INTRODUCTION

Everyone knows that there can be no trial nor court decision worthy of those names without due process as this principle is the very essence of a trial. But this does not apply to the reasoning of court decisions, which is not consubstantial to the proceedings. One has always stated that Divine justice need not be reasoned: indeed, the Lord need not explain, for His decision is necessarily right and fair. Those who speak in His name are not required to explain.

Human justice, for its part, must rely on reason, on the law, and cannot be subject to arbitrariness. Setting out the tribunal’s reasoning is, by far, the preferred way of showing that the solution was not arbitrary and that justice has been rendered: indeed, "Justice must not only be done, it must also be seen to be done."¹

However, it is only very recently that the practice of setting out reasons for court decisions has become widely accepted. Indeed, for a long time, it was usual practice that the judge simply offered a solution without explaining it. Thus, history teaches that setting out reasons for court decisions is relatively recent. Accursius, famous erudite of the Middle Ages (glossator), had even considered that court decisions should not in principle provide reasons.² In France, the duty to state reasons was imposed during the Revolution (laws of 16-24 August 1790), but it had appeared two centuries earlier in Florence or in Spain, in the province of Aragon.³ In England, as noted by J.-L. Delvolve, given the characteristics of the procedure and the work done by the parties in terms of fact-finding and evidence, the duty to state reasons is not as entrenched in legal tradition as it is in romanogermanic legal systems.⁴

Turning now to arbitration, whether to set out the legal reasoning of awards may appear to be one of those well-settled questions that do not raise any particular issue. An author even considered that, in recent years, rare were the subject matters that have raised as little interest as the reasoning of awards!⁵ Indeed, there is a certain unanimity in this regard, which can be easily explained: the basic principle is that awards should be set out their reasons. This reflects what most texts require, what parties or arbitration institutions wish for and what arbitrators do in practice. But if one takes a closer look and analyses the last few decades, one realises that things have, somewhat, evolved.

In order to encompass this evolution and assess the present situation, we can ask a few simple questions:

>> Should an arbitrator state his reasons? (see section II below);

>> What does setting out reasoning actually mean? (see section III);

>> Why should an arbitrator state his reasons? (see section IV); and

>> What is the appropriate standard of review in respect of the reasons stated in arbitral decisions? (see section V).

II. SHOULD AN ARBITRATOR STATE HIS REASONS?

The decision-maker is not always keen to disclose the reasons underpinning his decision. Indeed, being compelled to give reasons entails the risk of being exposed to criticism and, thereby, losing a bit of one’s authority. It is in this light that we should understand Lord Mansfield’s well-known advice to his fellow judges in the 18th century: "Consider what you consider justice requires and decide accordingly. But never give your reasons; for your judgment will probably be right, but your reasons will certainly be wrong."⁶ What is amusing is that, at that very moment, in France, Daniel Jousse, an 18th century prominent French lawyer, was saying exactly the same thing, and Talleyrand will not write anything different in his Mémoires⁷ one century later.

However, international laws and conventions now practically always include an obligation to give reasons (although French arbitration law does not require it).⁸ This obligation is sometimes expressly provided for (see e.g., Belgian Judicial Code (art. 1713(4°), Italian Code of civil procedure, (art. 823-5 CCP). It is also the case under French domestic law (art. 1482(2) CCP). But, under French law, in international cases, the provisions regarding the content of arbitral awards would seem to require reasons to be given in the award. However, as we shall see later, the issue is not as clear-cut as it would seem. The Washington Convention does impose the obligation to reason the decision.⁹

It is often the case that the parties are free to waive this obligation: in this sense, there is only a presumption that the award should include its reasons. This is the solution adopted by English Law, under the 1979 Arbitration Act, and later under the 1996 Arbitration Act (Art. 52(4)), by Swiss Law (art. 189(2) LDIP), just like the UNCITRAL Model Law (art. 31.2) or the Geneva Convention (art. VII).

One should stress that, although certain laws authorise the parties to relieve the arbitrators from having to state their reasons in their awards, in practice, they very seldom do. One would be hard put to find what would be the incentive for the parties not knowing the reasoning behind the solution imposed on them by an award, especially if this has no impact on the arbitrators’ fees.

Arbitration has been impacted by the way ideas have evolved with regard to due process. The right to a fair trial, as developed over time by the European Court of Human Rights, also implies the duty to state reasons. The latter has a somewhat special place because it is considered to be part and parcel of the fundamental right to a fair trial while, at the same time, being open to adjustments. Indeed, the European Court of Human Rights admits that the obligation to give reasons may vary depending on the nature of the decision and considers that it should be examined in light of all the circumstances of the case.

In short, the rule is easy to formulate: unless the parties decide otherwise (when such a decision is necessary), the rule is the obligation for the arbitrators to state reasons in their awards. Therefore, the interesting question is not whether arbitrators should articulate reasons in their award but, rather, what providing reasons actually means and then why provide reasons at all and, finally, how to review the reasons provided in the award.

III. WHAT DOES STATING REASONS ACTUALLY MEAN?

It might be useful to define what is meant by "stating the reasons" or the "motivation" for an award. In latin languages, the etymology is interesting: it comes from motus: movement, and movere: to move. Motivation or motivazione or motivación (which means "justification" in French or Italian or Spanish) is what encourages you to go in that direction, it is the expression of what leads your opinion in this direction rather than any other. Thus, reasoning is a rational approach of the mind. But reasoning, when we are judging, is not merely an internal approach because it requires that our thought process to find a solution to the question raised be externalised with the help of words.

This attempt at defining what "providing reasons" means already enables us to see what it is not: for instance, when an arbitrator sums up a party’s position and then writes: "the tribunal shares this opinion", there is no legal reasoning, it is simply sloth. Efficient arbitration institutions do notice this and share their observations with arbitrators so that they have to explain why they chose to uphold the arguments of one party, which usually also requires them to explain why they rejected the other party’s arguments. A review of arbitral awards shows that they are becoming longer and longer, even though this is not always due to the section relating to the "reasoning" but, rather, to the section which "sums up the positions of the parties". Be that as it may, on the one hand, the growing demands of the arbitration institutions and, on the other, the tendency to draft awards in such a manner that simply reading them will allow for a full understanding of the case have played a role in encouraging this trend.

There are different types of arbitrations or awards; are there also different types of reasoning or specific reasoning? This question should be answered in the affirmative and it is precisely in this regard that this matter has evolved.

Is there any point in giving reasons in an award by consent? One might wonder since, in this type of award, the parties themselves have found the solution. Explaining their own solution to them is useless: the rationale for stating reasons has disappeared. In concrete terms, an arbitrator who accepts to render an award which embodies the agreement of the parties need not give further reasons than a standard formulation which could read: "The parties have agreed to this or that solution; it does not, in my opinion, violate Public Policy and, therefore, I echo these elements as mine". This is why the 1986 Dutch Law, and also the UNCITRAL Model Law (art. 31.2), have expressly chosen to waive this requirement in relation to consent awards.

Awards rendered in trade associations arbitration cases (such as commodity trading and shipping) are worth analysing. Reading though one such award, one may have the impression that there is no legal reasoning. In fact, the reasoning exists, but it is extremely simplified given the specific nature of the dispute which relates to the quality of the goods. In this context, arbitrators are in essence technicians who can confine themselves to testing a set of requirements and drawing the necessary inferences. As Lord Bingham wrote: "There are some arbitrations, those of the ‘look-sniff’ variety in particular, where there is really no room for the giving of reasons: tapioca pellets either are, in the experienced judgment of a trade arbitrator, of fair average quality or they are not; whichever way his opinion goes there is probably not much that he can usefully add by way of exegesis."¹⁰

Does ex æquo et bono arbitration or amiable composition change anything? It does not, as far as the obligation to give reasons is concerned, but admittedly it influences the style of the reasoning. Indeed, the arbitrator no longer bases his reasoning solely on the applicable rules of law but must also explain the reasons why he or she chooses to depart from them, as the case may be. He must clearly set out the reasons why the outcome of the cause would be inequitable had he confined himself to applying the legal rules or give his reasons to hold that the legal solution appears equitable on the facts. In practice, the reasoning will tend to focus much more on all the circumstances of the case.

This publication devotes a chapter to the distinction between commercial arbitration and investment arbitration. In principle, the obligation to state reasons applies equally to both fields. But as regards the reasoning itself, one might perceive some differences, because, amongst other things, investment arbitration usually places a greater reliance on precedents, i.e. past awards addressing similar issues.

However, a careful review of awards shows that, whatever the type of award, certain set phrases and practices are routinely resorted to by arbitrators in order to avoid subsequent challenges. Such precautions often relate to the procedure applied by the arbitrators or their general approach. In this regard, I am not of the opinion that there should be a fundamental distinction in the legal reasoning expected of arbitrators depending on whether their background is one of civil law or common law.

There is another question which is beginning to stir up controversy and that I shall only mention: there is a growing trend in favour of publishing arbitral awards coupled with the development of artificial intelligence. Will this not lead to changes in the way arbitrators set out their reasoning, possibly towards a more frequent use of standard and formulaic reasons?

We have just seen that giving reasons for an award does not always mean the same thing depending on the type of award or the context. This far, however, a question has remained in the shadows and we need to tackle it: why should one give reasons for the decision?

IV. WHY SHOULD AN ARBITRATOR STATE HIS REASONS?

The question is: why should one give reasons? The answer lies in the answer to another question, i.e. whom are the reasons intended for? There are as many answers to this question as there are recipients. Indeed, the legal reasoning offered in support of a decision fulfils different functions that are common to both civil law and common law despite their different legal traditions

In this matter, a distinction should be made depending on the parties, the arbitration institutions involved and, later, the court that might be asked to review the award. But one should also not forget the arbitrator himself and that is where I suggest we start.

The arbitrator Indeed, we tend to forget one of the main functions of giving reasons, if not its central purpose, because it comes before the award itself and relates to its drafting. Fleshing out the reasons in writing helps the person who decides: it compels the decision-maker to formulate logically the thought process which has led him from the evidence and legal arguments of the parties to his solution. Jules Renard, who was well known for his wit, used to say: "Once I have decided, I hesitate for a long time." It is a joke, but not only a joke. As a matter of fact, we, arbitrators, very often have an idea of what the outcome should be, even before settling for a particular reasoning. Legal reasoning moves backwards: from the intuitive solution to the facts. Articulating legal reasons for a particular outcome enables arbitrators to check the soundness of the assumption. Facing seemingly insuperable difficulties in the drafting sometimes serves as a cue that the solution is not robust enough, or even mistaken and should be dropped.

The parties The main function the reasons stated in the award serve is to justify and explain the chosen solution to the parties and, first and foremost, to the losing party who wishes to know why this is the case, why his arguments or pieces of evidence failed to win over the tribunal. The winning party is more interested in the last page of the award and will soon forget being disappointed with the reasons for the award as long as the eventual award is favourable to him.

The reasons given in awards are more detailed than in court decisions. This is due to the fact that disputes often involve significant amounts and are complex, but also to the fact that, arbitrators being remunerated, parties expect them to make a conscious effort to show that they have considered every argument. Moreover, offering a thorough legal reasoning is certainly a pedagogic endeavour on the part of arbitrators. This is important and characteristic of arbitration because the end game is not only to render a valid award but to go beyond and guarantee its acceptability¹¹ by making it convincing.

The way reasons are set out in awards varies according to the arbitrator: they each have their own style, depending on the case or mode of thinking, each one may wish to emphasise legal arguments, factual elements or evidence, etc.

There are differences between awards and court decisions in terms of the purposes the reasons serve. Obviously, in both cases reasoning is a rampart against arbitrary decisions, but a court decision must be reasoned for reasons of its own that do not exist in arbitration, at least in commercial arbitration. Justice is public and stating reasons is one aspect of its democratic character. Arguably, as investment arbitration cases often touch on issues of public policy, there is a strong case for making awards in this field available to the public.

Individual opinions, when dissenting, raise specific problems. Indeed, as a matter of principle, dissenting opinions are not part of the award; they cannot, therefore, be considered part of the reasons for the award; moveover, even if they were to state reasons for anything, it would relate mainly to a solution which is the opposite of the solution ultimately upheld. Separate opinions, i.e. concurring opinions based on different grounds, seem to have a different status because they also justify the chosen solution, even though for reasons that differ from the majority award: they just form part of the award.

The arbitration institution Some rules—and the ICC rules feature prominently in this regard—provide that the draft award be submitted to the institution for a formal review. The scrutiny of ICC is very extensive and mainly relates to the legal reasoning. In so doing, ICC acts in the interest of the parties who will receive the award and in contemplation of a potential judicial review of the award. The objective is, therefore, twofold: make sure that the award is intelligible for the parties and that there are no irregularities affecting the validity of the award. The arbitral tribunal’s legal reasoning may be challenged for being deficient, insufficient, unrelated to the arguments raised, or inconsistent. There is obviously a fine line between objecting to deficient reasoning and disagreeing with the findings the tribunal made on the merits. It should be noted that in practice, arbitral institutions when scrutinising the quality of the reasoning in awards, tend not to take into account the applicable law as chosen by the parties, nor the law of the place of arbitration, but rather, insist on thorough reasoning, even on matters such as the quantification of the damages. While French law, for instance, does not require detailed explanations and considers that the determination of the amount of damages should be left to the discretion of the court or the arbitral tribunal, and while the Cour de cassation itself ruled that the judge is under no obligation to specify the factors that contributed to the assessment,¹² conversely, ICC takes the opposite view and, even in relation to awards rendered in France, upholds a similar rule to that which exists in English law since the 1966 Ward v James case.¹³

The judge with powers of judicial review The reasons stated in the award is of particular interest for the judge who will review the award, either within the framework of setting aside proceedings or that of enforcement proceedings of a foreign award. Stating reasons for the decision is, as we saw, a guarantee of a proper administration of justice for it allows for an efficient review of the decision and clarifies its scope. Indeed, the scope of res judicata often depends on the interpretation of the reasons which are, if one believes in the maxim Res judicata pro veritate habetur, an expression of the truth.

But, as far as the judge is concerned, the crux of the matter is to determine whether the reasoning can be subject to review and if so, what the standard of review should be. We will therefore end this whistle-stop tour with the applicable standard of review.

V. WHAT IS THE APPROPRIATE STANDARD OF REVIEW IN RELATION TO THE DUTY TO STATE REASONS?

Who reviews and how? Chronologically, it can be the arbitration institution and, more frequently, the court when dealing with setting aside proceedings or the enforcement of foreign awards. The role of arbitration institutions has already been depicted, all that remains to be addressed now is the review carried out by the court. I will, intentionally, deal only briefly with this important point because it would take a whole day to deal with it thoroughly.

At the outset, an important distinction to be drawn concerns the scope of the review: either a substantive review of the award on the merits or a more formal review comes to mind. The dividing line between the two is very blurred. In this regard, determining whether reasons are stated in the award is a special case: does this involve a review of the merits or simply a formal assessment taking the award at face value? If it suffices to confirm that reasons, whatever they may be, are in fact stated, it is obviously formal. However, to the extent that the essence of the reasons themselves— whether they are adequate and consistent—is challenged, it is a question of merits. Let us now take these two types of review one by one.

The review of the existence of reasoning, considered as a formal review, is assessed according to the applicable law which, as we saw earlier, often requires that reasons be given, but which could accept certain nuances or exceptions if it is provided for in the law or if the parties so wish. Total failure to state reasons is rare and the setting aside of awards on those grounds is just as rare.

It is worth noting that neither under the New York Convention nor under the UNCITRAL Model Law the failure to state reasons constitutes a ground for setting aside arbitral awards. Those jurisdictions, which require awards to state their reasons, do not necessarily consider that it would be against international public policy to enforce, in their country, an award rendered under a law that accepts to waive this obligation. The French Cour de Cassation has consistently taken this approach in its jurisprudence since the Elmassian case¹⁴ in 1960, and this rule has now found its way into art. 1483 of the Code of Civil Procedure according to which the failure to state reasons only constitutes a ground for annulment in relation to domestic arbitral awards. However, it is important to note a very recent decision by the Paris Court of Appeal dated 2 April 2019 which took the opposite view; holding that the obligation for

judicial decisions to state their reasons is an element of the right to a fair trial. Arbitrators who fail to state reasons for their decisions disregard the scope of their mission and recognition of an award devoid of legal reasons goes against the French conception of International Public Policy.¹⁵

It is too early to say whether this corresponds to a reversal of jurisprudence, all the more so since a June 2019 decision seems to revert to the traditional solution.¹⁶

Assessing the relevance of reasoning, in fact, is akin to revising the award: this review is generally excluded. Reviewing the quality of the reasoning is a delicate task. To put this issue into perspective, let us examine two different legal systems: the French and the English.

The French case law has undergone interesting changes in relation to conflicting reasons. Originally contradictory reasons in arbitral awards and court decisions alike were thought to cancel each other out so to speak: thus, the award was not deemed to be reasoned. French law later evolved and opted for the opposite solution but only in respect of arbitral awards. From now on, it is not only irrelevant whether the reasons provided were apposite to the case at hand but also, as we just saw, whether the law was correctly applied or whether the legal reasoning evinces potential contradictions. Under French law, it therefore suffices that reasons— whether or not legally correct—are stated in the award.¹⁷

More often than not, review of the merits is excluded, at least in most cases. However, some legal systems may, without allowing for a review of the merits, offer the possibility of striking down awards where the reasoning reveals a major legal mistake. English law is interesting in this particular regard: appeals on points of law (art. 69 Arb. Act) exist, but it is possible to waive this right, most notably by selecting arbitration rules in their arbitration agreement that exclude any right to appeal the award. In the USA, review of the manifest disregard of the law has become less frequent since the 25 March 2008 Hall Street v Mattel ruling.

Sanctions are different whether they relate to failure to state the reasons or to the contents of the reasoning itself. The lack of reasoning typically entails either the setting aside of the award (e.g. ICSID Convention), or remanding the case back to the arbitral tribunal so that the omission may be put right and the reasons given. The latter is the current solution under English Law pursuant to Art. 70(4) of the 1996 Arbitration Act).¹⁸

The substance of the reasons given might lead to the award being set aside when the outcome would be contrary to public policy but this sanction is not always applied. For instance, awards tainted by corruption will only be annulled on grounds of Public Policy under Swiss Law in the event that the reasons set out in the award show that the arbitrators chose wilfully to turn a blind eye to issues of corruption.

VI. CONCLUSION

My conclusion is that, at the end of the day, the practice of international arbitration shows that the duty to state reasons is generally understood in fairly similar ways across jurisdictions, both in terms of its recognition and ambit. It is as though the wind of practical exigencies has erased the footprints on the various routes taken by arbitrators during their training to leave but a single path. But may each one of us proceed at his own pace!

---

---