I. INTRODUCTION

I intend to raise a few questions on specific points regarding what should go into an award, and how it is to be said.

We all know what the reasons for providing reasons for arbitral awards are. Probably the most important one is to oblige the arbitrators to think carefully of every aspect of the dispute and to allow them to come to a decision that is fully thought through and respectful of the rules by which they have to abide in reaching their decision, and that has the cogency, intellectual rigour and persuasiveness that only writing can provide. Related to this is the need to ensure fairness and due process and avoiding arbitrariness or even just loose thinking. Other reasons are to explain to the losing party why it lost and to strengthen the legitimacy of a decision rendered by individuals who exercise a judicial function without any public authority and, moreover, whose decisions are subject only to very limited review.

These reasons apply both to commercial arbitration and to investor-State arbitration. In investor-State arbitration there is at least one additional reason, which is to clarify the law and possibly to contribute to its development, as well as to promote consistency and predictability. The need for convincing and elaborate reasoning is particularly strong in investor-State arbitration at a time when the appropriateness of having private arbitrators deciding disputes involving affairs of State is under severe and (to put it mildly) not benevolent scrutiny.

It is useful to note that the problem of the reasoning of the award is, at least conceptually, different from that of the decision-making process. The decision—and therefore laying down the rule that settles the dispute brought to arbitration—is obviously the principal task of the arbitrators. In this connection there are multiple issues: what rules and principles should the arbitrators apply and how should they go about ascertaining their content, how should they assess the evidence, should they simply act as referees between the parties and choose between their competing arguments or do they have to perform a higher function of administering justice or finding the truth, should they strive to ensure some form of consistency with other arbitral case law (this is certainly more a concern in investor-State arbitration, but may not be totally out of place in commercial arbitration) and so on. One could call this the intellectual process of reaching this decision.

At least in the sense in which the term is generally used in this context, the reasoning is the explanation and making explicit of the intellectual path followed by the decision-maker to come to its decision. Strictly speaking, a reasoning in this sense is not required for a decision, and in fact awards can in principle be unreasoned. This was actually very often the case in older days. However, in modern arbitration—both commercial and investor-State —reasoning is viewed as indispensable for the reasons mentioned above, and it must now be in writing. This is consistent with the fact that arbitration is no longer limited to the settlement of minor technical disputes as was primarily the case at the origin, but is a full-blown alternative to the judicial process (and often actually the only option) for the settlement of important disputes.

The question we are asked to focus on is how this explanation of the grounds of the decision is to be done. Although the term reasoning is usually referred only to the part of the award that deals expressly with the grounds for the decision, I think it is useful to consider, albeit more succinctly, all the other parts of the award, since everything that goes into the text of the award is relevant to explain the thought process by which the arbitrators arrive at their decision.

I will adopt a practical approach, which goes essentially to the technique of drafting an award.

There are obviously a number of elementary requirements that apply to any type of award, things such clarity, logic, consistency, avoidance of ambiguities (and to the extent possible, repetitions), a decent use of grammar, syntax and language: those are the essentials required to render intelligible whatever the arbitrators intend to say. Beyond this there are many other ingredients to writing an award.

Basically, the main issue is: what, and how much, should the arbitrators say, and how should they say it? What questions must the arbitral tribunal ask itself and address when going about drafting an award?

This question is common to commercial and investor-State arbitration, although the approach and the solutions may not be the same in the two cases. And clearly there is no single way to answer it and each tribunal has its own way of writing an award, which may also vary from one arbitration to another depending on a variety of factors. These will include the nature, the importance and the complexity of the dispute and of the individual legal and factual issues raised by it as well as the energy and resources expended by the parties and their counsel. Other factors may be the nature of the review to which the award is subject and the level of publicity that the award may ultimately have.

The arbitrators will obviously also have to give consideration to the tension between the exhaustiveness and completeness of an award and the pressure to ensure the efficiency and speed of arbitration and the control of costs. One criticism often heard is that awards are too verbose. Is this always a fair criticism?

The problem of what and how to say things applies to all the parts of the award: the procedural history, the description of the facts, the arguments of the parties and, most importantly, to the specific arguments based on which the Tribunal reaches its decision. I will mention briefly the first ones and concentrate on the actual reasoning.

II. THE PRELIMINARY PARTS OF THE AWARD

The procedural history is clearly the most boring part of the final award, the one arbitrators find most tedious to write and the one nobody tends to read, except in case of a challenge. It is also the simplest to write, partly because it is the one that most easily lends itself to be delegated to assistants and secretaries. The question is whether too much effort is spent writing the procedural history, and whether this sometimes constitutes a disproportionate part of the final award. One of the culprits of the prolixity may be arbitral institutions, including ICC, which seem to insist on every single incident and piece of correspondence being mentioned.

There is however no doubt that the rehearsal of the procedural history has considerable importance, particularly to ensure that all procedural requirements (in particular due process) were respected. Moreover, for the tribunal, reviewing the procedural history in detail can sometimes serve to bring back to memory aspects of the case that may continue to be relevant even at the end of the proceedings but may have been forgotten during the course of the arbitration.

In some cases it may be appropriate to devote a specific section of the award, which may be part of the procedural history or a section on its own, to issues of due process. This may make sense in case of particularly contentious proceedings rife with accusations of breach of due process by one or all parties. A detailed treatment of this aspect may help protect the award in case of challenge on that ground. I have personally experienced the usefulness of this in a case of mine, where a specific section of the award on due process was heavily relied upon by the court to reject a challenge to enforcement on that ground. Of course, this issue may already be the subject of procedural orders issued in the course of the proceedings, but consolidating it in the award may have its usefulness. Obviously the need for this exercise will depend on the circumstances.

The facts section is clearly of extreme importance. It is indeed essential that the arbitrators have, and be seen to have, a complete and detailed understanding and control of all the facts and evidence relating to them. Setting them out in detail in the award may be a way not only to ensure the reader that the arbitrators have properly understood the facts but, more importantly, to oblige the tribunal to look at them with the requisite attention.

As for all sections of the award, the challenge is striking the right balance between prolixity and conciseness or, put otherwise, between exhaustiveness and relevance. At the end of the process it may become clear that certain facts are irrelevant for the result. Should they nonetheless be recounted by the arbitrators just to show that they have not overlooked them, even if ultimately, for a good reason, they were not considered relevant and do not support the reasoning?

Essentially the same considerations apply to the arguments of the parties, which set out their vision of the facts and of the law, and are usually the subject of a specific section.

One practical question, that applies to both the discussion of the facts and of the parties’ arguments, is whether these should be treated exhaustively in the initial sections respectively devoted to them, or whether those sections should only provide a less detailed, more bird’s eye review of them, leaving a more specific discussion in the specific sections of the reasoning proper for which they are relevant. In my view, the second solution has the advantage of allowing the reader to obtain a general understanding of the dispute, without getting lost in excessive detail. It may also have the advantage of avoiding too much repetition, since the reasoning will usually require addressing the specific facts and party arguments that are relevant for each one of its passages.

III. THE REASONING

The actual reasoning is obviously the most important part, the core of the award, which builds on the material of the previous sections.

It goes without saying that this part has to be very precise and clear, and must allow the reader to understand exactly the reasons for which the arbitrators arrived at their decision. This implies that, at a minimum, the reasoning must address and explain all the arguments that are directly necessary to support the decision reflected in the dispositive part of the award. Very importantly, it must avoid contradictions between different parts of the reasoning, and between the reasoning and the dispositive.

This, however, leaves unanswered many questions that come to mind when setting about drafting an award.

The first one is that of the level of detail with which these essential arguments should be treated. The answer may be in some way similar to another particularly delicate question. This is whether the arbitrators should adopt a minimalistic approach by addressing only the arguments (or claims or defences) strictly necessary to come to their decision, or whether they should adopt a more comprehensive approach and discuss arguments, claims and defences and factual or legal points that go beyond what is strictly necessary. The bare statement that the tribunal has considered all the parties’ arguments even if it has not specifically alluded to them may not be sufficient.

Sometimes the decision on a given point or claim or defence, or even on the entire dispute, can be reached without addressing all of the arguments raised by the parties. Should the arbitrators nevertheless also consider some or all of those arguments? The paradigmatic examples are jurisdiction and statute of limitations, but there are many more.

The easy answer would be that arbitrators should avoid going beyond the strictly indispensable. It is faster, more efficient. And one could say: least said, soonest mended.

Obviously, nobody would probably argue that the tribunal should even touch upon the merits if its decision is going to be that it has no jurisdiction (unless the merits are in some way intertwined with and necessary to decide on jurisdiction). However, the example of statute of limitations may be different. A tribunal might be uneasy relying simply on what, at least in some cases, may appear to be a formalistic argument. Its unease might be even greater if the decision on the decisive preliminary argument is based on an interpretation of the relevant rules which is not undisputed or is even hotly contested (perhaps even the governing law may be the subject of uncertainty or disagreement), and even more so if the decision is not unanimous (regardless of whether there is a formal dissent).

There are many other situations where the parties have discussed a host of factual and legal arguments that may justify a comprehensive analysis and an "even if" type of reasoning showing that the same result would be achieved also on other considerations. This can assist in giving comfort to the tribunal, bolstering the correctness of the final outcome and rendering the award more persuasive, strengthening it in case of challenge and giving the losing party the comfort that it has not lost on a mere technicality. Parties may be more prepared to accept an unfavourable decision based on a curt and formalistic argument from a court than from arbitrators.

Matters obviously stand differently if the arguments, beyond those which allow the tribunal to come to a certain result, go in a different direction. In that case it would evidently be foolish to address them and to say that, but for the winning arguments, the outcome would be different. It would clearly weaken the award.

In considering whether to stick to the bare minimum that allows them to come to a decision or to expand their reasoning, and in deciding what more to say and how to say it, arbitrators must take into account a variety of other factors. Efficiency and speed are only one consideration, albeit an important one. Arbitrators must be able to gauge what is expected of them from the addressees of the award and to give satisfactory answers, not only in terms of the strict factual and legal basis of the decision. The applicable law or rules will usually provide little or no guidance since they typically prescribe simply that the reasons should be adequate. Article 47(1)(i) of the ICSID Rules is an exception in that it goes a little further by dictating that the award deals with "every question submitted", but this too is far from providing an exhaustive answer to the problems on the table.

If the dispute is simple and without broader implications, and even more so if it is low value, and the pleadings have not been particularly sophisticated, it is reasonable to assume that the parties do not expect an elaborate decision, so that a bare-bones reasoning will suffice.

I would hesitate to say, as a general proposition, that the same can be said when the dispute is complex or high-stake, or one in which the parties have invested much effort in marshalling abundant and sophisticated evidence and developing elaborate arguments, even if ultimately they turn out not to be outcome-determinative. In such cases, it seems to me that the tribunal owes it to the parties and their counsel to explain that even those arguments have been duly understood and considered.

This is particularly so in cases where there is a difference of opinion within the tribunal and of awards that are likely to become public, and even more so in investor-State arbitration. Here the addressees of the award are in practical terms not just the parties involved but, in addition perhaps to amici, also the public at large, which may well not have a clear understanding of the issues in dispute, and from whose viewpoint a broader perspective on them may be useful. The higher the stakes and implications, the stronger I believe is the case for erring on the side of exhaustiveness.

I am therefore not sure there should necessarily be a sharp dividing line between investor-State arbitration, for which exhaustiveness of the reasoning is almost taken for granted, and commercial arbitration. I think fairly similar considerations can apply to important commercial disputes, particularly those where States or State entities of public interests are involved—although one difference is that investor-State awards normally become public which is less likely for commercial cases, even involving States.

A related question can be the extent to which, in order to give the complete picture and to put the decision in context, the arbitrators may, or even should, refer to matters or arguments on which the parties have not particularly dwelt or have referred to only tangentially or not at all (and which the arbitrators have not raised with them). In principle, this is something the arbitrators should not do, particularly if it involves the risk of straying into controversial or (worse still) policy or political considerations. One could wonder, however, if in some situations the tribunal should at least mention that certain issues have not been raised, were it only to show that it has not missed them. What I instead think the tribunal can do is rely on authorities or theories not mentioned by the parties if they merely add further support to the solution which the arbitrators would in any event reach based on the theories or arguments developed by the parties, and which increase the persuasiveness of the decision. In such a case there should be no risk of the parties being taken by surprise.

Another situation which requires some reflection is the one where the arbitrators’ decision may have implications beyond the case at hand. This is obviously the case for investor-State awards, but may also be the case, for example, where the parties are involved in other similar relationships which may equally give rise to disputes or where the decision deals with an issue that may have broader repercussions in a particular industrial sector. Depending on the situation, these considerations may militate in favour of giving extensive reasons or, instead, greater restraint. From a somewhat different perspective, when the dispute is between parties involved in a broader relationship, the reasoning must not contain elements capable of negatively affecting this relationship and should actually, if possible and appropriate in the circumstances, try to facilitate that relationship.

In principle tribunals should refrain from addressing an issue simply because they have invested a lot of work on it, only to realise at one point that it is not really relevant to the flow of the reasoning. They should also avoid an overly scholarly approach. They should likewise refrain from needlessly castigating the parties or the lawyers for what they have said or not said, except perhaps where that is called for in order to explain that certain arguments negatively impacted the efficiency, speed and cost of the proceedings. But here also the arbitrators need to be careful because, depending on the circumstances and issues at stake, the parties may be justified in wanting to raise all possible arguments to support their case even if they end up not to be winning ones, or might have from the beginning appeared to be a long shot.

On the other hand, the tribunal must not succumb to the temptation of taking shortcuts to sidestep difficult issues by failing to give adequate reasons or providing only minimal reasons, perhaps relying on the fact that its reasoning is largely beyond the review of the courts. It is true that courts are very loath to set aside for failure to state reasons, and rightly so because such a review would be fatal. This has just been confirmed by the English High Court in The Islamic Republic of Pakistan and the National Accountability Bureau v Broadsheet LLC ([2019] EWHC 1832 (Comm)), which also held (at para. 47) that the standard is the same when a State is a party to the arbitration. I submit, however, that this definitely should not be taken to mean that arbitrators should not go beyond the bare minimum that suffices to avoid setting aside. Providing reasons for their decision is a fundamental part of their job and they should therefore put into it all the efforts and diligence commensurate to the type of dispute and issues before them.

Also for this reason tribunals should not cave in to the pressure of institutions breathing down their backs to issue the award quickly if, in the circumstances, the reasoning needs to be particularly elaborate. This of course does not justify inordinate delays in the issuance of the award.

IV. CONCLUSION

These are admittedly all fairly general considerations that need to be balanced and adapted to each case and that for the most part apply both to investor-State and to commercial arbitration. As I said before, and as is obvious, there is no single way to write an award. A lot depends on the circumstances, and much is a question of style that will depend on the personality of individual arbitrators. The greater or lesser experience and confidence, and reputation, of the arbitrators may also play a role in how they decide to deal with the different issues in the award.

Concretely, arbitrators need to use their broad discretion well and to strike a delicate balance between saying too much and saying too little, always keeping in mind the expectations of all those concerned by the arbitration, as well as the need to render an award capable of withstanding challenges— which may mean saying certain things and not saying other ones. Since they lack the same authority of judges, they should strive to be as exhaustive and convincing as possible.

One thing that definitely must be avoided is a reasoning in which the actual reasoning consists of a few pages sandwiched in between hundreds of pages of recital of procedural history, facts and positions of the parties. This type of award is totally unsatisfactory for the parties and goes against their legitimate expectations. The complaint of excessive length of certain awards, particularly in investor-State cases, may not be due to the length itself, but to the lack of proportion in some of them between the actual reasoning and peripheral material.

A point that would deserve some consideration is whether the content of the award, and in particular the level of detail of the reasoning, could be the subject of discussion between the tribunal and the parties at some, preferably early, stage of the arbitration. It is certainly open to the parties to flag to the tribunal any particular wishes or needs they may have in this respect, so that the matter can be appropriately codified in the terms of reference or appointment or in a procedural order.

A final consideration, to which little attention is devoted, is that in deciding what to write in their award arbitrators should also be particularly attentive to the implications of their reasoning on the res judicata effects of the award. These depend in part on the applicable law, the identification of which is itself a very difficult subject. To simplify, there are different conceptions on the res judicata of awards, under some of which also the reasoning may have certain preclusive and conclusive effects. Where that may be the case, the arbitrators have to consider with care how broad they want the effects of their decision to be.

As it is the case for so many issues in arbitration, the ultimate solution lies on a judicious exercise of discretion on the part of the arbitrators.

---