Akash Kumar Surya*

The paper explores the debate between using consent-based doctrines like Group of Companies (GOC) Doctrine and non-consent-based doctrines like alter ego to implead a non-signatory to an arbitration agreement. This debate is explored in the context of the Arbitration and Conciliation Act in light of the recently delivered Supreme Court judgement in Cox and Kings II. The paper attempts to justify the adoption of consent-based doctrines to the exclusion of others in the Indian jurisprudence by employing the framework of Exclusive Legal Positivism. It scrutinizes the present judgement to highlight its inadequacy in not even engaging with non-consent-based doctrines given the bench was a five-judge bench and had the opportunity to evaluate different approaches to the issue. It intends to fill this gap by resorting to the ELP framework using which it argues that the posited law did not permit the adoption of non-consent-based doctrines.

Introduction

A five-judge bench of the Supreme Court of India (SC) recently delivered its judgment in the case of Cox and Kings II. The case is a landmark one in the context of arbitration law as it settles the legal position of the Group of Companies (GOC) doctrine in India- a doctrine that seeks to implead non-signatories to the arbitration agreement by relying on the notion of implied consent. The case arises out of a reference made by a three-judge bench in the case of Cox and King I where the SC expressed reservations with respect to the adoption of the GOC doctrine in Indian jurisprudence. In this case, after a thorough discussion of previous decisions on the subject, the majority was of the opinion that the adoption of the GOC doctrine by virtue of the law laid down in the case of Chloro Controls, was based on considerations of economics and convenience rather than of law (para 55). It further observed that the doctrine was theoretically based on considerations of efficiency and that efficiency could not have been the sole basis to bind a party to arbitration in the absence of a legal basis for grounding the doctrine in Indian jurisprudence (para 17). Accordingly, it framed two issues for the five-judge bench to consider in Cox and Kings II, one of which questioned the legal validity of the GOC doctrine in the Indian context (para 54).

Settling the law on this question, the SC in Cox and Kings II relied on sections 2(1)(h) and 7 of the Arbitration Act (Act) to provide a legal foundation for the GOC doctrine. According to the SC, a literal interpretation of the said provisions was sufficient to justify the adoption of GOC in India. However, the court did not examine the possibility or efficacy of employing competing doctrines which could have also served the purpose of impleading the non-signatory to an arbitration agreement. The court seems to have engaged in an exercise of backward reasoning- proceeding with the conclusion that the GOC doctrine is what should be adopted in India and then determining the relevant legal provisions to justify it. In doing so, the court missed an opportunity to evaluate if the relevant legal provisions allowed the court to look past GOC towards other competing doctrines to determine if the Indian law also permitted the adoption of the latter.

In this paper, in order to address this gap, I argue that the Act did not provide any scope for the adoption of any other competing doctrines that are not based on the notion of consent. To do so, I firstly, analyse the key distinction between two groups of doctrines- while one group contains doctrines that are consent-based, the other group contains doctrines that rely on principles other than consent- in order to better situate the present judgement of the SC; and secondly, employ an Exclusive Legal Positivist perspective to the interpretation of the Act to substantiate my argument.

Non-Consensual Theories- The Road Not Taken

The GOC doctrine plays a key role in arbitration law and was developed to justify the imposition of liability arising out of an arbitration agreement upon an entity (non-signatory) that might not have originally signed the agreement itself. However, its adoption in legal systems across jurisdictions has been a subject of vehement scholarly debate primarily because commercial arbitration, as a means of dispute resolution, is characterised by two basic principles- party autonomy and a mutual intent to arbitrate a dispute i.e., consent.

In the contemporary commercial climate, where businesses are encouraged to invest, diversify, and limit their liabilities, the tendency to establish subsidiaries has grown immensely. In this context, a major question that has arisen is whether a parent company can be bound by the consent to arbitration provided by the subsidiary and vice versa. Ordinarily due to the operation of separate legal personality principle, one company or its shareholders are not bound by the actions of the other company. This is also termed as the principle of corporate veil- a protective shield which protects the shareholders from being liable for the actions of the company. So, if one subsidiary signs an arbitration agreement, the holding company or sister subsidiaries would, by implication, not be bound by any of the obligations of the first subsidiary. However, keeping commercial and economic realities in mind, arbitral tribunals have felt the need to extend their jurisdiction to even non-signatories (p. 35). Now, there are two different ways (p. 35) to achieve this objective- firstly, relying on doctrines like GOC which are grounded on the idea of implied consent, or secondly, resorting to traditional doctrines of contract law doctrines like the doctrines of veil-piercing, estoppel, or third party beneficiary^[1] that do not rely on the notion of consent.

The key difference between the two methods lies in the importance given to consent in the former. For instance, traditional doctrines like veil-piercing or alter-ego, etc. are based on principles of equity, fairness, and justice. When the courts are faced with situations wherein the parent company uses the subsidiary to commit fraud or evade liabilities, and realize the challenge thrown by the separate legal personality principle, they usually respond by piercing the corporate veil and holding the parent company liable for the actions of the subsidiary. In this approach thus, the court is not concerned with the consent of the parent company at all. However, in the GOC doctrine, the considerations of fairness and justice are not at all relevant and the only pertinent issue to impose liability upon a non-signatory parent company is to determine if the non-signatory had consented to be bound by the arbitration clause. In the context of arbitration law, which accords paramount importance to consent, it is reasonable to adopt the first approach and impute liability upon a non-signatory only if it is found to have consented to the agreement. Doing this would allow one to achieve a balance between two equally important considerations- preserving the foundational principles of arbitration law and acknowledging the modern approaches to business.

It is no surprise, then, that the SC chose to proceed with the first method and adopt the GOC doctrine to implead non-signatories in the arbitration agreement. While it seems logical to proceed with a consent-based theory to implead non-signatories in an arbitration agreement, the SC has not reasoned out why it preferred a consent-based doctrine over the non-consent ones. After all, even jurisdictions like the U.K., whose case laws are so enthusiastically cited by Indian courts, have shown significant reservations in adopting the GOC doctrine. English law places paramount importance on the notion of privity of contract and does not allow the extension of arbitration agreements to non-signatories. It requires the intention of the parties to be readily apparent and not inferred. Thus, what appears to be clear from this discussion is that there are two competing notions- the notion of privity of contract and the notion of consent in arbitration. The SC has evidently sought to preserve the notion of consent over the notion of privity of contract.

However, as mentioned above, the SC seems to have presupposed the existence of only consent-based doctrines in the Indian context, thereby not reasoning out the exclusion of the non-consent doctrines. Agreeing with the stance of the SC, I seek to fill this gap and argue that the Arbitration Act, the relevant law that governs the law on arbitration, itself privileges the notion of consent. To substantiate my argument, I adopt an Exclusive Legal Positivist perspective in interpreting in interpreting the relevant provisions of the Act which indicates that there was no scope for the SC to legally ground a non-consent-based doctrine in India.

An ELP view of the Arbitration Act- Justification for the adoption of the GOC Doctrine

The rationale behind using the ELP approach is two-fold- Firstly, employing this framework permits us to examine the judgment by tracing down its very reasoning to evaluate whether the judgement is a function of mere whims and fancies of the judge couched in legalese or is an outcome of legal reasoning arising from a valid source of law like a statute or a binding precedent. Secondly, by focusing on the source, the ELP framework helps ensure that the outcome arrived at by the court is as close to the most efficient outcome as possible. To illustrate, consider Ronald Coase’s law and economics theory on transaction costs which asserts that if there are no transaction costs in a market, the resources in the market would inevitably end up being allocated efficiently in a Pareto-Optimal outcome.[2] Further, Calabresi argues that there are hardly any allocations that are free from transaction costs due to which the fundamental aim of any legal system must be to allocate primary resources to reach a situation as close as possible to what would otherwise have existed had there been no transaction costs in the market. It is in this context that the importance of statutes can be appreciated. Statutes can be seen as instruments of intervention that are intended to address market inefficiencies and govern the relationships between entities in such a manner that the most efficient outcome is produced. Given that the legislature is the most well-suited organ of the government to frame laws, it is reasonable to assume that its assessment while enacting any law is correct and is intended to lead to the most efficient outcomes. Then, if a court’s judgement is not solidly backed by the posited law, it would reflect a reallocation of resources different from what was envisaged by the legislature. This would produce inefficient outcomes. Thus, it becomes important to test the judgements using the framework of ELP.

The ELP approach posits that all law is source-based based i.e., a judge is merely supposed to apply the law from the source which requires him to undertake legal reasoning, and not entertain moral considerations.[3] The idea therefore is that a law is valid if and only if its existence and contents can be determined without using moral arguments (p. 47). This idea of source is not merely limited to formal sources like legislation but also includes judicial precedents and interpretive sources arising out of such legislations.[4]

Using this perspective, one can justify the adoption of the GOC doctrine in India by referring to sections 2(1)(h) and 7 of the Arbitration and Conciliation Act (Act) as the court has also done. The court rightly reasons that section 2(1)(h) of the Act defines a party to mean a “party to an arbitration agreement”. Now, an arbitration agreement is a creature of contract, and it is only logical to argue that the principles of contract law need to be applied to determine who is a proper party to the arbitration agreement. In the context of contract law, one of the key principles that the courts rely on to determine if an entity is a party to a contract or not is the principle of free consent as contained in section 10 of the Indian Contract Act i.e., whether an entity had consented to being bound by the rights and obligations under the contract.

Obviously, the most common way of determining whether a party has submitted to the jurisdiction of the arbitral tribunal or not is to check if the party has consented to the agreement by putting in its signature. However, the court recognises that in contract law, the requirement of signature is not a necessary condition for the expression of consent.[5] Thus, we see that the posited law itself is based on the notion of a valid contract which is centered on the idea of consent. In Haji Mohammad, the SC had relied on the conduct of the defendant to observe that an implied contract existed between the parties. Extending this traditional principle of contract law and applying it to the posited law i.e., section 2(1)(h), it can be argued that one can be said to be a party to the agreement if one had expressed one’s consent to be bound by the agreement in any manner including by putting a signature.

This logically means that even a non-signatory can be validly argued to be a party to the arbitration agreement as long as its expression of consent can be established through any other means. A literal interpretation of the section makes it clear that neither it derogates from the principles of contract law, nor does it contemplate any additional requirements. So, viewing from an ELP perspective, since the provision itself is centered on the notion of consent, it would have been unreasonable for the court to adopt a non-consensual theory to implead non-signatories.

However, the Act itself also contemplates an additional requirement to be met for the arbitration agreement to be valid.[6] This is contained in sections 7(3) and (4). While section 7(3) stipulates that an arbitration agreement must be in writing, sub-section 4 postulates an exhaustive list of three different scenarios when the arbitration agreement can be said to be in writing. This requirement of a written arbitration agreement, according to the SC in Cox and Kings II, is also present to merely ensure that there is a clearly established record of consent between the parties.[7] Thus, according to the SC, even section 7, while imposing the additional requirement of a written arbitration agreement, is only primarily concerned with the determination of consent between the parties.

The conclusion reached by the court seems consistent with the established interpretation of section 7. Explaining the scope of this provision, the SC, in Jagdish Chander had observed that to constitute an arbitration agreement, the court was required to look at inter alia the following elements- the intention of the parties to be gathered from the terms of the agreement; the parties should have agreed (consensus ad idem) to refer the dispute to an arbitral tribunal; the parties should have mutually intended to be bound by the decisions of the tribunal.[8] Similarly, a number of other SC judgements- Bharat Bhushan Bansal v. U.P. Small Industriesand Bihar State Mineral Dev v. Encon Builders– have also observed that for any clause to be construed as an arbitration agreement, one of the relevant requirements needed to be seen was whether there was any express or implied agreement (consensus ad idem) to refer the disputes to arbitration or not.

Thus, a survey of the case laws on section 7 also clearly indicates that the provision itself does not diverge from the requirement of consent, and in fact, is based on the very idea of consensus ad idem. It requires the arbitration agreement to be in writing and merely provides the three different circumstances in which the Act considers an arbitration agreement to be in writing. So, all that this provision requires is that the arbitration agreement, to whom even a non-signatory can be made a party, needs to be in writing. This in no way derogates from the fundamental requirement of the consent of the parties, and in fact reinforces the notion that for an arbitration agreement to be valid, there must be an express or implied consensus ad idem.

What emerges from the above discussion is as follows: the GOC doctrine is not a function of moral or political considerations as its contents and existence can be traced back to provisions of the Act as interpreted by various precedents using the interpretive principles of contract law. Thus, a combined reading of sections 2(1)(h) and 7 in light of the settled principles of contract law suggests that the court would have faced a tough task in justifying the adoption of a non-consent theory to implead non-signatories. Employing this ELP lens, it can be argued that there is a firm foundation for the GOC doctrine to be adopted in India as its authority can be derived from the parent legislation and the precedents themselves.

Lastly, the Preamble to the Arbitration Act also clarifies that it is based on the UNCITRAL Model Law and therefore courts have referred to the latter as an aid in the interpretation of the former (para 26).  Although Article 7 of the Model Law provides that the arbitration agreement has to be in writing, it further specifies that its actual contents can be recorded in any manner including through oral means or through conduct. This provision thus seems to provide an exception to the general rule (p. 37) that every aspect of the arbitration clause needs to be written and explicitly consented to by the parties.

Therefore, the Model Law itself envisages a situation where, if an entity conducts itself in a manner that indicates the existence of implied consent to the arbitration agreement, it is considered to be a party to the agreement. Notably, the provision focuses on the conduct of the entities to determine if an arbitration agreement has indeed been recorded or not. Evaluating an entity’s conduct provides the courts with an opportunity to infer if the party intended to be bound by the agreement or not. Thus, this focus on the “conduct” signifies that the Model Law also relied on the notion of consent to determine if an entity was a party to an arbitration agreement in case the agreement was not reduced to writing. The Model Law could have very well adopted a different approach- one based on notions of fairness and equity as opposed to consent. However, the conscious choice of preferring consent over other values indicates that, at least, the Model Law does not provide us much wiggle room to justify the adoption of non-consent-based doctrines but provides us with sufficient scope to justify the adoption of a consent-based doctrine to implead a non-signatory to the arbitration agreement.

Conclusion

An ELP-based justification of the GOC doctrine, therefore, adequately responds to the reservations raised by the three-judge bench in Cox and King I. It addresses the valid concerns raised by the majority to the effect that the adoption of GOC seemed to be a function of the efficiency and convenience it brought by reducing the possibility of multiple litigations over the same dispute. However, an exercise of legal reasoning under the ELP framework clearly shows that a combined reading of sections 2 and 7 of the Act leaves no scope for one to incorporate any non-consent-based theory. The reliance on the UNCITRAL Model Law provides us with further reasons to substantiate the argument this paper makes. Since, the provisions themselves are centered on the notion of consent and consensus ad idem, they provide a solid foundation to justify the adoption of GOC doctrine, to the exclusion of non-consent-based doctrines, in the Indian arbitration law jurisprudence.

[1] William H. Park, ‘Non-Signatories and International Contracts: An Arbitrator’s Dilemma’ in Doak Bishop (ed), Multiple Party Actions in International Arbitration (OUP 2009) 4.

[2] R. H. Coase, ‘The Problem of Social Cost’ (1960) 3 Journal of Law and Economics 1.

[3] See the chapter on Exclusive Legal Positivism by Andrei Marmor beginning at page 83.

[4] Ibid.

[5] Cox and Kings II [66].

[6] Cox and Kings II [69].

[7] Cox and Kings II [75].

[8] Jagdish Chander v. Ramesh Chander (2007) 5 SCC 719 [8].

*Akash Kumar Surya is a third-year law student at the National Law School of India University pursuing the undergraduate B.A., LL.B. course.