Agency as a mechanism for compelling a non-signatory to join arbitral proceedings
By Hanna Roos for YIAG
International investors, and those who advise them, continue to be vexed by the question of when a non-signatory, such as a sovereign state, can be compelled to join arbitral proceedings.
A typical scenario involves a private investor who has entered into an arbitration agreement with a wholly state-owned entity, but seeks to join the state into the proceedings. This is particularly common in the field of energy and infrastructure projects, where the investor may anticipate difficulties in enforcing any award against the state entity alone if the entity’s assets are transferred away or frozen, leaving a mere corporate shell behind. The SPP v the Arab Republic of Egypt (February 1983), Westland Helicopters Ltd v the Arab Organization for Industrialization et al (March 1984), Joint Venture Yashlar and Bridas SAPIC v Turkmenistan (June 1999) and Bridas SAPIC v Turkmenistan (June 1999) ICC arbitrations concern variants of this scenario.
The legal mechanisms for compelling a non-signatory to join arbitral proceedings are numerous and include agency, assignment, group of companies, alter ego, estoppel and state responsibility theories. Agency is often thought to be the simplest and least controversial theory particularly under English law (which considers the identification of the parties to an agreement to be a question of substantive, not procedural law), and therefore the investor’s likeliest ally.
As a matter of English law, an agency relationship arises when one person, called the principal, authorises another, known as the agent, to act on its behalf, and the other agrees to do so. An entity may therefore be bound as the principal by an arbitration agreement which it has not physically signed, but which the agent has executed on its behalf. This was argued and agreed by arbitral tribunals for example in Antoine Biloune v Ghana Investment Centre, Ad Hoc Awards (October 1989 and June 1990).
Although agency is an established common law principle, there is little English law authority on the application of the theory in the context of an attempt to join a non-signatory to arbitral proceedings, with the possible exception of the High Court case Peterson Farms Inc v C&M Farming Ltd  EWHC 121. But even in that case, the question was whether an agency relationship existed within a group of companies, rather than whether one could be found between a state and a state-owned entity.
Against that background, it is perhaps time to shine a fresh light on the 1954 House of Lords decision in Bank Voor Handel En Scheepvaart, NV v Administrator of Hungarian Property  1 All ER 969, which is cited as legal authority for the principle that public corporations created by statute are sometimes regarded as agents of the Crown.
In Bank Voor Handel En Scheepvaart, gold belonging to a Dutch banking company was stored in a safe in the City of London. When Germany invaded the Netherlands in May 1940, the gold became “enemy property” under the Trading with the Enemy Act 1939 and was transferred to the Custodian of Enemy Property for England and Wales. In 1951, the Dutch banking company obtained a High Court judgment awarding it the proceeds which the custodian had derived from the gold. The company argued that it was entitled to the sums without deduction of the income tax paid by the custodian in respect of the profits. This was on the basis that Crown status attached to the custodian who therefore was not liable to pay income tax.
The House of Lords found the custodian to be an agent of the Crown with reference to the following reasons:
• Appointment. The custodian was appointed by statute.
• Nature of rights and duties. The House thought that “the definition in the statute of [the custodian’s] rights, duties and obligations is highly important”. The custodian’s duties were enshrined in the Trading with the Enemy (Custodian) Order 1939 made by the Board of Trade, which included holding money paid to him and transferring property vested in him under a vesting order. It was considered particularly significant that the custodian asserted these powers on behalf of, and for the benefit of, the Crown.
• Degree of discretion. The pivotal question in the view of Lord Reid was to what extent discretion was bestowed upon the custodian by statute: “the grant of any substantial independent discretion takes the officer out of the category of servants of the Crown”. The Court found the discretionary powers negligible as the custodian did not, for example, determine which properties were to vest in him and for what duration.
• Role within government. Finally, the House of Lords thought it significant that the custodian was a “necessary part of the machinery of modern government” and a “cog in governmental machinery devised in modern conditions for performing the former prerogative functions of the Sovereign”.
Even though the facts of the case are highly specific, the Court’s reasoning is perhaps worthy of consideration in a modern context where a party to an oil or infrastructure project seeks to introduce a sovereign state into the proceedings.
Applying the Court’s test, a state entity is often established by a legislative measure. Its rights and duties are usually clearly defined in an enactment, and entitle the corporation to assert its powers on behalf of, and for the benefit of, the government in overseeing foreign direct investment projects. The entity also typically collects taxes or other payments from the investor for the benefit of the government.
As to the entity’s degree of discretion, its board of directors commonly includes ministers who exercise governmental oversight. Statute may also define precisely the entity’s permissible discretion. For instance, the entity may be authorised to make by-laws but have a corresponding obligation to seek the approval of a ministry for such laws. Moreover, a power to enter into contracts may be curtailed by a requirement that contracts above a certain value be approved by the government.
Finally, a state-entity overseeing significant foreign direct investments characteristically performs a necessary part of the machinery of that government. Its website, brochures and other official materials may include statements to this effect.
Thus, where a party to an oil or infrastructure project seeks to introduce a non-signatory state into the arbitral proceedings, the authority of Bank Voor Handel En Scheepvaart suggests that a relationship of agency may be made out.
To make sure you do not miss out on regular updates from the Kluwer Arbitration Blog, please subscribe here. To submit a proposal for a blog post, please consult our Editorial Guidelines.
Profile Navigator and Relationship Indicator
Includes 7,300+ profiles of arbitrators, expert witnesses, counsels & 13,500+ relationships to uncover potential conflicts of interest.
Learn how Kluwer Arbitration can support you.
This is an interesting angle on a much discussed issue but I think ultimately the issue of agency or otherwise will be very much determined by the facts of any particular case.
The starting point is correct in the sense that the issue of whether or not a state is to be considered a party to a contract (and specifically to an arbitration agreement) is to be determined by normal principles of contract law and interpretation – although there has been much academic discussion of such principles as alter ego, estoppel and state responsibility, there are few or no actual examples of the application of such principles producing an actual result ( a lot of discussion but no decisions). The cases where the state has been bound to the contract of one of its organs such as was the case in one of the Bridas cases have been decided on the basis of a finding about the intention of the parties. The relationship of principal and agent (at least under English law) comes about as a result of a deliberate act by the Principal. Ms Roos makes that clear.
However the examples she gives of possible situations where this relationship may be established sre unlikely in practice. In modern government practice state owned entities tend to be established with the deliberate intent to distance government from the enterprise. One of the reasons is often to keep debt off the public balance sheet and in order to do this it is necessary to demonstrate a very clear degree of independence. Thus the situations where the statute or founding document demonstrates subservience to the degree that they can be read as showing a generalised agency relationship are likely to be rare.
I discussed some of these issues in an article in Arbitration International (Volume 20 Number 4 2004, p 287) and reached the following conclusions:
• The issue of whether a state is bound to a contract is entirely one of the intention of the parties;
• the parties whose intentions must be judged are the private counterparty and the state itself (not the state’s organ);
• the test of intention is not one of ‘reasonable expectations’ raised in the mind of the private counterparty by the behaviour of the slate during or after the negotiations;
• the only true test is one of subjective intention on the part of both parties. External ‘objective’ factors may serve to prove the existence of such subjective intention, but however many of these factors there may be, they will not make the state liable in the absence of an intention on its part to be bound;
• the state organ may be acting as the agent of the state, but if it is so doing, the organ itself cannot be liable for its acts to the extent that it is acting as agent:
• the concept of ‘alter ego’ meaning that the state itself is treated as the contracting party notwithstanding the absence of provable intention or a provable agency relationship, has not moved beyond being a theoretical possibility and it is unlikely ever to be the correct basis for a finding that a state is a party. The issue is more likely to be resolved in favour of the claimant by an ‘agency’ or ‘true intent’ argument. If it is not resolved in that way there is unlikely to be the basis for a finding that the organ is an alter ego of the state. If the agency and intent arguments fall to be resolved against the claimant, the claimant is unlikely to succeed in an alter ego argument;
One of the interesting results of the above conclusions is that if the Claimant succeeds in showing that the state organ is merely an agent of the state, it may lose its right to an award against the organ. Although, as Ms Roos correctly points out, the motivation for pursuing the state rather than the organ may be that it is concerned about its ability to enforce against the organ, it is also the case that it is often difficult to enforce against the state. Thus in successfully joining the state the claimant may be jumping out of the frying pan into the fire!
Mr Rosenberg: I take advantage of this way to ask you about what you think of the “Bridas” US Court decision (2006) which talks about the low capitalization of the state enterprise to consider the State as responsible.
Thank you for your time, best regards
I agree with most of Mr. Rosenberg’s comments, with the exception of the last. If the Claimant does succeed in demonstrating agency between the state and the state organ, and goes on win on merits, it will in fact have an award that will be enforceable in the alternative against the state and the organ. Their liabilities will be independent and co-extensive. I fail to see how Mr. Rosenberg reaches the conclusion that agency will absolve the organ of liability; it remains bound by the contract and its consequences. If anything, therefore, it will be a case of the Claimant jumping out of the frying pan, getting under the stove and watching the two burn!