1. Is Arbitral Jurisprudence anything more than a myth?
2. How does persuasiveness of past awards operate?
3. Is Precedent the product of the intrinsic qualities of one or more particularly well-reasoned awards?
4. Why do arbitral awards need to be available?
5. Why is reliance on arbitral precedents not frequent?
6. Should all awards be published?
7. Should awards be published with the names of the arbitrators?
8. How could a mass publication of complete, unabridged awards be achieved?
9. Is confidentiality a valid objection to the publication of arbitration awards?
10. Is there really an overriding principle of confidentiality?
1. Is Arbitral Jurisprudence anything more than a myth?
At the core of the question raised in the title of this blog is the much debated question of the existence of precedent in international commercial arbitration, the answer to which is in turn dependent on different philosophical conceptions of that means of dispute resolution. Do international arbitrators apply the law chosen by the parties in the same way a national court would do? Or is international arbitration a free-standing system of international justice relying on a body of legal rules of its own? Are international arbitrators only concerned with the case before them, or do they feel compelled to adhere to past arbitral solutions for the sake of consistency? In sum, is arbitral jurisprudence anything more than a myth? (see on these issues the seminal lecture of G. Kauffmann-Kohler, Arbitral Precedent: Dream, Necessity of Excuse? The 2006 Freshfields Lecture, Arb. Intern. 2007, Vol. 23, n°3, p. 357). One point in respect to which almost all authors seem to agree is that “persuasive precedent”, rather than precedent in the meaning of the doctrine of stare decisis, is the concept that can be applied to arbitration. Persuasive precedent can be defined as the de facto tendency for an international arbitrator to accept what has been consistently decided in a significant number of past arbitral decisions. It would therefore be misconceived to apprehend the concept of precedent in arbitration with the same perspective as that applied to courts. The jurisprudence of state courts present characteristics of homogeneity in a hierarchical system that arbitral case law does not and cannot have. Yet, international commercial arbitration produces decisions which are not the product of a given municipal judicial system: they are autonomous decisions issued by tribunals which have no forum and which are not rooted in the judicial system of the seat of the arbitration (see the recent Putrabali decision of the French Supreme court of 29 June 2007 which defines international arbitration award as an “international decision of justice, Rev. Arb. 2007.507 note E. Gaillard). These decisions are referred to by other arbitrators, and they may in certain cases persuade future tribunal to adhere to previous solutions. Arbitral precedent is no more and no less than this capacity of past arbitration awards to convince future tribunals to adhere to the solution they embody. The proper question should therefore not be whether arbitral precedent exists, but how and when it does operate.
2. How does persuasiveness of past awards operate?
The persuasiveness of past arbitration awards implies to a certain extent that international arbitrators see themselves as part of a group of international adjudicators which role and raison d’être is to fulfil the particular needs of the international business community, and perceive arbitration as a free-standing and autonomous system of international justice. If the idea of such a free-standing system of international justice is accepted, it is perfectly understandable that international arbitrators try to be as consistent as possible with past decisions of other international tribunals. Such effort of consistency is not driven by a structural homogeneity of arbitration as a dispute resolution system, or by the hierarchical situation in which arbitrators would find themselves. There is no such homogeneity or hierarchy in international arbitration. The driving force of arbitral precedent is rather the arbitrators’ desire to meet the parties’ legitimate expectation that their dispute will be resolved by international adjudicators according to internationally accepted procedures and from an international perspective. That is to say: resolved in a way that is not a mere imitation of what municipal judges would do. The idea that opting for arbitration as an international means of resolving business disputes implies the adhesion to a justice which is to a certain extent different from that of courts not only as regards procedure but also as to the perspective adopted for the resolution of substantive law issues. Accepting the dynamics of arbitral precedent as a tool for consistency and as a rule-making instrument cannot go without accepting the specificity of arbitration, not only as regards procedure, but also the way substantive issues are dealt with. On the other hand, arbitration cannot be thought as a truly autonomous system of justice without accepting the role and existence of arbitral precedent. Precedent in arbitration and arbitral autonomy are two closely intertwined concepts.
3. Is Precedent the product of the intrinsic qualities of one or more particularly well-reasoned awards?
The precedential effect of arbitral awards is a phenomenon which analysis is fraught with difficulties. However, arbitral case law is a reality in practice, albeit an imperfect one. Past solutions have some impact on the thinking of arbitrators having to resolve future cases, even though they may not be referred to in their awards. In this respect, the quality of the reasoning of a particular award may of course play a role in the thinking of future tribunals. Yet, in the views of the authors, good reasoning is no driving factor of arbitral precedent. Precedent in international arbitration is not – or not only – the product of the intrinsic qualities of one or more particularly well-reasoned awards. It is not, either, the product of the arbitrators’ own will, although some show a certain tendency to include obiter dicta in their awards. Arbitral precedent is a pure phénomène d’entraînement. And it is all the more difficult to define that it is most of the times difficult to analyse the exact role that reference to past cases played in the arbitral tribunal’s reasoning. While an arbitral tribunal might refer to a given solution adopted in one or two particularly well reasoned awards as a mere illustration of its reasoning, the same solution will, if adopted in similar terms by five, six or more awards rendered in comparable cases, have not only an illustrative value but also a compelling effect. This is not to say that solutions given in a consistent line of awards will always be perceived as binding in future cases. Their relevance will of course depend from the rules of law applicable to the case. From this perspective, procedural issues should be distinguished from questions of substance.
As far as issues of procedure are concerned, it is beyond doubt that solutions adopted in past arbitration awards are likely to be considered as precedents by arbitrators. Decisions on procedural issues or questions of arbitral jurisdiction are the natural ground for the emergence of arbitral jurisprudence because arbitral tribunals have the first say on these issues and, arbitral tribunals having no forum, will generally not resolve them by reference to any particular national law. A similar conclusion may be drawn as far as issues of applicable law are concerned, as it is generally admitted that, in absence of a choice of law, arbitral tribunals can resolve the dispute by referring to the rules of law they believe to be appropriate. As far as issues of substance are concerned, reference to arbitral precedents will be possible when, absent a choice-of-law, the arbitral tribunal decides to apply transnational principles, trade usages. If non-national rules of law are to play any role in the adjudication of international trade disputes, arbitral precedents cannot but be an important source – albeit not exclusive – of the same. Even in presence of a choice-of-law may arbitral precedents play a role in the resolution of the dispute? For instance, arbitral precedent may well play a role when a particular legal issue has not yet been settled in the particular applicable law. International conventions providing for substantive rules of law will also be the natural field for the application of arbitral jurisprudence. This is all the more so when said conventions, like the CISG, present the characteristic of being detached of national laws as far as their interpretation is concerned. The assumption that reference to arbitral precedents would not be conceivable in respect to substantive issues in presence of a choice-of-law is therefore incorrect. Arbitral jurisprudence may be a source of legal rules in a number of different fields, including in respect to issues of substantive law and when a national law is applicable to the dispute.
4. Why do arbitral awards need to be available?
The concept of arbitral precedent naturally raises the issue of the availability of arbitration awards. The fundamental importance of the publication of arbitration awards derives from the fact that, absent of a doctrine of stare decisis in arbitration, arbitral precedent will only operate in presence of a repetition of identical solutions in a number of different cases. Precedent in arbitration is, from that perspective, a rule-making mechanism comparable to that of trade usages. For that rule-making mechanism to operate, it is necessary that arbitration awards be available in sufficient quantity to permit the emergence of trends and the distinction of lines of identical or similar solutions. In other words, in order for past awards to be perceived as binding, there needs to be something close to what has been defined as path dependency for state courts, i.e. the accumulation of identical or similar solutions able to generate a phenomenon of imitation. The persuasiveness, which supposes an exemplary value and, as a consequence, a judgement on the value of a particular decision, often needs to be combined with quantity. Precedential value could only be given to a consistent line of decisions. The same applies to arbitral awards. In order for arbitral awards to have precedential effect, it is therefore necessary that awards be known and available. The main condition for arbitration awards to have a precedential effect is therefore that such awards be known and accessible in sufficient quantity, in other words that they be systematically published. If arbitration is to remain the normal avenue for resolving business disputes, it needs to provide the business community with greater predictability of the possible outcome of trade disputes. In turn, better knowledge of arbitral jurisprudence would allow the business community to have a clearer idea of the realities and advantages of arbitration. This is of course not to say that arbitrators should be deprived of their discretion in the resolution of each particular case. But such discretion in assessing the facts and determining the appropriate rules of law is in no way incompatible with the availability of a body of arbitral precedents upon which tribunals could rely if appropriate.
Yet, and although a precise study remains to be made on the question, it would appear from a superficial survey of published arbitration awards that arbitrators rarely rely on arbitral precedents.
5. Why is reliance on arbitral precedents not frequent?
The reason is in our view to be found in the lack of transparency of commercial arbitration as a dispute resolution system. How can an arbitral tribunal ever conclude that consistent past arbitration awards express a rule of law or a trade usage when the overwhelming majority of arbitration awards are unknown? The proportion of court decisions which are made available to the public through publications in official bulletins, legal publications and on the internet is quite representative of the overall jurisprudential production of a given judicial system. The same cannot be said of arbitration. Save a very limited number of exceptions, almost no ad hoc commercial awards are published, whereas such awards probably represent a very consistent part of the total volume of arbitration decisions rendered each year in the world. Arbitral institutions are therefore the exclusive source of published arbitration awards. Yet, only a small minority of arbitral institutions do publish awards. Yet, this publication policy of some institutions only covers a small minority of the total volume of awards rendered each year.
6. Should all awards be published?
It is of course true that an important part of rendered awards may not present any interest, as they only settle issues of fact. Likewise, decisions rendered in commodity arbitrations or in ex aequo et bono (amiable composition) do not present any interest to the effect of setting a precedent. It is nonetheless out of doubt that the volume of published cases is not representative of the global reality of international arbitration. Awards are published randomly, depending on whether they have been rendered under the aegis of one of the institutions having a publication policy. In addition, the availability of information depends on the editorial policy of these arbitral institutions. The issue is however not how arbitration awards are selected for publication, but whether there should at all be any such selection, except for awards that are manifestly deprived of any interest. Publications that are driven by the desire to treat certain specific issues of general interest that the editor has sought to cover will not, because of the subjectivity of the editor’s policy and the limited range of issues covered, allow the creation of a data base sufficient to treat a wider range of questions. It should also not be overlooked that awards are frequently published under the form of summaries or in extracts, which frequently happens to be insufficient to make a finding possible.
7. Should awards be published with the names of the arbitrators?
Most of the times, it is not the case. Yet, at the difference of court decisions, knowing the arbitrators’ identity may be relevant to the effect of a proper understanding of the decision’s reasoning. Judges are part of a hierarchical and unified judicial body, so that their decisions are more the emanation of the judicial system to which they belong than their individual creation: what matters is more the circuit, the court or particular section of the court which issued a decision. Conversely, awards are rendered by individuals selected for their personal credentials and reputation, who have no forum and whose decisions are not subject to the control of any superior court. Such individuals will frequently have published extensively, and expressed opinions in respect to issues addressed in their awards. Knowing who they are can therefore be important information for a proper understanding of their findings. The high reputation of certain arbitrators may enhance the value of an award in the eyes of their peers. It can of course be submitted that the publication of awards with the names of the members of the arbitral tribunal could have in turn entails a multiplication of obiter dicta by arbitrators desiring to promote their own “jurisprudence”. It could also be feared that the publication of arbitrators’ names could have the effect of dissuading arbitrators to take bold positions. Yet, the example of investment arbitration, were awards are made public with the names of the members of the tribunal, shows that these inconveniencies, as real as they may be, do not outweigh the advantages of putting a complete, unabridged information at the disposition of parties and arbitrators.
8. How could a mass publication of complete, unabridged awards be achieved?
A model could certainly be the CLOUT data base. CLOUT is an information system based on a 1988 Uncitral decision, established for collecting and disseminating information on court decisions and arbitral awards relating to conventions and model-laws that have emanated from the work of the Commission. The scope and purpose of such system, as explained by the Uncitral’s user guide, is “to promote international awareness of such legal texts elaborated or adopted by the Commission, to enable judges, arbitrators, lawyers, parties to commercial transactions and other interested persons to take decisions and awards relating to those texts into account in dealing with matters within their responsibilities and to promote the uniform interpretation and application of those texts“(Doc.A/CN.9/SER.GUIDE/1/Rev.14 February 2000). Why couldn’t a similar system be instated to promote the international awareness of arbitral precedent in commercial arbitration? A new data base with that precise scope could easily be organised under the aegis of the Uncitral, with the same successful system than that which has been used for CLOUT. Awards could be submitted to the secretariat, which would then ensure that the names of the parties and any non relevant or secret information be deleted, exactly in the same way as this is done for published decisions of the European Commission in the field of mergers. The secretariat would also ensure that there is no opposition from the parties to their award being published online a certain period of time after it was rendered. Such a system would allow the progressive constitution of a wide data base which, provided an efficient index and search system be available, would constitute the necessary basis of the elaboration of a true system of arbitral precedent.
9. Is confidentiality a valid objection to the publication of arbitration awards?
It may however be submitted that a systematic publication of complete awards would go against the privacy and confidentiality of arbitration. The argument is in our opinion not a decisive one.
Many arbitration rules provide that arbitration awards should not be published without the consent of the parties. (see Uncitral Rules at Article 32 § 5, Article 43 § 3) of the Swiss Rules, Article 27 § 4 of the AAA Rules, article 34 of the ICDR rules, Article 30 § 3 of the LCIA Rules). A similar provision can also be included in the arbitration agreement itself, or in the terms of reference. In presence of such a provision, it is out of doubt that requirement of consent has to be complied with prior to publication. This does not mean, of course, that publication will in practice be impossible. In fact, such rules do not in general provide that consent should be given in written form. Implied consent after proper notice to the parties may therefore be sufficient to the effect of permitting the publication of the award. Certain precautions should in any case be taken prior to any publication of an arbitration award. No publication should clearly take place if the parties are still litigating. A certain period of time should also have elapsed after the award is rendered (the ICC practice of waiting at least three years before publishing an award seems in this respect particularly healthy). In addition, the award, although published in its entirety, should be expurgated of any potentially confidential or secret information contained therein, which inclusion is not necessary to the comprehension of the decision. This certainly includes the names of the parties involved in the arbitration, the names of third parties, as well as – unless necessary to the understanding of the award, such as data relating to market shares and turnover in antitrust cases – most of the economical and financial information contained in the award.
10. Is there really an overriding principle of confidentiality?
In absence of a provision expressly requiring the parties’ consent to publish the award, the issue is whether a rule to that effect can be deducted from an express or implied rule of confidentiality applicable to the arbitration. Most arbitration statutes do not expressly provide for a general principle of confidentiality. Arbitration statutes may provide for rules applicable to the protection of business secrets, or the secrecy of deliberations (as Article 1469 of the French new code of civil proceedings), but they do not embody a general rule preventing the publication of arbitration awards. As to arbitration rules, those which do not provide for specific rules applying to the publication of awards do not either, in general, contemplate a general principle of confidentiality. The ICC Rules refer, in Article 20.7, to the protection of business secrets and, in Article 21.3, to the privacy of the hearings. The rules of arbitration of the Vienna Chamber also limit themselves to the protection of business secrets. These provisions would not, for themselves, stand in the way of the publication of awards. Certain authors have however submitted that arbitration would be subject to an implied general principle of confidentiality. Such an implied principle of confidentiality would be a necessary consequence of the parties’ consent to arbitrate, be part of the lex mercatoria, or constitute a transnational rule of international arbitration.
The fact however that most arbitration statutes do not embody such a general principle of confidentiality could be seen as an indication that there is no such general principle. Quite on the contrary, the solutions adopted with respect to confidentiality are very different from one jurisdiction to another. In fact, far from expressing a general acknowledgement of the implied confidentiality of arbitration, case law seems to be oriented on the opposite direction in many jurisdictions.
The principle according to which hearings are held in camera is for example justified by the parties’ desire to protect the serenity of the debates. The confidentiality of the documents produced in the arbitration is justified by the need to preserve business secrets, etc. From that perspective, one could wonder what would be the rationale of preventing the publication of an award years after it was rendered if the names of the parties and any potentially secret or confidential information has been removed. There is no uniform conception of confidentiality in arbitration. The notion varies with the situations and functions which it is supposed to cover and does not even apply equally to all participants in arbitral proceedings. Positing that arbitration is la chose des parties and that the award belongs to the litigants is clearly not sufficient to prevent its publication. An award is not only the ultimate product of the parties’ arbitration agreement. It is not solely a private document. It is also a jurisdictional decision which may, to a certain extent, affect the public, and in which the business community at large has an interest. There are many instances in which disclosure of information relating to arbitration is required and permitted. Statutes applicable to listed companies may require the parties to publish financial information or when parties are compelled to disclose the award for the purpose of enforcement or annulment proceedings. A party may find itself in the obligation to produce the award to defend a claim, or to protect its interest or image. These are cases in which a counterbalancing interest imposes disclosure rather than secrecy.
Likewise, the public interest in the development of arbitral case law, in the enhancement of the quality of arbitration, and in providing transparency and predictability to the business community should override the principle of confidentiality as far as the publication of arbitration awards is concerned.
Alexis Mourre/Alexandre Vagenheim
________________________
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.
