Party autonomy and choice of law: is international arbitration leading the way or marching to the beat of its own drummer?

Author:Karton, Joshua D.H.

When a choice of law issue arises, who decides: the parties (i.e., party autonomy) or the adjudicator? If the adjudicator decides, by what rules will the decision be made? As Catherine Walsh described in her 2010 Rand Lecture, party autonomy has been a flashpoint in discussions of domestic choice of law regimes. Over the past few decades, party autonomy has generally expanded, but a variety of limitations remain. If an adjudicator automatically applies the law or rules of law chosen by the parties, (1) it would seem that their rights and interests have been protected. However, acceding to contractual choices of law may harm third parties or contravene the public policy of a country. Accordingly, courts have restricted party autonomy in choice of law. The advisability of such restrictions has led to deep philosophical debates about the interests that choice of law should serve and to earnest discussions of the proper scope of party autonomy. Party autonomy may be on the march, but it continues to face opposition.

International arbitration provides an interesting counterpoint. One might think that similarly earnest discussions and deep debates would have taken place within the international arbitration community. After all, parties to international arbitrations by definition come from different states, and arbitral tribunals are not tied to any particular state. As a result, choice of law is always implicated and there is no default law to serve as a starting point for determinations. Nevertheless, the march of party autonomy in international arbitration has been unhindered. The discussions have not been philosophical but rather practical: how best to promote party autonomy and expand its scope.

National courts have abetted this process in the international arbitration context, even as they have maintained restrictions on party autonomy in choice of law within their own legal systems. This phenomenon, combined with the gradual loosening of restrictions on party autonomy in national legal systems, has led some to claim that international arbitration is leading the nations of the world into a global choice of law regime in which party autonomy reigns. Such triumphalists look forward to the day when party autonomy is as unrestricted in national courts as it is in arbitration.

That day is unlikely to arrive. The different pressures and interests that shape national court litigation on the one hand, and international arbitration on the other, are likely to generate different levels of support for party autonomy in choice of law. As a result, regardless of whether more party autonomy in choice of law is preferable in all contexts or whether it ought to have a wider scope in arbitration than in litigation, arbitrators are likely to maintain greater deference to party autonomy than will national legislatures or courts. (2)

Part I considers the choice of law process in international arbitration and how it differs from that prevailing in national courts, with particular attention to party autonomy. Part II proposes some reasons why international arbitral tribunals and national courts have accepted party autonomy to different degrees, and why this disparate treatment is likely to continue. Part III concludes and suggests a few lessons that national courts and international arbitral tribunals might learn from each other.


International arbitral tribunals have no inherent lex fori. Therefore, they have no inbuilt predisposition toward any particular national law and no nationally generated choice of law regime. They begin the choice of law process from the expectations of the parties, rather than from the point of view of any national system of law. If the contract contains a clear choice of law clause, the tribunal will respect that choice. The rules of procedure of most arbitral institutions require the tribunal to apply any laws or rules of law chosen by the parties. (3) There are only two generally applicable exceptions: first, choices in fraudem legis, that is, made only to circumvent an undesired law, will not necessarily be enforced. Second, even when the parties have chosen a particular national law, some tribunals have also applied other national laws that claim extraterritorial effect. (4) By contrast, as Professor Walsh has explained with respect to Canada, national courts continue refuse to enforce choice of law clauses in a variety of situations.

The doctrine that parties to arbitrations have freedom to choose the governing substantive law is fairly recent. Under the traditional theory, sometimes called the "lex fori," "siege," or "jurisdictional" doctrine, by selecting a particular country as the site of an arbitration, the parties also impliedly chose that country's law to govern the merits of the dispute. This doctrine is now all but abandoned; however, as recently as 1968, English courts took the position that the choice of England as the site of arbitration constituted the choice of English law. (5) The lex fori doctrine found adherents among arbitral tribunals even into the 1990s. (6)

If the contract does not designate the governing substantive law, current arbitral rules empower the tribunal to select the choice of law rule it considers most appropriate, and then to apply that rule to determine the governing law. Moreover, the tribunal need not choose any particular choice of law and, indeed, need not even apply a choice of law rule. Instead, most arbitral rules of procedure permit the tribunal to designate an applicable law directly, a doctrine usually called voie directe. (7) For example, the Rules of Arbitration of the International Chamber of Commerce ("ICC Rules") permit the tribunal, in the absence of a choice by the parties, to "apply the rules of law which it determines to be appropriate." (8) This power is regularly exercised and, when exercised, will not render the award unenforceable in national courts. (9) In short, arbitral tribunals will enforce a contractual choice of substantive law and, if no clear choice is made, often avoid the (allegedly) complex and uncertain choice of law process by applying voie directe.

Voie directe gives an arbitral tribunal "an almost unlimited freedom in the choice of rules of substantive law." (10) In practice, however, voie directe most often "consists in finding a significant connecting factor between the contract and the law which the arbitrator decides to apply." (11) Accordingly, choice of law under voie directe frequently yields the same law that application of a country's choice of law rules would provide. However, vole directe has also been employed in a manner designed to give voice to party autonomy in cases where the parties have not agreed on an applicable law. A 1984 ICC award provides an example. (12) There, the contract did not specify the governing law, and the law of Switzerland and of another country were the only laws connected to the contract. The tribunal directly applied Swiss law, on the ground that:

... the law of country X might partially or totally affect the validity of the agreement. It is then reasonable to assume that from two possible laws, the parties would choose the law which would uphold the validity of the agreement.... In these circumstances, the arbitrators definitely decided to choose Swiss law as the applicable law, assuming that this choice corresponds to what the parties had in mind. (13) The tribunal used its power to make a direct choice of the governing substantive law in order to enforce the parties' reasonable expectations at the time that they entered into the contract. (Of course, one of the parties might well have wished to have the contract declared invalid after the dispute arose.) In this way, choice of law under voie directe is more likely than application of choice of law rules to accord with the parties' presumed intentions.

Party autonomy manifests itself in arbitral choice of law in another important way. Rules of law are frequently applied in arbitration that are either unavailable or unlikely to be applied in national courts. In the context of international private law, "rules of law" must be distinguished from "laws." The latter means the laws applicable in a state, whatever their source--statute, case law, or treaty. Rules of law, on the other hand, are any statements of principle that do not have the force of law in any state. (14) Rules of law cannot be applied in any national court, but are regularly applied as the governing law in arbitrations. The two most prominent sources of rules of law are international contract law instruments and "general principles" of international private law.

International contract law instruments take one of two forms: conventions that mimic domestic statutes (15) or compilations of principles in the manner of the American Restatements. (16) The conventions constitute "rules of law" until they are ratified, whereupon they become "laws" and are equally applicable in state courts and in international arbitrations. Parties may choose to be governed by one of the conventions by including an express term in their contracts applying it, regardless of whether they are domiciled in states that have ratified the convention. More commonly, however, the conventions apply when the parties are domiciled in different states that have each ratified the convention and choice of law analysis leads to the law of one of those states; by ratifying the convention, each state agrees that it will apply to contracts between parties from that state and parties from another state party. (17)

The most prominent example of this type of international instrument is the U.N. Convention on Contracts for the International Sale of Goods ("CISG"), (18) a convention that applies to all contracts for the sale of goods concluded between parties situated in different signatory states. There are now 74 states party to the CISG...

To continue reading