Introduction
There is no decided formula which guides the arbitrators regarding which legal principles apply in determining when to join non-signatories. Allegations of implied consent implicate a different legal framework from arguments asserting lack of corporate personality. National law can come into play, particularly the place of incorporation of a signatory company. In addition, arbitrators often look to transnational norms established in other arbitral awards, as well as the law at the arbitral situs and the enforcement forum. The general trend is that non-signatories will generally not be bound to (nor able to compel) arbitration because there is no privity of contract. However, there are exceptions to this general rule of privity of contract. As an arbitration agreement is governed by the ordinary principles of contract law, non-signatories may be compelled to arbitrate in a number of circumstances.
A range of legal theories have been developed to facilitate this determination either for or against including such non-signatories. Parent companies, subsidiaries, contract assignees, governmental and quasi-governmental entities, beneficiaries and other non-signatories to an underlying arbitration agreement may find themselves bound by an arbitration agreement, and by the subsequent arbitral award. Compelling a non-signatory to arbitrate involves complex theories of contract and corporate laws that are themselves unsettled or at least highly fact dependent in their application. Various national courts and arbitral tribunals, from both, common law and civil law systems have evolved some of these theories:
- Group of Companies doctrine
- Veil-piercing or Alter-ego
- Agency
- Incorporation by Reference
- Estoppel
- Assignment
- Third-Party Beneficiary
Consenting and Non-Consenting Non-Signatories
When a non-signatory asks to arbitrate against a signatory, the threshold for extending the arbitration clause thus may be set at a lower level from its resting place when one side never abandoned the right to present claims or defences in otherwise competent courts. The signatory resisting joinder of the third party might argue that it never agreed to arbitrate with the particular affiliate seeking to enter the proceedings. The argument has some force, albeit limited in nature as the resisting party did agree that disputes related to the subject in question would be settled by arbitration. On the contrary, when a non-signatory is sought to held bound by an arbitration agreement, the very basis of arbitral jurisdiction – consent to arbitrate – is generally absent. The party sought to be bound would argue that it never agreed to arbitrate with anyone at all, thus requiring arbitrators to look fully for clear manifestation of consent – express or implied.
Consequently, arbitrators and judges often draw distinctions between what might be called “consenting non-signatories” (which seek to arbitrate) and “non-consenting non-signatories” (which resist arbitration). It is understandably easier to justify allowing a willing party to join an arbitral proceeding than the converse. The position can be summed up by Professor William W. Park’s opinion:
“Policy reasons as well as practical considerations make it difficult to compare a situation where the non-signatory does want to arbitrate with one where the non-signatory does not want to arbitrate. In the latter instance, the drawbacks of parallel proceedings must be weighed against the serious countervailing considerations of imposing arbitration clearly unwilling entities. When the non-signatory has never consented to arbitration more analytic rigor and hesitation are in order before extension should be ordered. The very basis of arbitral jurisdiction is prima facie absent.”
Group of Companies
The most widely known principle for the extension of arbitration agreements is the group of companies’ doctrine, which evolved in the famous Dow Chemical case. This doctrine provides that several companies that form part of a larger corporate group may be regarded as a single legal entity or “une réalité économique unique”. Under the doctrine, a group of companies constitutes one and the same economic reality - despite the legal independence of the individual entities from one another - where the circumstances of the contract's conclusion, its performance, its (possible) subsequent termination, and the degree of control executed among the group companies warrants such an inference. However, in all “group of companies” decisions, the finding of an express or implied consensus of the parties remains key in binding non-signatories to an arbitration clause, and this doctrine is not applied due to mere affiliation of the companies. In essence this doctrine states that when a non-signatory company of group of companies is an active participant in the contractual relationship, then the arbitration agreement can be extended to it.
TIn Dow Chemical, an ICC Tribunal sitting in Paris decided that the parent company, Dow Chemical Company (USA) should become a party to an agreement applying to its subsidiary Dow Chemical (France), for the following reasons:
Considering that it is indisputable – and in fact not disputed – that DOW CHEMICAL COMPANY (USA) has and exercises absolute control over its subsidiaries having either signed the relevant contracts or, like DOW CHEMICAL (FRANCE), effectively and individually participated in their conclusion, their performance and their termination.”
In a similar decision, a Singapore International Arbitration Centre tribunal extended an arbitration clause to a parent non-signatory company on the basis of “the true intent of the parties on the basis of the evidence before it”. Further, in SARHANK V. ORACLE CORPORATION, a tribunal sitting in Cairo decided, as a matter of Egyptian law:
“Despite their having separate juristic personalities, subsidiary companies to one group of companies are deemed subject to the arbitration clause incorporated in the contract because contractual relations cannot take place without the consent of the parent company owning the trademark by, and upon which transactions proceed.”
However, the common law countries such as England and Australia have been reluctant to apply this principle to bind the non-signatories. Peterson Farms , for instance, is a case where an English Court rejected this doctrine in an outright manner. In that decision, the Commercial Court held that “the group of companies’ doctrine . . . forms no part of English law”. In Australia and Switzerland too, this doctrine has been rejected for the purpose of extending the arbitration agreement to non-signatories. In India, the Supreme Court came across this situation in Indowind Energy Ltd. v. Wescare (I) Ltd. & Anr. However, it held that the mere fact that the Signatory and Non-signatory companies had common shareholders or common boards of directors did not make the two companies a single entity. The Court further held that the mere existence of common shareholders or directors could not lead to an inference that one company was bound by the acts of the other. Thus the extension of arbitration agreement was refused by the court reasoning that only signatories are bound by the same.
Professor Habegger notes that scholarly debate of this issue continues, particularly among French commentators. While one school of thought favours a stricter approach relying predominantly on national laws, the other is more lenient, allowing extension based on principles such as bona fides or good faith, lex mercatoria or other principles of private international law.
Veil-Piercing or Alter-Ego
A number of arbitral tribunals and national courts have often had to consider whether an arbitration agreement concluded by a company may be binding on its group affiliates or even a natural person who is the group's ultimate controlling shareholder. Although main limiting principle for the application of this doctrine is that corporate personality is created precisely in order to contain liability within a particular corporate entity - in practice, construction of the arbitration agreement in question, as well as the circumstances surrounding the entry into, and performance of, the underlying contract need to be considered.
Under the alter ego doctrine, a corporation may be bound by an agreement entered into by its subsidiary, regardless of the agreement’s structure or the subsidiary’s attempts to bind itself alone to its terms, “when their conduct demonstrates a virtual abandonment of separateness”. Alter ego determinations are highly fact-based, and require consideration of the totality of the circumstances. No single factor is final and conclusive. The courts have developed extensive lists of circumstances to guide alter ego determinations. In determining whether a non-signatory is the alter ego of a signatory, a court or a tribunal failing to take into account all of the aspects of the relationship between the non-signatory and the signatory commits an error of law. If it is established that corporate form was used to effect fraud or another wrong on a third party, alter ego determinations then revolve around issues of control and use. The courts explore the totality of the environment in which the party and non-signatory operate.
In ALPHA S.A. v. BETA & Co., State Company of Ruritanian Law, the tribunal pierced the corporate veil of the signatory state-owned entity to bind its non-signatory corporate parent to the arbitration agreement entered into by its subsidiary. The tribunal discussed case law and doctrine developed in connection with Swiss company law, in particular for so-called “one-man companies”, and summarized that piercing the corporate veil was only warranted where (i) a shareholder had total control over an entity, evinced by insufficient capitalization, confusion in the administration and management, and confusion of assets, and (ii) the totality of circumstances constituted an abuse of rights.
The veil-piercing issue generally includes those factors normally explored in the context of parent-subsidiary alter ego claims, such as whether:
- The parent and subsidiary have common stock ownership;
- The parent and subsidiary have common directors or officers;
- The parent and subsidiary have common business departments;
- The parent and subsidiary file consolidated financial statements;
- The parent finances the subsidiary;
- The parent caused the incorporation of the subsidiary;
- The subsidiary operates with grossly inadequate capital;
- The parent pays salaries and other expenses of the subsidiary;
- The subsidiary receives no business except that given by the parent
- The parent uses the subsidiary’s property as its own;
- The daily operations of the two corporations are not kept separate; and
- The subsidiary does not observe corporate formalities.
Additional factors to consider in an alter ego determination include:
- Whether the directors of the “subsidiary” act in the primary and independent interest of the “parent”;
- Whether others pay or guarantee debts of the dominated corporation; and
- Whether the alleged dominator deals with the dominated corporation at arm’s length.
The main limitation of an alter ego argument is that piercing the veil is an exception to a deeply entrenched rule in our legal and economic system. In an arbitration setting, the weight of various national authorities suggests that the veil should be pierced seldom. Moreover, an arbitration is created by contract, unlike a court proceeding. It might, therefore, be argued that arbitrators should be especially hesitant to pierce the veil. That is to say, since the doctrine is rarely applied by even the higher national courts which possess inherent jurisdiction, arguably arbitrators should be even more reluctant to pierce the veil.
Agency
Agency is “the fiduciary relation which results from the manifestation of consent by one person to another that the other shall act on his behalf and subject to his control, and consent by the other so to act.” An agency relationship may be demonstrated by:
“. . . written or spoken words or conduct, by the principal, communicated either to the agent (actual authority) or to the third party (apparent authority).”
A principal will generally be bound by an arbitration clause in a contract signed by its agent. As a result, arguments about whether a non-signatory should be compelled to arbitrate will only arise where there is no explicit contract between the principal and its agent, and the principal does not wish to be part of the arbitration. In China National, for example, the Federal Supreme Court of Switzerland upheld the decision by the arbitral tribunal against a signatory agent based on the fact that agent and principal held themselves out as one indistinguishable entity. In Thomson-CSF, S.A. v. American Arbitration Association , considerations such as mutual benefits derived from affiliation were rejected as insufficient to bind a non-signatory to an arbitration agreement signed by an affiliate, on the basis of agency principles.
A United States district court had compelled Thomson-CSF to arbitrate with Evans & Sutherland Computer Corporation (E&S) on the basis of an arbitration agreement between E&S and Rediffusion Simulation Limited (Rediffusion), a Thomson-CSF subsidiary. The district court applied a “hybrid” approach to compel arbitration by Thomson-CSF, a non-signatory to the arbitration agreement, relying on various factual factors. The Second Circuit Court of Appeals held that the district court had improperly extended the limited theories on which an arbitration agreement can be enforced against a non-signatory. It also stated:
“The district court’s hybrid approach dilutes the safeguards afforded to a non-signatory by the ordinary principles of contract and agency and fails to adequately protect parent companies, the subsidiaries of which have entered into arbitration agreements. Anything short of requiring a full showing of some accepted theory under agency or contract law imperils a vast number of parent corporations. This Court did not intend such an outcome in prior opinions and does not adopt such an approach here.”
A non-signatory cannot compel arbitration merely because he is an agent of one of the signatories. Courts, however, differentiate between whether one signatory may compel the principal of another signatory agent to arbitrate under an agreement that the agent signed as an authorized representative of its principal:
“... and he who would bind the principal is bound to ascertain, not only the fact of agency, but the nature and extent of the authority.”
The binding effect of an arbitration agreement concluded by an agent on behalf of a principal involves questions of authority (i.e. the agent's ability to bind the principal to such agreements) and allied questions of necessary form. Thus, an ICC tribunal while deciding whether a principal was bound by an arbitration agreement concluded by its agent distinguished between the law governing the arbitration agreement (in that case, the law of the seat of the arbitration), the laws which governed the agent's capacity to conclude an arbitration agreement on behalf of the principal (the law of the principal's registered office) and the form in which such capacity should have been conferred on the agent (the law of the jurisdiction in which the agreement between the agent and the principal was concluded).
National laws feature substantial differences on questions of necessary form (i.e. whether the principal's written authorisation is required) and content (i.e. whether the principal's authorisation need expressly envisage the conclusion of an arbitration agreement). For example, both Swiss and Austrian law require the principal expressly to authorise an agent to enter into an arbitration agreement on its behalf in order for a principal to be bound by such an agreement, but only Austrian law requires such express authorisation to be in writing. Under Italian, French and German law no particular form of authorisation is required.
Professor Sandrock, opines that an arbitration agreement concluded by an agent or representative without the principal's written authorisation would bind that principal only if in the circumstances third parties’ legitimate expectations required protection. To conclude, in an agency analysis, courts and tribunals generally look at firstly, whether a non-signatory party was made a party by formal representation or subsequent ratification and, as a secondly, whether the appearance of an agency relationship between signatory and non-signatory and a good faith reliance on this appearance to the counterparty's detriment warrants extending the arbitration clause to the non-signatory. Courts and tribunals will reject good faith reliance where the alleging party did not at least investigate the non-signatory's purported powers of representation.
Incorporation by Reference
It is possible to incorporate an arbitration agreement by reference. For example, standard conditions may be described in a document separate from a contract, but referenced in that contract. A party may therefore be a signatory to the main contract which references a secondary document containing an arbitration clause, but a non-signatory to that secondary document. However, under the theory of incorporation by reference that non-signatory may be subject to the arbitration agreement and compelled to arbitrate. Both English and Canadian courts have consistently held that the general words “all terms, conditions and exceptions governing the charter party are hereby incorporated herein” in a bill of lading are insufficient to incorporate an arbitration clause included in the charter party.
Where there is clear and specific wording, an arbitration agreement that is incorporated into a bill of lading will be enforceable as in the case of Nanisivik Mines Ltd. v. Canartic Shipping Co. The Canadian and British position is in accordance with the American jurisprudence concerning incorporation by reference. Non-signatories have been compelled to arbitrate under bills of lading and in the wider context.
Estoppel
Equitable estoppel prevents a party who knowingly accepts the benefits of a contract containing an arbitration agreement from avoiding the obligation to arbitrate. This theory has so far been recognized only in the United States and Canada, where two theories have been recognized for holding a party bound by an arbitration agreement under estoppel. The first theory is that a non-signatory who knowingly accepts the direct benefits of a contract containing an arbitration agreement can be compelled to arbitrate by a signatory. A leading case on this point is American Bureau of Shipping v. Tencara , where Tencara contracted with a syndicate to build a yacht. The contract required the American Bureau of Shipping (ABS) to classify the ship. Tencara entered into a contract containing an arbitration contract with the ABS to classify the ship. The ship sustained serious hull damage due to poor design and construction. Tencara sued the ABS in Italy. The yacht’s owners sued the ABS in France. The ABS brought suit in New York to compel all parties to arbitrate their claims together. The owners argued that they were not party to the contract between Tencara and the ABS and, therefore, not a party to the arbitration agreement. Following the first theory, the Court held that the owners were compelled to arbitrate as non-signatories. Since the owners had received the benefit of the ABS’s classification (e.g., lower insurance rates), they were estopped from claiming that the arbitration provision did not apply to them.
The second theory is that a non-signatory can compel arbitration with a signatory when the issues the non-signatory is seeking to resolve are inherently inseparable or inextricably intertwined with the agreement and the non-signatory is closely related to the signatory. The test to determine when non- signatories will be allowed to compel arbitration was enunciated in MS Dealer Serv. v. Franklin as follows:
“Existing case law demonstrates that equitable estoppel allows a non-signatory to compel arbitration in two different circumstances. First, equitable estoppel applies when the signatory to a written agreement containing an arbitration clause must rely on the term of the written agreement in asserting its claim against the non-signatory. When each of the of the signatory’s claims against a non-signatory makes reference to or presumes the existence of the written agreement, the signatory’s claims arise out of and relate directly to the written agreement and arbitration is appropriate. Second, application of equitable estoppel is warranted when the signatory to the contract containing an arbitration clause raises allegations of substantially interdependent and concerted misconduct by both the signatory and one or more of the signatories to the contract...”
The second American theory of equitable estoppel can further be sub-divided into two categories :
(a) Permitting non-signatories to compel arbitration by signatories
Sunkist Soft Drinks, Inc. v. Sunkist Growers Inc., is an example of a non-signatory compelling arbitration because the signatory had to rely on the terms in the arbitration agreement in asserting its claim against the non-signatory. The defendant Sunkist had entered into a licence agreement. The licensee was then acquired and absorbed by Del Monte. Del Monte, a non-signatory, was allowed to compel arbitration for breaches of the license agreement because the claims were “intimately founded in and intertwined with the license agreement.” Therefore, Sunkist was estopped from avoiding arbitration.
a) Permitting signatories to compel arbitration by non-signatories
The case of McBro Planning and Development Co. v. Triangle Electronic Construction Co., Inc. concerned a hospital renovation. Both McBro and Triangle had arbitration provisions in their respective contracts with the hospital, but none with each other. Despite the lack of a written agreement between the two parties, McBro nevertheless tried to compel an arbitration proceeding between itself and Triangle. The Court held that Triangle was estopped from refusing to participate in the arbitration because its claims were intimately founded on and intertwined with McBro’s underlying contract with the hospital.
Learned academicians have opined that the first theory decision violates arbitration’s general premise of consent as the resisting non-signatory never consented nor later consents to being bound by the arbitration agreement , and moves away from the two traditional requirements of estoppel, namely, misrepresentation and detrimental reliance by the party seeking the estoppel. Commentators also note that in many estoppel cases, the same result could have been obtained by applying the ordinary laws of contract.
Assignment
The issue of whether an arbitration agreement remains valid if the contract containing it is assigned to another party is not dealt with in any of the main international conventions – the New York Convention, the UNCITRAL Model Law and the European Convention on International Commercial Arbitration. Consequently, the issue remains one for domestic laws, to be determined by reference to the law governing the assignment, as well as the law governing the arbitration agreement.
Given the separability of the arbitration agreement from the rest of the contract, it might be supposed that special requirements apply to the assignment of the arbitration clause. This is the judicial point of view in English law . By contrast, German law makes the assumption that the arbitration clause will be assigned along with the main contract. Other laws such as French law and New York law, an arbitration agreement will be considered as producing mostly duties - rather than rights - and consequently requires express agreement on the part of the assignee in order to produce effects for that party. The Swedish Supreme Court appears to have adopted a middle position, namely that an arbitration clause will be presumed to be assignable if the parties have not expressly agreed otherwise, but once assigned it will operate vis-à-vis the assignee only if that party has actual or constructive knowledge of the arbitration clause.
Third-Party Beneficiary
While very similar to estoppel, the third-party beneficiary doctrine is somewhat distinct. Under a third-party beneficiary theory, a court must look to the intentions of the parties at the time the contract was executed. Parties are presumed to be contracting for themselves only. This presumption may be overcome only if the intent to make someone a third-party beneficiary is clearly written or evidenced in the contract. Primary indicia of a third-party beneficiary interest will be whether the non-signatory files a claim against one of the signatory parties.
In the leading case of London Drugs Ltd. v. Kuehne & Nagel International Ltd., the Supreme Court of Canada ruled that the strict rules of privity could be incrementally relaxed in order to conform to the commercial reality and justice. The Court ruled that if an employer’s limitation of liability clause extends its benefits to employees, then the employees are entitled to benefit from the clause if they are acting within the scope of their duties and performing the services provided for in the contract. It must be noted that the Court was cautious in its decision and refused to overthrow the often criticized rule of privity of contract in its entirety.
In Mississippi Fleet Card v. Bilstat, Inc., a Federal United States Court compelled the non-signatories to arbitrate as they were third-party beneficiaries of a contract containing an arbitration clause. The Court stated the American test for establishing that an entity is a third-party beneficiary as follows:
“(1) that the terms of the contract are expressly broad enough to include the third party either by name or as one of a specified class and (2) the said third party was evidently within the intent of the terms so used, the said third party will be within its benefits if (3) the promise had, in fact, a substantial and articulate interest in the welfare of the said third party in respect to the subject of the contract.”
Similar to agency and piercing the veil, even the outcome of arguments based on the third-party beneficiary theory are fact driven. However, it appears well settled as a matter of principle that non-signatories can compel arbitration through the third-party beneficiary theory.
Conclusion and Suggestions
Thus on analysis of various principles and case-laws, it can be inferred that in order to determine whether a non-signatory is bound by an arbitration agreement, courts andtribunals usually look at theories which revolve around implied consent or lack of corporate personality. Generally, transnational norms are invoked which are based on decisions in landmark cases or opinions of learned scholars. When the joinder is based on implied consent and related doctrines, the basic contract and agency principles ofthe applicable lex contractus or lex situs are followed, whereas when corporate law doctrines are used, the law of the place of incorporation of the company is considered.
While most of such cases are fact-dependant, a general flow and considerations have evolved from decided cases on the discussed doctrines. In essence, however, the issue boils down to, as other issues in international arbitration, to the true and genuine intention of the parties in light of good faith and equity. Perhaps acknowledging the controversy surrounding this issue, UNCITRAL has recently suggested the following possible amendments to the UNCITRAL Rules:
“The Swiss Rules, for instance, expressly provide, under Article 4, paragraph (2) that: ‘Where a third party requests to participate in arbitral proceedings already pending under these Rules or where a party to arbitral proceedings under these Rules intends to cause a third party to participate in the arbitration, the arbitral tribunal shall decide on such request, after consulting with all parties, taking into account all circumstances it deems relevant and applicable. The Working Group might wish to consider whether an express provision on third party intervention should be included in any revised version of the UNCITRAL Rules.”
A similar provision can be found in Article 22(5) of the Rules of the Chamber of Arbitration of Milan. It is suggested that lawyers engaged in drafting contracts which contain arbitration clauses must be sensitized to the fact that a non-signatory may be added to the arbitration. If this risk exists, then clients must be advised of this risk, and, language be added to the contract and, arbitration clause, to minimize the risk of a related non-signatory party being bound by the arbitrator’s decision. Finally, it is submitted that with exceedingly frequent non-signatory decisions, moreover, this controversy in international arbitration is on a path to gradually approach the most precious of all grounds in dispute resolution: predictability.
|