Edited by Jürgen Basedow, Giesela Rühl, Franco Ferrari and Pedro de Miguel Asensio
Chapter A.13: Arbitration, international commercial
I. The role of states and of domestic law in international commercial arbitration
Arbitration is an alternative to litigation before domestic courts for the settlement of private disputes, which is subject to the consent of the litigants. In relation to international commercial disputes, the principal advantage of arbitration is its neutrality compared to courts, sometimes viewed by international litigants as parochial or even biased. Another advantage is that arbitration offers litigants a broad freedom to designate the applicable procedural and substantive rules, thereby avoiding the application of national rules and procedures. These are often considered ill-suited to the settlement of transnational disputes.
Largely because of these and other perceived advantages, arbitration is the most widely used dispute settlement mechanism for international commercial transactions, to the point that it can almost be considered a transnational system of justice. According to one commentator, it should even be considered the default mechanism for such transactions (Gilles Cuniberti, ‘Beyond Contract: The Case for Default Arbitration in International Commercial Disputes’ (2008) 32 Fordham Int’l L.J. 417). One of the main factors behind the diffusion of international commercial arbitration is the near universal acceptance of the New York Convention (New York Convention of 10 June 1958 on the recognition and enforcement of foreign arbitral awards, 330 UNTS 3), now in force in more than 150 states. The Convention lays down two fundamental obligations, to uphold arbitration agreements (art II) and to enforce arbitral awards (art III). The combination of these obligations puts international arbitration on a par with litigation before national courts, with the added advantage that by virtue of the Convention awards can circulate internationally, unlike judgments.
The New York Convention’s pro-arbitration bias has unleashed a competition amongst national legal systems. As a result of this, and of the influence of the UNCITRAL Arbitration Model Law (United Nations Commission on International Trade Law, UNCITRAL Model Law on International Commercial Arbitration as adopted on 21 June 1985, and as amended on 7 July 2006, UN Doc A/40/17 and A/61/17; →Arbitration, (UNCITRAL) Model Law), the past decades have witnessed a surge in the pro-arbitration attitude of states and a convergence, and in some respects almost a de facto harmonization, amongst national arbitration laws. One of the main features of these laws is the broad recognition of the role of →party autonomy, which enables the parties to arbitration to derogate from national law in significant measure. The upshot is that arbitration is one of the areas of law where party autonomy is capable of playing the greatest role and where state influence has receded most conspicuously. In parallel to these developments there has also emerged a body of rules, mostly procedural, but also substantive, of non-state origin which can apply in international arbitration as an alternative or a supplement to national law.
Because of these developments, international arbitration might seem to have become largely detached from and independent of national legal systems and able to operate without regard to them. Some scholars postulate the existence of an international arbitral legal system, which would constitute the framework of international commercial arbitration (Emmanuel Gaillard, Legal Theory of International Arbitration (Martinus Nijhoff 2010)). Under such a vision, →choice of law would have little or no role in arbitration, and an analysis of international commercial arbitration from a choice-of-law perspective would be almost pointless.
Despite its theoretical appeal and the undeniable role of non-state rules (see Stavros L Brekoulakis, ‘International Arbitration Scholarship and the Concept of Arbitration Law’ (2013) 36 Fordham Int’l L.J. 745), that theory does not withstand empirical scrutiny, since in fact arbitration is not detached from national legal systems. Even if it rests on party autonomy, in the sense that it can be resorted to only if the litigants have consented to it, arbitration ultimately can exist and function effectively provided it is tolerated and supported by states. Parties cannot escape litigation before national courts, unless these forgo exercising their jurisdiction (→Jurisdiction, foundations) when arbitration agreements are contested. State authority is also indispensable to enforce awards not voluntarily complied with and to support dysfunctional arbitral proceedings. All domestic laws contain bodies of statutory or jurisprudential rules (generally referred to as arbitration laws) specifically applicable to arbitration, as well as p. 87other substantive and procedural rules which may aspire to be applied to arbitrations and which the parties may invoke.
Unlike state-to-state and investor–state arbitration, international commercial arbitration maintains substantial links with national law, which, together with party autonomy, provides the source of its authority (see Luca G Radicati di Brozolo, ‘The Impact of National Law and Courts on International Commercial Arbitration: Mythology, Physiology, Pathology, Remedies and Trends’  Paris J.Int’l Arb. 663).
II. Conflict of laws and of jurisdictions in international arbitration
The rules governing the multiple issues liable to arise in relation to an international arbitration – which can pertain to the merits of the dispute, to the conduct of the arbitral proceedings and more generally to the basis and functioning of the arbitration – may be contained in international conventions, different national laws and a composite body of non-national rules, usages and practices. Thus, for each international arbitration, which by definition has contacts with more than one law, the body of rules capable of providing the solution to each one of such issues must be identified. This poses a choice-of-law problem.
The impact of choice of law in international arbitration is only partially reduced by the convergence between national arbitration laws, the vastly increased scope of party autonomy, and the body of rules, usages and practices of varying nature and origin that may be applicable to arbitration. The convergence between national laws is far from complete, also due to the absence of any generally applicable international instrument harmonizing arbitration law. The UNCITRAL Model Law on International Commercial Arbitration is merely a model, adopted only by some states, and in any case with several variations. The New York Convention does not harmonize arbitration law comprehensively. Even with regard to the two obligations that it does lay down, its wording leaves broad scope for differing interpretations and there is no mechanism to ensure its uniform interpretation or application. →Party autonomy, despite its breadth, is not unlimited, since there are restrictions to the freedom of the parties to arbitrations to derogate from domestic procedural and substantive rules. In any event, party autonomy is not always exercised to its fullest extent. Also the rules of non-national origin, which can fill gaps left by national legal systems and the agreements of the parties or even provide alternative solutions, are applicable only insofar as permitted by states.
National law therefore remains the ultimate standard for the solution of all problems related to arbitration. It not only determines the extent to which party autonomy or non-national rules can be relied upon in a given case. It is relevant even when the applicable rules originate from an international convention, since conventions are not necessarily applied in the same way in all countries.
The role of national law is most prominent when matters relating to an arbitration are brought before the courts, particularly to determine the validity and effects of an arbitration agreement, to seek measures in support of the arbitration and in setting aside and enforcement proceedings. National law is also regularly resorted to by arbitrators for the solution of a variety of problems, not only substantive ones.
Thus, despite the very significant disenfranchisement of arbitration from national law, →choice of law retains a crucial place in international arbitration. A choice-of-law analysis is in principle always called for to identify the rules governing each specific issue relating to an arbitration.
Due to the possible involvement of courts in matters relating to it, arbitration also poses issues of conflict of jurisdictions (→Jurisdiction, foundations; →Jurisdiction, limits under international law), when it comes to identifying the national courts having authority to decide such matters.
The present contribution offers an overview of the private international law of arbitration. It cannot seek to analyse in detail the solutions to the scores of problems that arise. Indeed, despite a convergence in the approaches to conflict problems, an exhaustive analysis of each legal system and of non-national rules would still be required.
III. Sources and principles of conflicts of laws and jurisdiction in arbitration
1. Choice of law
The parameters for the choice of the governing law of the individual questions likely to arise in arbitration derive from a variety of sources and are for the most part not expressed in precise p. 88rules. Some rules are contained in international conventions, notably art V of the New York Convention; for the most part they derive from domestic legal systems, and a few derive from non-national sources, particularly the rules of arbitral institutions. Some principles must be deduced from systemic considerations.
The applicable choice-of-law framework varies depending on whether the issues arise before the courts or the arbitrators. Furthermore, the solution to specific issues may vary depending on the context in which they arise, for instance at the stage of the enforcement of the arbitration agreement, of the enforcement of the award, of a challenge to the award, during the arbitration, in proceedings for the support of the arbitration, and so on.
Courts, as always when faced with questions of private international law, apply the conflict principles and rules of their →lex fori. These may on occasion be influenced by international sources, notably art V of the New York Convention. Some conflict rules applied by courts are specific to arbitration (see for example art 178 of the Swiss Private International Law Act (Bundesgesetz über das Internationale Privatrecht of 18 December 1987, 1988 BBl I 5, as amended, henceforth Swiss PILA)). In relation to certain issues most systems adopt similar approaches, for instance the role of the seat (see IV.1.) and to the autonomy of the arbitration agreement also as to its governing law (see V.2.). In the absence of arbitration-specific conflict, rules courts apply ordinary conflict rules also in matters relating to arbitration.
Arbitrators, unlike courts, have no lex fori and, thus, no predetermined system of private international law. Initially it was held that arbitrators had to rely on the system of private international law of the seat of the arbitration and to apply the same conflict rules as courts. That position has been largely abandoned, although some commentators posit that, for instance, the conflict rules contained in the Rome I Regulation (Regulation (EC) No 593/2008 of the European Parliament and of the Council of 17 June 2008 on the law applicable to contractual obligations (Rome I),  OJ L 177/6) (→Rome Convention and Rome I Regulation) and the →Rome II Regulation (Regulation (EC) No 864/2007 of the European Parliament and of the Council of 11 July 2007 on the law applicable to non-contractual obligations (Rome II),  OJ L 199/40) must be applied in arbitrations seated in the European Union, because of the prevalence of European Union law over national law (→European Union and private international law). This position has no foundation, also because there is no indication that those regulations are intended to apply to arbitration proceedings. The legal systems of some states contain specific conflict rules applicable to arbitrations seated there, but such rules usually leave arbitrators a great freedom in the choice of the applicable law (see V.4.). Conflict rules for arbitrators are contained in many rules of arbitral institutions and these, too, accord a broad freedom to arbitrators (for instance arts 21 ICC Rules, 35(1) UNCITRAL Arbitration Rules (as revised in 2010, UNGA A/RES/65/22 and UN Doc A/RES/68/109, adopted on 6 December 2010), 22(3) LCIA Rules) (see Luca G. Radicati di Brozolo, ‘Party Autonomy and the Rules Governing the Merits of the Dispute in Commercial Arbitration’, in Franco Ferrari (ed), Limits to Party Autonomy in International Commercial Arbitration (Juris 2016), 331 ff).
In practice arbitrators adopt a variety of choice-of-law techniques. These include relying on the choice-of-law rules of the seat or of the legal system considered most appropriate or most closely connected to the matter at bar and applying the →choice of law principles common to the legal systems with relevant connections to the subject matter. Sometimes they follow the so-called ‘voie directe’, which consists in the direct application of the substantive rules considered appropriate, without resort to a choice-of-law analysis.
The arbitrators’ freedom to choose the applicable law is compounded by the fact that their decision on that point is usually considered a legal question. As such, in most legal systems it is not subject to court review in setting aside and enforcement proceedings. The only case in which an error by the arbitrators in the choice of the applicable law can conceivably be a ground for a challenge or refusal of enforcement (→Arbitration, recognition of awards) is when the arbitrators have disregarded an express designation of the governing law by the parties, thereby exceeding their mandate.
The paucity of clear and uniform conflict rules applicable to arbitration by courts and arbitrators breeds complexity, uncertainty, costs and potentially conflicting outcomes and is at odds with the aspiration of arbitration to be a transnational system of adjudication. The solution lies in the development of uniform approaches to choice-of-law problems and in the p. 89replacement of the classic approach of ‘bilateral’ choice-of-law rules in favour of the direct application of substantive rules specifically governing transnational situations. An example of these are the pro-arbitration rules developed by French case-law (→France) in relation to the validity of arbitration agreements.
An additional element of complexity is the absence of consensus as to how arbitrators are to deal with characterization (see Gabrielle Kaufmann-Kohler, La qualification en arbitrage commercial international, Travaux du Comité Français de droit international privé 2010–2012 (Pedone 2013) 299).
2. Conflict of jurisdictions
The jurisdiction of national courts in matters related to arbitration is determined, like for all other issues, by the →lex fori. There are no relevant international rules. Jurisdiction in relation to arbitration is in particular not governed by the Brussels I Regulation (recast) (Regulation (EU) No 1215/2012 of the European Parliament and of the Council of 12 December 2012 on jurisdiction and the recognition and enforcement of judgments in civil and commercial matters (recast),  OJ L 351/1; →Brussels I (Convention and Regulation)). Article 1(2)(d) explicitly excludes arbitration from the scope of the Regulation. The precise scope of the exclusion is clarified by Recital (12) of the Regulation, which must be read in conjunction with art 73(1), pursuant to which the Regulation does not affect the application of the New York Convention (see Trevor C Hartley, ‘The Brussels I Regulation and Arbitration’ (2014) 63 ICLQ 843).
The main jurisdictional connecting factor in arbitration is the seat of the arbitration (see IV. below). Under most contemporary legal systems, courts will generally exercise jurisdiction in ancillary proceedings (notably appointment, challenge and removal of arbitrators and assistance in taking of evidence) and in supervisory (ie setting aside) proceedings only in relation to arbitrations having their seat in the state of the court. Also →anti-suit injunctions in aid of arbitration can usually be applied for only at the seat.
There are exceptions. The courts of certain states exercise jurisdiction in setting aside actions also in relation to arbitrations seated in other states in the presence of certain →connecting factors, notably if one of the parties is a national of that state or the applicable substantive law of the dispute is the law of that state. This approach, followed in particular in →India but seemingly in the process of being abandoned in that country, is often criticized, although it is not manifestly contrary to any international rule.
Occasionally states are prepared to exercise their jurisdiction in aid of arbitrations seated elsewhere. This may be the case in relation to assistance in the taking of evidence (→Evidence, procurement of) to arbitral tribunals in proceedings seated abroad (for instance in the USA (→USA) under 28 US Code § 1782). Another example is art 1505 of the French Code of Civil Procedure ((Nouveau) Code de Procédure Civile of 1 January 1976, décret No 75–1123, 5 December 1975, henceforth French CCP) which allows the French courts (→France) to appoint an arbitrator in foreign arbitrations when the courts of the seat decline to do so. Courts are also generally prepared to grant interim and preliminary measures also in relation to arbitrations seated abroad, on the basis of their normal jurisdictional criteria (for instance presence of assets in the forum).
Disputes on the validity and effects of arbitration agreements, which usually arise from objections to jurisdiction in ordinary proceedings, are not subject to specific jurisdictional rules (→Arbitration, recognition of awards). According to the prevailing approach, proceedings for the recognition and enforcement of awards can be brought in any jurisdiction where the award creditor has an interest in obtaining recognition or enforcement. However, in some states such actions may be subject to jurisdictional conditions, such as the presence of assets or the existence of personal jurisdiction over the award debtor, or defences such as →forum non conveniens. Such conditions are arguably not permissible under the New York Convention.
Disputes between the parties and the arbitrators relating to the mandate of the latter are likewise subject to ordinary rules on jurisdiction.
IV. The law governing the arbitration
1. The law of the seat as the law governing the framework of arbitration
Like court litigation, arbitration is a process for the settlement of disputes through a binding decision. There is, however, a fundamental difference between the two processes, which relates to their legal basis and framework. Litigation before state courts is by its nature p. 90rooted exclusively in the law of one state, the forum, which governs all matters relating to the process. The only question subject to a different law is that of the effects of judgments in other countries, which are governed by the law of the state where recognition or enforcement (→Recognition and enforcement of judgments (civil law); →Recognition and enforcement of judgments (common law)) is sought.
Because it is in the first instance the product of →party autonomy, arbitration instead lacks a predetermined exclusive link to a single legal system. All questions relating to a given arbitration can be considered from the perspective of a multitude of legal systems. This is evident for international arbitration, by definition linked to more than one law. It is true also for arbitrations with links only to one state, since issues relating to them can nonetheless be submitted to the courts of other states.
Owing to its hybrid nature – dispute settlement mechanism, on a par with the adjudication by courts, and creature of contract – arbitration and in particular international arbitration could be assessed differently by different legal systems, depending on which aspect is considered prevailing. If prevalence is given to the first one, and arbitration is in some way assimilated to court litigation and considered an emanation of a given legal system, the tendency may be to link it to a single legal system. If, on the other hand, arbitration is considered essentially the product of private autonomy, like a contract, there is less justification for viewing it as exclusively or prevailingly linked to any one legal system as its only legal basis and there is greater scope for a variegated choice-of-law approach.
Traditionally the focus was placed on arbitration’s quasi-jurisdictional function and each arbitral proceeding was considered linked to a given national law which would provide the framework for it, rather like the law of the forum for litigation before state courts. In keeping with that approach, in most modern legal systems there is an identical (usually unwritten) conflict rule which considers the law of the seat, generally referred to as lex arbitri, as the one that provides the basis and the framework of arbitration.
The seat of the arbitration does not indicate the physical and geographic place where the activities related to the arbitration take place (hearings, deliberations, issuance of the award etc). Rather, it is a legal fiction in the sense that it designates the ‘legal home’ of the arbitration. In technical terms, the seat is the connecting factor between an arbitration and a given legal system considered as its governing law.
The role of the seat as the principal connecting factor to determine the law governing an arbitration is acknowledged also by the New York Convention, in art V(1)(d) and (e). In the latter provision the Convention also recognizes the possible relevance of the law of the country ‘under which’ the award was made. That alternative addressed the, now uncommon, situation where, instead of, or in addition to, designating the seat, the parties designated a national procedural law to govern the arbitration (→Proceedings, law governing).
The seat of the arbitration is also the prevailing jurisdictional →connecting factor to identify the courts having jurisdiction to deal with most matters relating to an arbitration (see III.2.). Consequently, for a large number of issues relating to an arbitration there is a coincidence between forum and applicable law. In relation to them states will thus exercise jurisdiction only over arbitrations they consider seated on their territory and to those they will apply their own law.
The seat determines what is often referred to as the ‘nationality’ of the arbitration, as well as of the resulting award, and, therefore, which arbitrations and awards are considered domestic (and thereby governed by the local law) and foreign from the standpoint of each legal system, including for the purposes of the New York Convention. When referred to an arbitration, nationality is merely shorthand for the lex arbitri.
2. The location of the seat
Under most legal systems the seat is determined in the first instance by party autonomy, either directly or through the mechanisms foreseen by the arbitration rules they may have opted for. The designation by the parties is dispositive, regardless of a significant connection between the selected seat and the arbitration. The role of the seat in determining the arbitration’s legal regime, combined with the parties’ freedom to select the seat, entails in practice that the parties are under no restriction in the selection of the law governing the arbitration, ie the lex arbitri. The situation is thus similar to the one of contracts, the governing law of which is determined by the parties. The only difference is that, whilst p. 91for contracts the choice of the law by the parties is straightforward, for arbitration it is indirect because the parties designate the seat and it is the seat that defines the governing law.
The existence in practically all legal systems of an identical conflict rule designating the governing law of an arbitration by means of the same connecting factor, the seat, in turn designated by the will of the parties, entails that the governing law of a given arbitration will generally be the same from the perspective of all legal systems. This greatly reduces →choice of law, at least in this respect.
Problems can arise where the parties do not (directly or indirectly) designate the seat or the designation is ambiguous. The laws of most states contain default criteria to establish whether in such cases the arbitration is to be considered seated on their own territory. This can lead to conflicts if a given arbitration is not considered by any state to be seated on its territory according to its criteria, or if more than one state considers it seated there. There are no international criteria to resolve such conflicts.
3. The scope of the lex arbitri
Although its significance has diminished over the years, the lex arbitri continues to play a very significant role. It is the law that is applied by the courts of the seat, and often by the arbitrators, at least as a starting point or as a fallback, to decide the majority of issues relating to an arbitration.
The lex arbitri determines, in the first instance, whether arbitration can be resorted to in relation to a given dispute. Although arbitrability may be assessed under other laws (see V.1.), inarbitrability under the lex arbitri is an almost insuperable impediment to the holding of the arbitration.
The lex arbitri also governs the creation of the arbitral tribunal and the qualifications, the appointment, removal and replacement of arbitrators, the powers and duties of the arbitrators, the rules of procedure before the arbitrators, the language of the arbitration, the qualifications that may be required of the parties’ counsel, the time limits for issuing the award and the formal requirements (→Formal requirements and validity) of the award. The law of the seat is often also considered applicable to a broad range of questions that are not strictly of procedure, but which concern the overall functioning of the arbitral process. Examples are the issue of the ascertainment of the content of the applicable law, the principles of res judicata, the powers of the arbitrators, the admissibility of dissenting opinions, the relief that can be granted by the arbitrators, the power of arbitrators to grant interim relief, the joinder of parties and the consolidation of different arbitral proceedings, the apportionment of the costs of the arbitration between the parties, and possibly whether interest is due on amounts owed and the relevant interest rate, the confidential or public nature of the arbitration, the treatment of third party funding, the waiver of the right to arbitrate, the rules on privilege. The justification for a blanket application of the law of the seat to this type of question is open to debate. In part this solution is based on an implicit assimilation of the lex arbitri to the lex fori in litigation, in part it can be explained by the lack of obvious alternative connecting factors to other laws. In relation to questions of this type a conflict-o-laws approach that seeks solutions in national law, including the law of the seat, is unsatisfactory. Indeed, most national laws, including those of many sophisticated jurisdictions, lack arbitration-specific rules for these problems. The application by analogy of the rules applicable before domestic courts will often lead to results not tailored to international arbitration and at odds with the goal of harmonization of the ground rules of arbitration which is essential if arbitration is to perform its role of transnational system of justice. The solution would be the development of harmonized arbitration-specific rules, such as those on evidence developed by the International Bar Association’s Rules on the Taking of Evidence in International Arbitration (see Luca G. Radicati di Brozolo, ‘Res judicata’, in Pierre Tercier (ed), Post Award Issues (Jurisnet 2011), 127 ff.). In principle, the lex arbitri also provides the choice-of-law rules to be applied by the arbitrators, although today most arbitration laws do not impose specific conflict rules on arbitrators (see III.1.).
Finally the lex arbitri governs the grounds for the challenge, revision and revocation of awards (see V.5.), as well as the possibility to exclude challenges altogether and to reduce or expand the grounds for challenge. Specifically it governs the extent to which the merits of the award can be reviewed (→Arbitration, recognition of awards), in particular for errors in the application of the law, and the consequences of p. 92setting aside, as well as the effects of the award, including, albeit only to a certain extent, its res judicata and preclusive effect.
Very importantly, the lex arbitri also defines the boundaries of the autonomy of the parties and therefore of their right to fashion the arbitration as they choose. Today under most arbitration laws that autonomy is very far-reaching. This is because most national rules governing arbitration, in particular those dealing with the matters mentioned above, are not mandatory, but are designed to apply only where the parties have failed to govern a matter themselves, either directly or by reference to arbitration rules. Accordingly, today the role of the lex arbitri is relatively residual, since it is limited essentially to filling gaps in relation to matters not dealt with by the parties.
The broad scope of →party autonomy under most arbitration laws, coupled with the parties’ freedom to choose the seat, and consequently the most favourable lex arbitri, entails that the parties enjoy extensive freedom in defining the rules of their arbitration. However, since that autonomy is not unlimited, varies from one legal system to another and ultimately depends on how the lex arbitri is applied by the courts of the seat, the roles of that law and of those courts can never be completely discounted.
4. The limits of the lex arbitri
The lex arbitri does not have an exclusive role in matters related to arbitration. Courts, including those of the seat, and arbitrators frequently apply other laws to certain issues on the strength of a choice-of-law analysis (see V.). However, because of the enduring resilience of the choice-of-law and choice-of-jurisdiction rules which emphasize the role of the seat, the lex arbitri and the courts of the seat still retain a pivotal role, albeit tempered by the role of →party autonomy.
This may ensure uniformity and predictability of solutions. Yet, it creates a strong link between arbitration and a national legal system, treating arbitration and awards as if they were products of the law of the seat and assimilating them to court litigation. It also confers on the courts and on the law of the seat a de facto quasi-monopoly on matters relating to the arbitration, and ultimately on its fate, even where the seat’s standards are less arbitration-friendly than those of other countries and the commonly accepted ones. Although the role of the seat is defused for the reasons mentioned above, this is antithetical to the conception of arbitration as a creature of the will of the parties and to the widespread expectation that it operate as a true transnational system of justice, detached from states to the greatest possible extent.
The question is therefore whether it is appropriate to defer almost by default to the law and decisions of the seat. Indeed, contrary to the position of some commentators, neither under the New York Convention nor under any other legal basis must the seat of the arbitration be considered to have ‘primary jurisdiction’ over arbitrations seated there, requiring other states to defer to its law and decisions (for contrasting positions see M Reisman and B Richardson, ‘Tribunals and Courts: An Interpretation of the Architecture of International Commercial Arbitration’ (2012) 16 ICCA Congress Series 17; Luca G Radicati di Brozolo, ‘The Control System of Arbitral Awards: A Pro-arbitration Critique of Michael Reisman’s “Normative Architecture of International Commercial Arbitration”’ (2012) 16 ICCA Congress Series 74–102).
Today it is unrealistic to postulate that national courts and arbitrators should disregard national law completely when dealing with arbitration. The main reason is that there does not exist a sufficiently comprehensive and universally accepted body of non-national rules capable of dealing with all the issues that can arise in relation to a given arbitration. From this it does not follow that the lex arbitri should be deferred to as passively as it often is. This is true even where the lex arbitri is designated by the parties. Indeed, that designation may be due to many reasons, including precisely the persistence of the role of the law of the seat and the consequent desire to avoid the uncertainties stemming from a failure to choose the seat. It does not necessarily indicate unreserved acceptance by the parties of the standards of that law and of the way they are applied by local courts. Since the role of the lex arbitri is not consecrated by international law and there is no intrinsic link between it and arbitrations seated there, when assessing arbitrations, awards and all related matters, courts and arbitrators are free to apply their own parameters. The situation is not different from when they deal with the other prototypical instrument of party autonomy, contracts, which likewise have no predetermined and objective legal framework and can be assessed by each legal system according to its own parameters. As regards arbitration, the p. 93parameters can be drawn from other national laws than the one of the seat. Courts may be inclined to apply those of their own law if they consider them more arbitration-friendly that those of the lex arbitri. A proper consideration of the role and nature of international arbitration may also lead to abandoning where possible a strict choice-of-law approach and to the elaboration of ad hoc rules eventually capable of more general acceptance and application.
V. Specific choice-of-law issues
The need to identify the applicable law arises for almost any matter relating to an international arbitration. This section outlines some of the more common approaches to recurrent choice-of-law issues, bearing in mind that there are significant variations amongst legal systems and that for the most part the solutions are not encapsulated in written rules or well defined jurisprudential trends.
The preliminary issue is whether a dispute contemplated by an arbitration agreement can be decided by arbitration. In most legal systems the scope of arbitrability is now very broad and usually encompasses all disputes involving commercial relations, economic interests and property (→Property and proprietary rights), even if implicating mandatory rules. Yet, the precise contours of arbitrability vary amongst legal systems, thereby potentially raising choice-of-law issues.
Lack of arbitrability may be a ground for an objection to jurisdiction before a court or the arbitrators, a challenge of the award and refusal of recognition and enforcement (→Arbitration, recognition of awards), particularly under art V(2)(a) of the New York Convention.
When raised to contest the jurisdiction of a national court, arbitrability will generally be assessed in the first instance under the →lex fori (see art VI(2) of the European Arbitration Convention (European Convention of 21 April 1961 on International Commercial Arbitration, 484 UNTS 349)), particularly if the preservation of the exclusive jurisdiction of the local courts is at stake (→Jurisdiction, foundations; →Jurisdiction, limits under international law), or under the law governing the arbitration agreement. Article II(3) of the New York Convention does not specify the law under which arbitrability is to be assessed by courts confronted with an objection to jurisdiction. When an objection to arbitral jurisdiction is raised before the arbitrators, these will usually take into consideration alternatively the law governing the arbitration agreement (see the following subsection) or the law of the seat or transnational rules favouring a very broad notion of arbitrability.
In setting aside actions arbitrability will generally be assessed according to the lex arbitri, although the justification for this approach is not cogent where the dispute is not otherwise connected to the seat and would not fall under the exclusive jurisdiction of the local courts. In recognition and enforcement proceedings the applicable law to determine arbitrability under art V(2)(a) of the New York Convention is the one of the requested court. This solution attracts the same objections as the application of the law of the seat in setting aside actions (see Stavros Brekoulakis, ‘Arbitrability and Conflict of Jurisdictions: The (Diminishing) Relevance of lex fori and lex loci arbitri’ in Franco Ferrari (ed), Conflict of Laws in International Arbitration (Sellier 2010) 117ff).
The prevailing approach today is to accord no relevance to the limitations to arbitrability contained in the national law of the parties, including states and state entities (eg art 177(2) Swiss PILA).
2. The arbitration agreement
According to the prevailing view on the ‘severability’ of the arbitration agreement, such agreements are considered self-standing, also from the standpoint of their governing law, even when they are contained in the contract which governs the merits of the dispute. Some legal systems contain specific conflict rules on the substantive validity of arbitration agreements (eg art 178(2) Swiss PILA). French law (→France) shuns a choice-of-law analysis, applying, instead, a substantive rule applicable to all arbitration agreements which relies on international public policy (règle de droit international privé materiel). In other legal systems the conflict rule for the determination of the governing law of such agreements is the one that applies to contracts in general, although the Rome I Regulation is not applicable to arbitration agreements by virtue of its art 1(2)(e). This means that the arbitration agreement will be governed by the law chosen by the parties, if p. 94such a choice has been made. Since an express choice is usually lacking, resort is to be had to the other →connecting factors applicable to contracts, particularly the tacit or implied choice and the closest connection. This leaves considerable discretion. Most frequently, those indicators are held to lead to the law of the contract governing the merits or to the law of the seat. This solution is in line with art V(1)(a) of the New York Convention, which refers to the law chosen by the parties, or failing a choice, to the lex arbitri for the determination of the validity of the arbitration agreement when this is raised to contest recognition or enforcement of an award. The same rule is contained in art VI(2) of the European Arbitration Convention, which adds that, if neither of those laws can be determined, the law governing the arbitration agreement falls to be determined ‘under the competent law by virtue of the rules of conflict of the court seized of the dispute’. Because the results of these approaches to the choice of the applicable law of the arbitration agreement are often unsatisfactory, courts and arbitrators increasingly embrace the ‘validation principle’. This focuses principally on the purposes of such agreements and on the need to give effect to the parties’ commercial interests and the New York Convention’s pro-arbitration regime (see Gary B Born, International Commercial Arbitration (Kluwer Law International 2014) 542ff).
The range of issues potentially governed by the law so identified is broad, although courts and arbitrators may sometimes apply other laws to some of those issues. The issues include the formation of the arbitration agreement, its existence, its material validity, interpretation, its positive and negative effects, the personal scope of the agreement and its applicability to non-signatories, the scope of the submission to arbitration, its termination, expiration and assignability (→Assignability/Assignment of claims), the waiver of right to arbitrate. That law does not govern matters such as the capacity of parties (→Capacity and emancipation), usually governed by their respective personal law (see for example art VI(2) of the European Convention) and the form of the arbitration agreement (see eg art 178(1) of the Swiss PILA and art II of the New York Convention which lays down a maximum standard of formal validity).
The arbitration agreement must be distinguished from the agreement between the arbitrators and the parties. This agreement governs in particular the mandate of the arbitrators and their rights and obligations towards the parties, including the standards of their liability and their →‘immunity’ from liability for their decisions and their liability for conflicts of interests or failure to disclose matters relevant to these. Under some laws the liability of arbitrators is based on tort (→Torts), in which case the governing law falls to be identified by the conflict rule on non-contractual liability. The agreement between the parties and the arbitrators, frequently not contained in a formal document, is subject to its own governing law to be identified according to the ordinary conflict rules for contracts. These include the Rome I Regulation, because that agreement is not covered by the exception of art 1(2)(e). Since usually there is no express choice and a tacit choice is difficult to discern, recourse must be had to the subsidiary connecting factors. In case of a sole arbitrator, the law of the latter’s habitual residence could be applicable (notably under art 4(1)(b) of the Rome I Regulation). That solution is unworkable when there is more than one arbitrator, unless the relations with each arbitrator are held to be governed by their individual law. However, this too is unsatisfactory because it potentially entails the subjection of the rights and obligations of the members of the tribunal to different laws. Thus, in most circumstances, including pursuant to art 4(3) and (4) of the Rome I Regulation, recourse will have to be had to the closest connection. In most cases this will probably be with the lex arbitri, which, being the law governing most other matters relating to the arbitration and to the arbitrators (see IV. above), avoids the need to parse the overall relationship to identify the law governing different aspects. Alternatively, the closest connection could be with the law governing the arbitration agreement or, in the case of institutional arbitration, with the law of the seat of the arbitral institution.
A further relationship, which is governed by its own applicable law, is the one between the arbitrators and the parties and the arbitral institution. In this case a significant connection could be held to exist with the seat of the arbitral institution.
3. The arbitral proceedings
National law nowadays plays a relatively marginal role in relation to the conduct of the proceedings before arbitral tribunals. The prevailing assumption is that procedure – in particular the form and sequence of written p. 95and oral pleadings, the taking of evidence (→Evidence, procurement of), the language, the powers of the arbitrators to decide specific matters and to organize the proceedings and to adopt procedural decisions and so on – is governed by the law of the seat, which governs the majority of issues relating more broadly to the arbitral process (see IV.).
However, most laws leave almost complete freedom to the parties and, absent agreement between them, to the arbitrators. The only limitations are due process and equality of the parties, which are usually not interpreted with reference to the peculiarities of local law. More generally, the rules of the lex arbitri governing court proceedings are not considered applicable to arbitral proceedings, unless specifically referred to by the parties.
The freedom to determine the rules governing the proceedings means that the parties are free to apply to the procedure a national law of a state other than that of the seat. In practice, parties very seldom follow this approach, which is at odds with the trend to avoid the application of national procedural rules to arbitral proceedings.
In matters of procedure an increasing role is played by non-national rules, particularly the rules of arbitral institutions and those promoted by institutions, such as the International Bar Association, which constitute an embryo of transnational law of arbitration procedure widely followed in international arbitration. Sometimes, however, such rules refer back to national law, as is the case of art 9.2(b) of the International Bar Association’s Rules on the Taking of Evidence in International Arbitration, which allows the exclusion of evidence for reasons of ‘legal impediment or privilege under the legal … rules determined by the Arbitral Tribunal to be applicable’.
There are other matters, broadly speaking related to the conduct of arbitral proceedings, which may be affected, at least in part, by other laws. These include whether information relating to the arbitration or introduced into it can or must be disclosed, including legal and similar privilege, the duties and ethical obligations of counsel. These matters may be governed respectively by the law of the place where the information is held or to which the parties for which the information is relevant are subject, and by the law of the country to which counsel is subject. Also in relation to these situations a strong case can be made for the development of arbitration-specific harmonized rules, given the unsatisfactory results of the recourse to national law via conflict-of-law rules (see IV.3).
4. The merits of the dispute
The rules applicable to the merits, ie the ones to be applied by the arbitrators to the decision of the dispute, are the subject of one of the few explicit conflict rules contained in most arbitration laws and rules of arbitration institutions (Linda Silberman and Franco Ferrari, ‘Getting to the Law Applicable to the Merits in the International Arbitration and the Consequences of Getting It Wrong’ in Franco Ferrari (ed), Conflict of Laws in International Arbitration (Sellier 2010) 257ff). According to those rules, the primary connecting factor is the choice of the parties (see art 28(1) UNCITRAL Arbitration Model Law, art 187(1) Swiss PILA, art 1511 of the French CCP, s 46(1) of the English Arbitration Act 1996 (c 23), art 21(1) ICC Rules, art 35(1) UNCITRAL Rules; see also art VII European Arbitration Convention). Unlike the conflict rules on contracts applicable in litigation before state courts, under most of those rules the parties are free to choose non-national rules (for instance general principles of law, the →lex mercatoria and the UNIDROIT Principles (International Institute for the Unification of Private Law/ Institut international pour l’unification du droit privé, UNIDROIT Principles of International Commercial Contracts 2010 (3rd edn, UNIDROIT 2010))) to the exclusion of any national law. In the absence of an express choice, the individual conflict rules use different →connecting factors, such as the closest connection, or approaches, such as leaving the arbitrators to apply the rules deemed to be most appropriate (art 21(1) ICC Rules, art 1511 French CCP). In all cases the arbitrators have a broad freedom in determining the applicable rules. One question is whether, in the absence of a selection of the applicable law or rules by the parties, the arbitrators must apply a national law or are permitted to apply non-state law to the exclusion of a national law. There seems to be no reason of principle why this should not be possible, save in cases where it is clear that the parties had in mind only the application of a national law. Another point is whether, in applying national law, arbitrators are bound to apply it rigidly, in the same way as national courts, or can apply it more flexibly, taking into p. 96consideration other elements, including the explicit agreement of the parties embodied in the contract, even where it is contrary to ‘simple’ mandatory rules. Given the absence of any obligation to apply national law to the merits in arbitration, there is no reason in principle why arbitrators should not be able to apply national law as they consider most in keeping with the expectations of the parties. Some rules (see eg art VII of the European Arbitration Convention and art 21(2) of the ICC Rules) explicitly provide for the application of the provisions of the contract and of trade usages, which may be interpreted as giving a prevalence to these sources over national mandatory rules (see Luca G. Radicati di Brozolo, ‘Party Autonomy and the Rules Governing the Merits of the Dispute in Commercial Arbitration’, in Franco Ferrari (ed), Limits to Party Autonomy in International Commercial Arbitration (Juris 2016), 331 ff); Luca G. Radicati di Brozolo, ‘Ethics and law applicable to the merits in arbitration’ in Mélanges en l’honneur du Professeur Pierre Mayer (LGDJ 2015), 754 ff.).
The arbitrators have to resort to a choice-of-law approach also to identify the law applicable to issues traditionally falling outside the domain of the lex contractus. In so doing they may have to determine whether a →choice of law by the parties may operate (if so intended by the parties) also for matters for which traditional choice of law does not permit the parties to designate the applicable law. Also in this case there is no reason of principle why the arbitrators should not defer to →party autonomy.
It is now generally acknowledged that, in deciding the merits, arbitrators must apply overriding mandatory rules (→Overriding mandatory provisions) (competition law, anti-corruption, →embargoes, sanctions, protection of financial markets and natural heritage and so on), since arbitration cannot be a means to circumvent such rules. The criteria for the application of overriding mandatory rules by arbitrators are not laid down by any explicit rules. Since mandatory rules have their own sphere of application, neither those of the lex arbitri nor those of the lex causae, even if chosen by the parties, are applicable merely by virtue of their being part of such laws. Rather, the correct approach seems to be that arbitrators should apply, or take into consideration, those mandatory rules, whatever legal system they belong to, that have a particular entitlement to apply to the dispute, taking into consideration the nature of the dispute and its link to the state from which the rules emanate, and whether such rules would have had a realistic prospect of being applied absent the agreement to arbitrate (see Luca G Radicati di Brozolo, ‘Mandatory Rules and International Arbitration’  Am.Rev.Int’l Arb. 65–72; Luca G. Radicati di Brozolo, ‘When, Why and How Must Arbitrators Apply Mandatory Rules: the Problem and A Proposal’, in Franco Ferrari (ed), Arbitration and EU Law (Juris, 2017, forthcoming)).
It is sometimes said that, in deciding the merits, arbitrators must respect ‘truly international public policy’. This is defined as a set of generally recognized core principles and values that are less parochial that the →public policy standards applied by courts in reviewing awards. Often these principles will be embodied in the overriding mandatory rules that are applicable according to the standard mentioned above.
5. Setting aside and enforcement
In setting aside proceedings (→Arbitration, recognition of awards) courts will for the most part apply their →lex fori (which in such cases will be the lex arbitri), in particular as to the grounds of annulment. In deciding on certain grounds of annulment they may, however, have regard to other laws, such as for instance to decide on the validity of the arbitration agreement and on the capacity of the parties to enter into it (see art IX of the European Arbitration Convention; →Capacity and emancipation). As mentioned above, at the annulment stage there is little scope for a conflict of laws analysis regarding the law applied by the arbitrators to decide the merits, because this is a legal question generally outside the scope of judicial review of awards.
As to the recognition and enforcement of foreign awards, the procedure is governed by the law of the recognizing state (art III of the New York Convention). However, the permissible grounds for refusal of recognition are laid down by art V of the New York Convention, which provides for the application of laws other than the one of the recognizing state for certain matters, as mentioned above (art V(1)(a) and (d)). The enforcing state is permitted to rely on its own law to decide whether the award deals with an arbitrable matter or violates public policy.
Luca G Radicati di Brozolo
Gary B Born, International Commercial Arbitration (Kluwer Law International 2014); Stavros Brekoulakis, ‘Arbitrability and Conflict of Jurisdictions: The (Diminishing) Relevance of lex fori and lex loci arbitri’ in Franco Ferrari (ed), Conflict of Laws in International Arbitration (Sellier 2010) 117;
Stavros L Brekoulakis, ‘International Arbitration Scholarship and the Concept of Arbitration Law’ (2013) 36 Fordham Int’l L.J. 745; Giuditta Cordero Moss, ‘Arbitration and Private International Law’ (2008) 11 IntALR 153;
Gilles Cuniberti, ‘Beyond Contract: The Case for Default Arbitration in International Commercial Disputes’ (2008) 32 Fordham Int’l L.J. 417;
Emmanuel Gaillard, Legal Theory of International Arbitration (Martinus Nijhoff 2010); Gabrielle Kaufmann-Kohler, La qualification en arbitrage commercial international, Travaux du Comité Français de droit international privé 2010–2012 (Pedone 2013) 299;
Jean-François Poudret and Sébastien Besson, Comparative Law of International Arbitration (Sweet & Maxwell 2007); Luca G Radicati di Brozolo, ‘The Impact of National Law and Courts on International Commercial Arbitration: Mythology, Physiology, Pathology, Remedies and Trends’  Paris J.Int’l Arb. 663; Luca G Radicati di Brozolo, ‘The Control System of Arbitral Awards: A Pro-arbitration Critique of Michael Reisman’s “Normative Architecture of International Commercial Arbitration”’ (2012) 16 ICCA Congress Series 74;
Luca G Radicati di Brozolo, ‘Mandatory Rules and International Arbitration’  Am.Rev.Int’l Arb. 65; M Reisman and B Richardson, ‘Tribunals and Courts: An Interpretation of the Architecture of International Commercial Arbitration’ (2012) 16 ICCA Congress Series 17;
Linda Silberman and Franco Ferrari, ‘Getting to the Law Applicable to the Merits in the International Arbitration and the Consequences of Getting It Wrong’ in Franco Ferrari (ed), Conflict of Laws in International Arbitration (Sellier 2010) 257.