THE ARBITRAL SITUS AND THE LEX LOCI ARBITRI*
There must be no Alsatia in England where the King’s writ does not run.
Lord Justice Scrutton1
A company that submits a controversy to arbitration may later regret having abandoned recourse to the courts. On the day of reckoning, the sage chosen to decide the dispute may no longer seem so wise to the losing party, and the loser might consider refusing to comply with the arbitrator’s decision. Some legal system, therefore, must legitimize the arbitrator’s authority. Otherwise, the award remains an unenforceable conciliation attempt that the parties are free to treat as mere foreplay to litigation.
The country where the award is rendered traditionally has legitimized arbitral authority subject to conditions, in the form of mandatory procedural rules imposed on the arbitral proceedings. The proper scope of these local curial norms—commonly known as the lex loci arbitri—has been problematic. In England, prior to 1979,2 national courts supervised arbitral proceedings to ensure legally correct results. According to another model, however, arbitration is subject to judicial control only to safeguard its fundamental fairness. Yet a third model would free arbitration from any constraints imposed by the legal system of the place where the award is rendered.
These alternative patterns for judicial intervention implicate competing values that do not yield to facile analysis. The commercial community desires finality in private dispute resolution. Yet, national judicial systems may wish to further rival goals, such as the integrity of the adjudicatory process and respect for the rights of third parties.
Scholars and practitioners alike have questioned the traditional view that the lex loci arbitri governs the validity of an arbitration. Proponents of “denationalized” or “floating” arbitration assert that arbitral awards may be detached from the law of the country of the proceedings and yet remain enforceable.3
The first part of this chapter explores the control normally exercised over international commercial arbitration by the place of the proceedings. The second part examines the way that English arbitration law encapsulates the struggle to reconcile the rival goals of fairness or finality in private dispute resolution. The author concludes that control of international commercial arbitration by the lex loci arbitri commends itself if the local judge limits his role to ensuring the integrity of the arbitral proceedings and respect for the interests of third parties.
In England prior to 1979 the High Court could force an arbitrator to submit a point of law for judicial determination under the so-called “special case” or “case stated” procedure. This practice was not introduced into Scottish law until 1972.4 Even thereafter, the parties to an arbitration in Scotland could exclude the procedure by pre-dispute agreement between the parties. Such “ouster” of judicial jurisdiction was not possible in England.5
In the case of James Miller v. Whitworth Street Estates6 an English company contracted with a Scottish construction company for work on a building in Scotland. English law governed the interpretation of the contract. Scotland, however, was the “seat” of the arbitration, where the proceedings were held. After evidence had been heard, the English company asked that the arbitrator “state a case” to the High Court. The arbitrator refused, since under the law of Scotland he was the final adjudicator of legal as well as factual questions. The House of Lords upheld the arbitrator’s refusal. The arbitration was governed by the law of Scotland, the place of the proceedings, notwithstanding that English law governed the contract.
Two assumptions about the interaction of national law and commercial arbitration inhere in this decision. The first is that arbitration is controlled by some national law, a lex arbitri.7 The lex arbitri is not necessarily the law governing the substance of the dispute, nor the procedural rules applied by the arbitrators. Rather, the lex arbitri governs the validity of the arbitral process itself.
The second assumption underlying Whitworth is that the law of the arbitration is the law of the place of the proceedings: the lex arbitri is the lex loci arbitri. Thus an arbitrator must bow to mandatory norms of the country in which he sits.8 Parties may choose the law governing the contract. They may even choose some of the procedural rules applied by the arbitrator to a matter such as cross-examination of witnesses. They do not, however, choose the law governing the arbitration, except indirectly through choice of its situs.
F.A. Mann has articulated this traditional view with force, arguing that the pronouncements of an arbitral tribunal are not binding unless linked to a specific system of national law: “Every right or power a private person enjoys is inexorably conferred by or derived from a system of municipal law which may conveniently and in accordance with tradition be called lex fori, though it would be more extact (but also less familiar) to speak of the lex arbitri.”9
The mandatory rules imposed by the lex loci arbitri do not yield to neat classification. Many legal systems prohibit arbitration of disputes involving sensitive public interests, such as the protection of investors in corporate securities10 or contracts with state agencies.11 Some require arbitrators to state the reasons for their awards,12 or provide for the removal of arbitrators who are inept13 or unfair.14 A few legal systems have provided for appeal from arbitrator error on matters of law.15
English judges traditionally have given the lex loci arbitri greater significance than their French or American brethren. One Court of Appeal decision held that the selection of London as a situs for arbitration implied that English substantive law governed the issue of contract damages.16 This decision contrasts with American17 and French18 cases that have minimized the influence of the law of the place of arbitration.
The traditional role of the lex loci arbitri has been questioned by scholars and practitioners who suggest that an international commercial arbitral award may “float” free from the constraints of the national law of the place of the proceedings.19 Such denationalized arbitration, producing an “anational” arbitral award,20 marries well with the commercial motive behind the trend toward greater arbitral autonomy in modern arbitration law to increase a country’s attractiveness as a situs for arbitral proceedings.21 This approach also permits national tribunals to concern themselves less with disputes not implicating national interests, and accommodates international business transactions in which the parties’ divergent nationalities create a special need for a neutral and private forum for dispute resolution.22
“Denationalized” arbitration, however, has come to mean more than self-restraint by the country of the proceedings in the imposition of local procedural law. “Floating” awards, it has been argued, are and should be enforceable outside the country of proceedings despite annulment where rendered. For example, suppose arbitration is conducted in Azania, where the award is set aside on the ground that the parties did not validly consent to arbitration. Should a court in Ruritania, where the loser has assets, nevertheless enforce the award if it comes to a different conclusion about the validity of the arbitration agreement? What if Azanian courts set aside the award for error of law, a ground for annulment unknown to Ruritanian law?
Varying degrees of floating arbitration might be contemplated. For instance, Ruritania might take the position that annulment in Azania is never relevant to enforcement in Ruritania. A less extreme position might take annulment of an award in Azania as an impediment to enforcement in Ruritania if the annulment was made for reasons considered appropriate under Ruritanian law. For example, Ruritania might accept annulment for arbitrator corruption, but not for arbitrator error of law.
Ruritanian courts, of course, may always deny recognition to an arbitration defective under their own standards. So much is accepted by both traditionalists and proponents of floating arbitration. The divisive issue is whether Ruritanian courts should also defer to Azanian nullification for violation of the latter’s procedural norms.
Critical to the viability of “anational” arbitration is the legal effect of an arbitral award not linked to a national legal system. The distinguished arbitration lawyer Jan Paulsson has written recently that “the binding force of an international award may be derived …without a specific legal system serving as its foundation.”23 Focusing on the French case of Gotaverken v. Libyan General National Maritime Transport Co.,24 Paulsson demonstrates that an international arbitral award may be enforced even if not subject by the lex loci arbitri to the same appeal procedures as a domestic award.
The arbitration in Gotaverken involved three tankers constructed by a Swedish shipyard for a Libyan government agency. The Libyans refused delivery, inter alia, on the ground that ship components had been made in Israel in violation of Libyan boycott law. The arbitrators ordered the Libyans to take delivery of the vessels and to pay the outstanding portion of the price. The Libyans challenged the award, arguing that it violated the public policy (ordre public) of France because of its failure to respect Libyan boycott law. The Paris Cour d’appel refused to hear the challenge on the ground that the award was not French. Such an award, Paulsson maintains, is enforceable “without a specific legal system serving as its foundation.”25
The paradox of a legal obligation independent of a legal order suggests Athena springing full-blown from the head of Zeus: a binding commitment, free from any municipal law, just appears. Grasp of Paulsson’s thesis requires a conceptual leap to a document labeled “obligation” enforced without respect to whether the document constitutes a valid obligation under the legal system normally selected by the enforcement forum’s choice of law principles. In other words, the document receives contractual force from the enforcement forum itself regardless of the otherwise governing law.
Gotaverken and its progeny lend themselves to a less radical analysis than suggested by Paulsson. An international arbitration, rather than being “detached from its country of origin,” receives substantially greater autonomy, and is subject to fewer constraints, than a domestic arbitration. The award in Gotaverken was not annulled. Rather, the French court found the award not subject to a challenge procedure (appel en nullité) available only to French awards. Moreover, if the facts of Gotaverken are changed slightly, so that the Swedes deliver cocaine rather than tankers, one wonders whether the French court would exhibit a similar laissez-faire attitude towards alleged public policy violations.
Several months after Gotaverken, the Paris Cour d’appel affirmed the vitality of the traditional role of the lex loci arbitri with respect to another international award. In Berardi v. Clair26 the balance sheet of a Gabonese company was at issue, its shares having been sold by a Canadian (Berardi) to a Frenchman (Clair). The arbitrator awarded the seller 23 million French francs, and the Paris Tribunal de grand instance granted leave to enforce the award in France. Three months later, the Geneva cantonal Cour de justice annulled the award as “arbitrary.” The Cour d’appel of Paris thereafter quashed the lower court decision, refusing recognition (exequatur) in France to an award set aside under the law of the place where rendered.
In both Gotaverken and Berardi v. Clair, the parties had stipulated arbitration under the Arbitration Rules of the International Chamber of Commerce, but with a significant difference: in Gotaverken the 1975 version of the Rules was applied, whereas in Berardi v. Clair the arbitration was subject to the 1955 version of the Rules. The latter provides for gaps in the procedural rules to be filled according to the law of the country in which the arbitrator holds the proceedings. The 1975 Rules, by contrast, provide for such gaps to be filled by the arbitrator. One may speculate about the enforceability of the award in Berardi v. Clair had it been rendered under the 1975 Rules. It would seem misleading, however, to suggest that such awards may be “detached” from the lex loci arbitri.
The present author knows of no award enforced after explicit annulment where rendered. Post-annulment recognition would, however, seem appropriate in two circumstances. The first would require that the arbitration arise under the European Convention on International Commercial Arbitration (commonly referred to as the Geneva Convention of 1961), covering disputes between nationals of different contracting States. Annulment of an award in its country of origin constitutes a ground for refusal of recognition under the European Convention only when annulment is for specifically enumerated reasons, such as the invalidity of the arbitral agreement or lack of proper notice to the parties. Thus if Canada had adhered to the European Convention, the award in Berardi v. Clair, discussed earlier, might have been enforceable in France, since “arbitrariness” is not a ground for annulment under the Convention.
The second situation in which one might approve enforcement of an award annulled in its country of origin is where the local judiciary is corrupt or biased. For example, a judge of a country without a tradition of judicial independence might set aside an award rendered against its own government merely to please the bureaucracy, without regard to the fairness of the proceedings. Enforcement of such an award by a country where the debtor government has assets would seem neither improper nor inappropriate.
The 1958 New York Arbitration Convention leaves open the theoretical possibility that the lex arbitri may be other than the law of the place of the proceedings. Under Article V(1)(e), a non-domestic arbitral award may be refused recognition on proof that it has been set aside by “a competent authority of the country in which, or under the law of which, that award was made.” Annulment by an authority of the country “under the law of which” an award is made could be construed as a supplementary, rather than substitute, ground for refusal of recognition.
Debate over the nationality of awards is more than an academic quibble. The critical importance of the lex loci arbitri was strikingly illustrated by the saga of Société Européenne d’Etudes et d’Entreprise (S.E.E.E.) v. Yugoslavia.27 Before the Second World War S.E.E.E. built a railroad in Yugoslavia for which it was paid in devalued French francs. A panel of two arbitrators found an implicit currency stabilization provision and rendered an ex parte arbitral award in favor of S.E.E.E. in Lausanne on 2 July 1956. In an action before the cantonal court of Vaud, Yugoslavia argued that an even number of arbitrators violated mandatory dispositions of local law. The Swiss court, reasoning that the arbitration was not governed by Vaud cantonal law, refused deposit of the award and returned it to the parties.
In a subsequent action in the Netherlands to enforce the award, the Dutch Supreme Court rejected the Yugoslav argument that the award was invalid because it was not linked to any national legal system.28 The victory of the “floating award,” however, was short-lived. In a second decision the Dutch Supreme Court held that the Swiss refusal to hear the Yugoslav challenge had the same consequence as annulment of the award, which therefore was refused recognition.29 The Dutch court considered the arbitrator’s decision to constitute a Swiss award notwithstanding the Swiss Court’s contrary view.30
A legal system that subjects international arbitration to its judicial control must decide which procedural norms are to be applicable. Judicial intervention might be limited to ascertaining the award’s basic integrity, breached by an arbitrator who is not impartial, or who decides matters that the parties never entrusted to his adjudication. On the other hand, the courts also might concern themselves with the legal merits of the dispute. The extent of judicial control will depend upon conclusions as to whether it is more important that an arbitral award be correct, or that it be final.
Prior to 1979, English law had erred on the side of legal certainty, and attempted to ensure that all disputes decided within the borders of England and Wales would be subject to a uniform application of legal principles. Although parties to a business dispute had agreed to give the responsibility for its resolution to an arbitrator, the arbitral process still could be vitiated whenever one party, fearful of losing, decided to break its agreement to settle the dispute privately, and to try its chance instead before a judge.