1. INTRODUCTIVE PROPOSITIONS

I wish to offer the following propositions as a general frame for this Institute Dossier.

First, I never experienced a case of total absence of reasons in an award. Pursuant to most arbitration laws and institutional rules in force worldwide, the reasoning is now strictly required by mandatory norms and only a few legal orders allow the parties to dispense the arbitrators with the duty to give reasons, if the parties so agree at the outset of the proceedings. A further, obvious, exception is the "award by consent", in which the reasoning is replaced by the terms of the settlement agreement stipulated by the parties and endorsed by the tribunal.

Apart from these rare exceptions, we should therefore take for granted that the reasoning is a constant presence in all awards. Accordingly, the articles in this publication will not, presumably, wonder whether the reasons must be part of an award, but rather discuss what adequacy standard the reasons should meet. In other words, we will assume that "failure of reasons" is tantamount to "failure of adequate reasons" to then embark on the search for the minimum threshold to meet the adequacy or sufficiency standard in the reasoning that speakers and participants consider necessary.

Second, arbitral reasoning should be viewed distinctly from judicial reasoning. A badly or wrongly reasoned judgment remains subject to appeal and will be reviewed on both facts and law. It is because of this circumstance that the quality of the Court’s reasoning may be relatively lower without causing irreparable harm to the parties. However, the appeal is not a remedy available to review or rectify an award, which remains "final" although potentially unsatisfactory or deficient on facts or law. It is this essential distinction that makes the arbitrators’ task, when drafting the reasoning, significantly more burdensome and sensitive than the equivalent task of a State lower Court.¹

The difference is no surprise for the following reason. The Court is a State organ committed to render justice in accordance with the State legal order. It is obviously accountable to the public for improprieties or wrongdoings in administering justice, but not bound by any duty towards the parties in the individual case to frame the reasons of the judgment according to defined quality standards. The arbitrator is instead a "private judge" owning to the parties a duty to fulfil the mandate with diligence and care, which includes the duty to provide reasons understandable to the parties, by which he/she is remunerated. His/her principal duty is to render a decision that, in conformity with the parties’ expectations, shall be exhaustively responsive to their claims, arguments and defenses.

In simpler terms, the arbitrator’s task is greater, but so is his/her paycheck. Third, for what concerns the legal reasoning, arbitrators should be reminded that the principle jura novit curia is of limited application in international arbitration. Whereas the factual reasoning requires no more than a careful analysis of the evidence in the record, the legal reasoning may face a cultural difficulty which is a typical feature of arbitration. By common experience, the members of a tribunal may be educated in a legal order different from the applicable law chosen by the parties so that the latter will be viewed as a "foreign" law from the arbitrators’ perspective. This peculiarity is currently much less troubling than when I was young. At that time, application of foreign law fluctuated between two alternative approaches, both drawn from State court proceedings: either the foreign law was treated as a question of fact (common law jurisdictions) left essentially to the parties’ burden of proof; or was treated as a question of law (civil law jurisdictions), left to the autonomous determination of the arbitrators under the adagio juria novit curia enabling them to apply whatever law ex officio and even irrespective of the parties’ submissions in law.

However, neither one proved suitable in international arbitration. Being a creature of the parties’ consent, the arbitral process may be shaped more flexibly than the process before courts and, over the years, a betterbalanced approach was specifically tailored for arbitration. This consists of a constructive cooperation and dialogue between parties and tribunals to correctly ascertain the contents of the applicable law: the parties provide the tribunal with the sources of law and their reading of the legal texts supported by case-law; the tribunals are not confined to the information received from the parties and have the power to sua sponte inquire about the contents of the law they must apply, however with an important limitation that directly affects their legal reasoning.

The limitation consists in the following. When autonomously interpreting the law they have to apply, before framing their reasoning by adopting a legal interpretation which was never raised nor pleaded by the parties, the arbitrators must give them a reasonable opportunity to be heard. By failing to so do, the tribunal will take the parties by surprise in the final award and this would be tantamount to a breach of due process and lead to annulment of the award, particularly when the arbitrators’ ex officio construction of the law proves to be decisive for the outcome of the case.

In my view, this is one of the most crucial issues when drafting the reasoning section of an award: it indeed requires particular caution by the arbitrators, who should timely (i.e. before final deliberation) consult the parties on the legal issue concerned, to avoid finding themselves too late to do so in light of the time-limit to render the award.

Fourth, another mantra is the "sufficiency" or "adequacy" of the reasoning as an inseparable component of the duty to give reasons. Does this mean "exhaustiveness" implying that the award should address not only all questions, but also all allegations, arguments or defences raised by the parties? It is common practice for the parties to support a claim or a defence or objection on multiple alternative or sometimes cumulative arguments. Does a well-reasoned award need to take a position on every and all such arguments?

In investment disputes, Article 48(3) of the ICSID Convention provides that "The award shall deal with every question submitted to the Tribunal, and shall state the reasons upon which it is based". Article 47(1)(j) specifies that one of the essential elements that must appear in the award is "the decision of the Tribunal on every question submitted to it together with the reasons upon which the decision is based". According to legal literature, a similar approach is shared in international commercial arbitration

However, the "every question" requirement is to be construed reasonably. As case-law and legal literature show, a distinction should be maintained between "question", on the one hand, and "argument" or "allegation" or "plea", on the other hand. The "question" is the issue that has a bearing on the resolution of the case and cannot be left undecided, otherwise giving rise to an infra petitum which in most legislations is a ground for annulment.² In order to resolve the specific "question", a tribunal must of course consider all arguments and counter-arguments, but is not bound to decide them all where it finds that some of the arguments are prevailing or absorb the others.³

The only admissible exception is the case in which two "questions" are apparently distinct but in realty strictly intertwined. In such a case, a prudent tribunal does not omit to consider both "questions" but may validly rule, providing the relevant reasons, that "question A" is related to, or encompasses, "question B" and the decision on "B" is absorbed by the decision on "A". This would be a perfectly adequate reasoning, allowing readers to understand that both "questions" were duly considered and that the necessary reasoning is not missing. This was the approach recommended by the International Court of Justice and adopted by ICSID ad hoc Committees in annulment proceedings.⁴

Fifth, the reasoning is required since it provides the only tool allowing for control over the legitimacy of the arbitral process: was due process respected? Were the parties provided with a reasonable opportunity to be heard? Did the tribunal decide by applying the law it was mandated to apply? And so on: I will provide below a longer checklist of all the "musts" that a reasoning should satisfy for various purposes, however it appears evident that procedural legitimacy cannot be checked when reasons are so vague as to leave unanswered the above questions. In similar cases, the reasoning is missing because no controller may infer whether the award withstands the legitimacy test or is, on the contrary, the product or arbitrariness.⁵

Sixth (and last), several chapters will make the case that the purpose of a good reasoning is to make the award "convincing" to the readers, and even "persuasive" to the losing party. This is correct, but we should recall that two-thousand-four-hundred years ago Aristotle had already come to the same conclusion. In his view, rhetoric is the typical "art of persuasion" or the "ability of observing in any given case the available means of persuasion", an art implemented in three modes: ethos, that is the "power of evincing a personal character which makes oneself credible"; pathos, that is the "power of stirring the emotions of the recipients"; and logos, that is the "power of proving the truth, or an apparent truth, by means of persuasive arguments".⁶

Approximately three centuries after Aristotle, Cicero underlined in his monumental work De Oratore, also known as De Eloquentia, that advocacy was typically an intellectual art to persuade the audience through elocutio or loquendi elegantia.⁷

Therefore, we are not "reinventing the wheel" so to speak. Legal techniques have progressed (or perhaps regressed?), but legal reasoning has maintained the fascinating components of a rhetorical mission exactly as depicted centuries ago by Aristotle and Cicero.⁸

2. THE CHECKLIST FOR ESTABLISHING WHETHER THE AWARD IS WELL-REASONED

I strongly hope that a consensus will be reached to qualify an award as adequately reasoned if it meets the prerequisites enumerated in the following checklist.

An award is not satisfactorily reasoned in one or more of the following cases:

i. The reader, especially the Court which is approached for either enforcing or annulling the award, is unable to understand the logical itinerary (the ratio decidendi) followed by the tribunal, i.e., is unable to understand how the tribunal has proceeded from one point to the other and eventually to its conclusions.⁹

ii. The award fails to identify at the outset the issues to be decided.

iii. The award does not allow one to infer whether the tribunal has duly considered both parties’ positions on points of fact or of law, including the determination of the applicable law when this is a disputed issue.

iv. The award does not permit one to ascertain whether the tribunal has taken into consideration only the evidence provided by the parties, as it should, or whether it was influenced by pieces of information known to the arbitrators but unknown to the parties, or never pleaded during the proceedings, or whether it simply failed to consider and assess the evidence provided by the parties.¹⁰

v. The award does not allow one to understand whether the tribunal has applied the agreed governing law, or what law it has applied, if any.

vi. The award does not allow one to understand whether the tribunal has exceeded its mission by omitting a relevant issue (infra petita), or deciding an issue which was not referred by the parties to the tribunal (extra petita), in short whether the final decision exactly mirrors the claims and/or counterclaims as lodged by the parties.

vii. The award does not clarify whether the contractual provisions from which the dispute arises, and the parties have discussed before the tribunal treating them as relevant to the issues in dispute, have been taken into account or applied.

viii. The award simply mentions the contract clauses that the tribunal has read but fails to address the tribunal’s findings concerning their interpretation, or the reasons for those findings.

ix. The reasoning does not show internal consistency and does not appear congruent with deliberation process recorded in the award.¹¹

x. The award leaves unclear whether the claims or objections submitted by the parties have been understood or misunderstood.

xi. Even more importantly, the award leaves unclear whether the decision is influenced by a legal or contractual interpretation internally debated within the tribunal ex officio at the deliberation stage, however never pleaded by the parties before the tribunal, thus breaching the principle of contradictoire.

3. SOME CONCLUSIVE REMARKS

To close my introduction, I offer to readers the following preliminary conclusions.

3.1 In the search for an adequacy standard

I take the liberty to share what I consider to be the best definition of adequate reasoning. One was given by the Soufraki Ad Hoc Committee when requested to annul a previous award under Article 52 (1) of the ICSID Convention. According to this annulment committee, the expression of reasons is in substance

the only way by which compliance with the fundamental prohibition of manifest excess of powers and with the critical duty to apply the proper law may be observed… the more lucid and explicit the reasons set out by a Tribunal, the easier it should be to observe what a Tribunal is in fact doing by way of compliance.¹²

The second was given by another ICSID Ad Hoc Committee in M.I.N.E., according to which, "as long as the award enables the reader to follow how the Tribunal proceeded from point A. to point B. and eventually to its conclusions", the requirement to state reasons is at least satisfied at its minimum level and the award stands.¹³

Thus, a reasoning which has evident lacunae on important points of fact or law, prevents one from understanding the Tribunal’s ratio decidendi. The lacuna makes it impossible to understand how the Tribunal arrived at certain conclusions instead of others. It particularly prevents checking whether the Tribunal has applied the law that it had to apply, or whether it was inspired by self-made considerations, manifestly exceeding its mission. In such conditions, the award cannot stand.

3.2 Avoiding arbitrariness

A reasoning which is laconic in some parts and opaque or puzzling in others, does not guarantee that the award is based on objective and rational grounds. On the contrary, it sheds doubts on the possibility that the Tribunal’s findings be arbitrary on points of fact or of law, including the determination of the applicable law. This is why providing express reasons is the only means to test the arbitrators’ compliance with the fundamental duty not to exceed their powers and to safeguard due process, as lucidly stated in the Soufraki decision on annulment.

Personally, I would never advise a party to accept renouncing to the reasoning, which is tantamount to empowering an arbitral tribunal to adjudicate "as it deems fit". Too many risks are run when taking a similar approach, first and foremost, the absolute arbitrariness of the decision in plain disregard of the parties’ positions as argued before the Tribunal.

3.3 The educative function of the reasoning—contribution to improving the law

At a time when the arbitration community was still discussing whether to favor reasoned awards or unreasoned awards, Professor Carbonneau stated, quite rightly in my view, that reasoning in the awards contributes to form "a better basis for the elaboration of a common law of international transactions than national court decisions".¹⁴

Justice Bingham and Toby Landau share the above opinion and add a further benefit potentially provided by a good and credible reasoning. They, in particular, agree that the giving of reasons may provide an essential guide not only to the disputing parties in the individual case, but also to all other interested parties involved in future similar disputes to allow them to "learn from the forensic experience of others".¹⁵

I find this view attractive: qualifying the reasoning as a guide for future conduct is, at the same time, a lesson to the arbitrators soliciting their utmost care in drafting the reasons, and a lesson imparted to arbitration practitioners in comparable situations. In investment arbitrations, this is the regular case: as we know, awards in investment disputes are frequently crafted in a didactic way to progressively create uniform international investment law, but also some important commercial arbitral awards are quoted as "leading" or "guiding" decisions in subsequent awards or in doctrinal comments.

---

---