the seat, in a role analogous to that of a national judge. Another understands international arbitration through the relationship between multiple legal orders. Gaillard ultimately favours a third representation which understands international arbitration as an autonomous legal order.[iv] Although Gaillard develops his theory in the context of international arbitration, his account offers a useful way to understand the character of domestic arbitration. The relevance of his theory lies in recognising that arbitration is a distinct form of adjudication and not merely a stage of ordinary litigation.
This perspective challenges the reasoning in Gayatri Balasamy, which treats the additional proceedings only as a cost to be avoided. This way of counting the stages of dispute resolution risks treating the arbitral award as a first-instance decision awaiting judicial correction, rather than as the result of the adjudicative function entrusted to the tribunal. Parties choose arbitration as a means of resolving their disputes, and the courts retain the powers conferred on them by law. But the existence of those powers, and their permissible scope, must be established by interpreting the Arbitration and Conciliation Act, 1996 and not simply by the desirability of avoiding further proceedings.
The majority’s reasoning in Gayatri Balasamy, particularly paragraphs 40 to 42, places considerable emphasis on the practical objectives of arbitration. It reasons that allowing modification may avoid the expense and delay of another round of proceedings. If a court can modify an award rather than set it aside, the dispute may be brought to an end more efficiently.
That is understandable, but notice how arbitration is counted in this reasoning. It becomes the first of several stages in a process that runs through the courts, and that process is judged by its cost and delay. The principle of efficiency is one of the several advantages of arbitration, and not the legal underpinning of it.
The issue is not whether efficiency matters. It does. The issue is whether the desirability of a remedy establishes the existence of the power to grant it. A court cannot assume that it possesses a power simply because exercising it would save the parties time and money. Its authority must be found in the law.
III. Section 5, Section 34 and the dissent
The legal justification offered by the majority is equally unsatisfactory too. Whatever the circumstances, putting the cart before the horse cannot be justified. A legal framework must lead the reasoning, not be made to follow a conclusion already reached. Pulling in one direction is almost always fruitful; creating a situation in which the Court appears to be right whichever direction it takes, however, inevitably leads to absurdity.
Section 5 of the Arbitration and Conciliation Act provides that no judicial authority shall intervene in matters governed by Part I except where so provided in that Part. It does not eliminate judicial supervision of arbitration. It makes the statutory basis for that supervision central to answering the question in Gayatri Balasamy is whether the powers of the Court under Sections 34 and 37 of the Arbitration and Conciliation Act 1996 will include the power to modify an arbitral award?
Justice K. V. Viswanathan’s dissent develops the objection that the statutory framework does not confer a general power to modify awards. It distinguishes between setting aside an award and modifying it, and rejects the suggestion that a power to do the former necessarily includes a power to do the latter. The majority reaches a different conclusion and recognises a limited power of modification under Sections 34 and 37.
Setting aside removes an award, wholly or in part. Modification changes its terms while preserving it, at least to some extent. The two remedies are not the same legal act. The distinction is reflected in the English Arbitration Act 1996. Section 67(3), which deals with challenges to an arbitral tribunal’s substantive jurisdiction, expressly empowers the court to confirm, vary, or set aside an award. Section 68(3), by contrast, deals with serious irregularities in the arbitral process and provides that the court may remit the award to the tribunal for reconsideration, set it aside, or declare it to be of no effect. Section 71 addresses the effect of the court’s order, including the effect of a variation made under the Act. These provisions illustrate that setting aside, varying and remitting an award are distinct forms of judicial intervention.
The English provisions do not establish the meaning of the Indian Act, and Section 71 is not itself a free-standing power to modify. But the English legislature expressly distinguishes between these two forms of intervention and specifies their legal effects. Modification is a distinct power that needs to be addressed directly by legislation. In the absence of such an address, the Court cannot speak this power into existence simply because it is more efficient.
IV. Why non-intervention matters
Judicial intervention is often criticised on the grounds that it can increase costs, cause delay and undermine the efficiency of arbitration. Here, the judiciary intervened precisely because it wanted to prevent this from happening, so from the Court’s perspective, it would make sense to let courts have the power to modify an award. But if we think about it, there is something much deeper at stake here.
The deeper issue is the relationship between the parties’ commitment and the authority of the court. Arbitration is chosen to resolve disputes through a particular process. Its value depends, in part, on the parties being able to understand the legal consequences of that choice and to rely on the framework governing the resulting award.
If courts treat awards as outcomes that may be adjusted whenever another result appears preferable, the allocation of risk changes. The parties may have agreed to arbitration, but the consequences of that agreement become less predictable. The award begins to resemble an intermediate result in a continuing dispute over what the court considers appropriate.
H.L.A. Hart’s account of obligation helps us broaden the scope of this issue. Hart distinguishes between being obliged to do something, as one is when a gunman makes a threat, and having an obligation.[v] The difference lies in what he calls the internal point of view. A person has an obligation when they accept a rule as a standard for their own conduct and as a reason for acting, rather than merely as a prediction of what will happen if they do not.[vi]
This is where the question stops being only about courts. Despite efforts to promote India as an arbitration-friendly jurisdiction, courts and parties alike continue to approach arbitration as though it were merely an extension of the judicial system i.e. a different forum for the first round of adjudication. A party who signs an arbitration clause, and then gives the mechanism importance only as long as it feels advantageous, has not accepted the obligation in Hart’s sense. They comply with it for as long as it suits them, and treat a challenge as the next round of the same dispute.
Courts do the same when they honour the limits in Sections 5 and 34 only when the outcome seems acceptable. In both cases a framework that is meant to bind is treated as something to be worked around. This diminishes the institution and the role it is, designed to play.
An arbitration-friendly jurisdiction cannot be built through legislation, institutional reform, procedural amendments, or promotional campaigns alone. It also requires a change in judicial and professional attitudes towards arbitration, and in the attitudes of the parties who choose it. We must be willing to examine not only whether our laws formally support arbitration, but also whether our reasoning, and conduct, gives meaningful effect to that choice.
None of this requires courts to abandon their role. Arbitration is still subject to the statutory framework. Hart himself accepted that judges must sometimes fill gaps where rules run out,[vii] but the Act’s structure, which provides for setting aside and not modification, reads less like a gap than a deliberate choice. The point is that the parties must choose arbitration with a clear understanding of the consequences of that choice, and then keep to it, and courts must hold the parties to that choice even if it means undesirable outcomes.
Otherwise, arbitration risks becoming like a sheltered child under the constant supervision of an overbearing parent. If arbitration is to develop as a credible and genuinely distinct means of dispute resolution, courts must recognise that supporting the institution sometimes requires resisting the impulse to supervise it, and parties must resist the impulse to invite it.
Conclusion
An arbitration agreement is not simply an agreement to begin a process. It is a commitment to accept the legal consequences of that process, including an outcome that may disappoint one or both parties. This does not make arbitral awards immune from judicial scrutiny, nor does it place arbitration beyond the reach of law. It does, however, require courts to distinguish between correcting an award on a legally recognised ground and altering its outcome because another result appears more satisfactory.
If India is serious about becoming a successful arbitration jurisdiction, it must do more than enact arbitration-friendly legislation or establish institutions to administer arbitrations. It must also reconsider the attitudes through which arbitration is understood. Parties must be able to choose arbitration with a clear understanding of its legal consequences, and courts must respect that choice within the framework the law provides.
[i] Gayatri Balasamy v ISG Novasoft Technologies Ltd 2025 INSC 605.
[ii] Randy E Barnett, ‘A Consent Theory of Contract’ (1986) 86 Columbia Law Review 269.
[iii] Patrick S Atiyah, The Rise and Fall of Freedom of Contract (Clarendon Press 1979).
[iv] Emmanuel Gaillard, Legal Theory of International Arbitration (Martinus Nijhoff Publishers 2010).
[v] HLA Hart, The Concept of Law (2nd edn, Clarendon Press 1994) 82.
[vi] ibid 88-89.
[vii] ibid 124.
Leave a Reply