I. INTRODUCTION

Traditionally, the reasoning of awards rendered in international commercial disputes is expected to meet a standard that is sufficient to explain the basis upon which the tribunal decided. The purpose is usually to convince the parties that the decision is sound. Another purpose is to make sure that the award resists any scrutiny by the courts, either if the validity of the award is challenged or in case of enforcement. Increasingly, however, the reasoning of international arbitral awards is expected to meet further criteria that are not necessarily compatible with each other.

On the one hand, the significant attention devoted in the international arbitration community to investment arbitration, and particularly the regular publication and analysis of most arbitral awards that are rendered in investment disputes, seems to be exercising a force of attraction on the style in which commercial awards are written. Due to the presence of general interests in investment disputes, and certainly also due to the publicity that surrounds most investment awards, such investment awards tend to contain lengthy and thorough reasons. Investment awards are written with the awareness that they are going to be studied and discussed in the legal community, and used as precedents. Therefore, investment tribunals devote considerable attention to the compilation of reasons that not only solve the dispute at hand but may also be useful to the development of the law. This also contributes to the proliferation of dissenting opinions. Commercial awards, on the contrary, are traditionally confidential and not meant to be perused by anyone else but the parties. However, the calls for transparency are increasingly affecting commercial arbitration as well, and the publication of commercial awards, albeit not as systematic as it is the case for investment awards, is considerably more widespread than in the past. The prospects that the award will be analysed in the legal community is an incitement to write extensive reasoning that may have general interest and increase the possibility for the award to be used as a precedent.

On the other hand, the growing desire to enhance the efficiency of arbitral proceedings seems to suggest that even the need that awards state reasons is questioned. Various forms of expedited arbitration are being developed, and in some instances the reasoning is one of the features deemed dispensable for the sake of enhancing efficiency.

In summary, the reasoning in commercial arbitral awards is under a double and inconsistent pressure: on the one hand, an expansive pressure aiming at making awards with general relevance and potential of being used as precedent; on the other hand, a restrictive pressure aiming at enhancing efficiency.

In evaluating how to state the reasons of an award, in my opinion, one should not lose sight of the legal constraints that may affect the validity and enforceability of the award. Court control on the validity of awards and court assistance to enforce awards represent an indispensable legal framework, and awards should state reasons to the extent that reasons are required to permit the exercise of court control. Anything less is not advisable, because it leads to rendering awards that potentially may be set aside or refused enforcement. Anything more is not necessary for the purpose of ensuring the validity and enforceability of an award. It is disputable whether it is useful, for the purpose of ensuring the credibility of arbitration, to exceed the standard for reasons required for passing court scrutiny: arguably, an award with reasons that are sufficient to meet the requirements for court control should be sufficient to satisfy the standard for credibility that is expected of a dispute resolution method.

In this paper, I will go through the requirements that awards have to meet in order to resist court scrutiny, and I will make some considerations on what is needed and what is advisable or not advisable in terms of reasoning.

II. REASONING OF THE AWARD AND COURT CONTROL

A court who controls the validity or the enforceability of an award derives its jurisdiction from the applicable law. In case of challenge to the award’s validity, the applicable law is the arbitration law prevailing at the place of arbitration. In case of the award’s enforcement, the applicable law is, in the 159 countries having ratified it, ¹ the New York Convention. ²

National arbitration law differs from country to country, therefore the court’s jurisdiction in case of challenge to the award’s validity needs to be analysed on the basis of the national law that is applicable in the specific case. For the sake of simplicity, however, we will assume here that the national arbitration law corresponds to the 1985 UNCITRAL Model Law on International Commercial Arbitration (revised in 2006),³ as it does, more or less literally, in the 80 countries that are considered to be "Model Law countries".⁴

As the court’s powers, to the extent that is relevant here, are equivalent in the New York Convention and in the Model Law,⁵ the analysis can be made without distinguishing between court control carried out in connection with challenge to the award’s validity and court control carried out in connection with enforcement of the award. However, it should be emphasised that, if the law applicable in a specific case does not belong to a Model Law country, or if it implements the Model Law with some discrepancies from the original, it will be necessary to verify whether the law applicable to annulment has diverging regulation.

The applicable provisions make it clear that court control is not meant to be an appeal. Court control is not the same as a review of the award on the merits, neither in respect of the assessment of facts nor in respect of the application of law. The direct consequence of this limitation of court control is that an award is final and binding, even if it contains errors of fact or errors of law. This is the basis upon which the system of arbitration, as we know it today, rests: international conventions, national laws, courts of law, legal doctrine and practitioners support the aim that arbitration is to be an effective and efficient means of dispute resolution. To achieve this aim, they widely recognise that awards must be final and binding. Effectiveness and efficiency of arbitration are important principles of arbitration law, and at the origin of the widespread arbitration-friendly attitude that has characterised legislation and case law in the past decades. This seems to suggest that the controlling court shall not be concerned with the award’s reasons: even if the reasons show that the arbitral tribunal got the facts or the law wrongly, the court does not have the authority to react.

However, as will be seen below, the award’s reasons may be useful or even necessary in the context of court control. The provisions regulating court control on arbitral awards represent the limit of tolerance that legal systems have in respect of arbitral awards. As seen above, an award containing errors of fact or errors of law shall be confirmed as valid and shall be enforced.

However, an award rendered by an arbitral tribunal whose jurisdiction did not rest on a valid and binding arbitration agreement is not valid (article 34(2)(a)(i) of the Model Law) and not enforceable (article 36(1)(a)(i) of the Model Law and article V(1)(a) of the New York Convention); an award rendered as a result of a proceeding that did not give each of the parties the possibility to present its case is not valid (article 34(2)(a)(ii) of the Model Law) and not enforceable (article 36(1)(a)(ii) of the Model Law and article V(1)(b) of the New York Convention); an award rendered in excess of the jurisdiction granted on the arbitral tribunal is not valid (article 34(2)(a)(iii) of the Model Law) and not enforceable (article 36(1)(a)(iii) of the Model Law and article V(1)(c) of the New York Convention); an award rendered by an arbitral tribunal that was not constituted in accordance with the parties’ agreement or the applicable law, or as a result of proceedings that did not comply with the parties’ agreement or the applicable procedural rules is not valid (article 34(2)(a)(iv) of the Model Law) and not enforceable (article 36(1)(a)(iv) of the Model Law and article V(1)(d) of the New York Convention); an award rendered on a non-arbitrable subject-matter is not valid (article 34(2)(b)(i) of the Model Law) and not enforceable (article 36(1) (b)(i) of the Model Law and article V(2)(a) of the New York Convention); an award infringing fundamental principles (public policy) is not valid (article 34(2)(b)(ii) of the Model Law) and not enforceable (article 36(1)(b)(ii) of the Model Law and article V(2)(b) of the New York Convention).

There is a certain tension between the principle that court control is not a review of the award on the merits, on one hand, and the courts’ power to set aside an award or refuse its enforcement, on the other hand.

This becomes clear particularly when the court exercises control on a matter that already has been considered by the arbitral tribunal. As was seen above, there is an exhaustive list of issues the court may evaluate: whether the arbitration agreement was valid or the parties had legal capacity to enter into it, whether the principle of due process was respected, whether the tribunal’s constitution or procedure was in accordance with the parties’ will and the applicable rules, whether the tribunal exceeded its power, whether the object of the dispute was arbitrable and whether the award infringes public policy. If the tribunal has not considered the matter at all, the tension does not become evident: the court exercises its power to determine these issues, and this does not interfere with an evaluation already made by the tribunal. It may interfere with the award if the outcome is that the award is set aside or not enforced. However, it does not interfere with the tribunal’s evaluation of the particular issues regarding the validity of the arbitration agreement, the parties’ legal capacity, etc., because the tribunal has not evaluated these issues.

All the above-mentioned issues underlying the courts’ power to control arbitral awards may, however, conceivably have been already evaluated by the arbitral tribunal. The tribunal may have considered whether the arbitration agreement met the applicable form requirements or whether a party had legal capacity to enter into it, and it may have concluded in the affirmative, thus proceeding to solving the dispute in the merits and rendering an award. Yet the courts may have a different opinion of the same issues and may conclude that the award shall be set aside or refused enforcement. The same reasoning may be made in respect of the other grounds for setting aside or refusing enforcement: the tribunal may have considered its constitution, the procedure followed under the dispute, the scope of its power, the arbitrability of the disputed object, or the conformity of the award with fundamental principles; the tribunal may have concluded that there were no violations. Yet the court may have a different opinion, and it may exercise its power to set aside the award or refuse its enforcement.

The above-mentioned tension between the principle of the award’s finality and court control becomes, therefore, particularly evident in case of concurrent, and diverging, evaluations of the same issue carried out by the tribunal and by the court. Is the court reviewing the arbitral tribunal’s evaluation of these issues? This is at odds with the principle of no review of the award in the merits. Is the court independently carrying out its own examination of these issues? This is compatible with the court’s power to exercise control on the award. The court’s activity is the same in both situations, but the description of this activity is different and leads to different results. Where does the border go between these two conflicting descriptions of the same court activity? And which role do the reasons stated in the award play?

From the foregoing it appears that the award’s reasons are necessary to determine whether the arbitral tribunal has examined an issue. If the matter was not addressed in the award, and if it falls within the above-mentioned grounds for annulling or refusing enforcement, the court will independently consider the matter and decide the issue. The reasons of the award, therefore, are necessary to prevent annulment of refusal of enforcement on the basis that the arbitral tribunal did not consider issues such as the validity of the arbitration agreement, the legal capacity of the parties, the respect of due process, the regulatory of the procedure or of the constitution of the arbitral tribunal, the scope of power conferred on the arbitral tribunal by the parties, the arbitrability of the dispute or the compatibility of the award with public policy. That the arbitral tribunal did not evaluate these issues is not, in itself, a ground for annulling the award or refusing its enforcement, but the court will have no basis upon which to develop its independent evaluation of these issues. If the court, on the basis of its evaluation, determines that the award is not in compliance with the relevant criteria, it will set aside the award or refuse its enforcement.

But what happens if the arbitral tribunal has considered these issues, has written a reasoned award that explains the arbitral tribunal’s views thereon, and the court disagrees with the award’s reasoning?

The matter has been perhaps mostly discussed in connection with two of the above-mentioned issues: whether the arbitral tribunal has jurisdiction to hear the dispute (particularly in connection with whether there is a valid and binding arbitration agreement), and whether the award violates public policy. There seems to be an inconsistency in the answer to the question, depending on the context in which it arises. This, and the significance that it has for the reasons of the award, will be addressed in sections IV. and V. below.

The controlling court does not necessarily direct its attention to the award’s reasons: with the exception of the minimalist doctrine relating to the public policy rule, that will be discussed in section V. below, the court forms its own opinion of whether the award meets the above-mentioned criteria for validity and enforceability. In theory, therefore, a court does not need to read the award’s reasons. However, as will be seen in section VI., knowing the reasoning upon which the tribunal’s decision was based is necessary when the court has to decide whether the tribunal exceeded its power. The award’s reasons play a role also in respect of the remaining criteria for validity and enforceability, discussed in sections IV., VII. and VIII. below, as well as the maximalist doctrine relating to the public policy rule, that will be discussed in section V. The award’s reasons provide a useful insight into the factual elements, the legal sources and the arguments that led to the decision. If the award did not state any reasons, it is not certain that the court would have in front of it all elements useful to make its own determination of the validity and enforceability of the award. Therefore, the award’s reasons contribute to a sound court control by permitting the court to examine the elements that were considered by the arbitral tribunal.

III. REASONING OF THE AWARD AND CREDIBILITY OF ARBITRATION

In addition to being necessary or useful to the exercise of court control, the reasons of the award have an important function in respect of the credibility of arbitration as a mechanism for settling disputes.

As known, arbitration is often praised for being an autonomous dispute resolution mechanism that is based on the parties’ will, that is only minimally subject to national laws, and only exceptionally subject to court control. This suggests that it is completely up to the parties or the tribunal to decide whether and to what extent to write reasons, as long as the (few) requirements for validity and enforceability are met. This is confirmed by the circumstance that under some sources, such as the Model Law, parties may agree that the award shall not have reasons. As will be seen in section IX. below, however, this possibility creates some difficulties when the lack of reasons prevents a proper court control.

Quite irrespective of the importance of the award’s reasons for the exercise of court control, it is my opinion that permitting an award to be without reasons may affirm the autonomy of arbitration in the specific case, but in the long run it creates the risk that arbitration loses its effectiveness. An award without reasons is meant to create legal effects for the parties and must be enforced by the courts, but at the same time gives no basis to determine whether the decision was sound or arbitrary. By not stating reasons, the tribunal does not explain the legal basis for its order, the evidence upon which it has reached its conclusion, etc. Courts are expected to nevertheless recognise and enforce the decision, although they are not capable of verifying the basis upon which the decision was taken. While it is true that the effectiveness of arbitration requires that courts do not review the merits of the award, depriving the award of the basic elements to ensure a minimum of transparency seems to be an unnecessary statement of autonomy. As long as arbitration is meant as a way of solving disputes on legal terms, it should comply with the essence of an adjudicative system, which requires that the basis for the decision is known or understandable.⁶ It has been suggested that, if the parties wish to avoid that the arbitral tribunal be constrained by legal rules, including the necessity to write reasons for the purpose of safeguarding the quality of the adjudication process, they may instruct the tribunal to decide ex aequo et bono.⁷

By permitting that awards have the legal effects of a court decision, but without having to comply with the most basic notions of the adjudication process, the risk is that the trust in the arbitration system will be eroded, and that this leads courts being more restrictive in determining which disputes may be subject to arbitration.

As known, lack of arbitrability is a ground for setting aside an award and for refusing its enforcement; and what is arbitrable is determined by the courts under their national law. Even the New York Convention specifies, in article V (2) (a), that arbitrability is determined under the court’s law. Therefore, it is relatively easy for a court to restrict the scope of what is arbitrable—easier than if arbitrability was a category defined autonomously on an international level.

A parallel could be drawn with the relationship between the possibility to exercise court control and the scope of arbitrability. Elsewhere⁸ I have pointed out that restricting the scope of court control creates the risk of reducing the scope of arbitrability. The close link between court control and arbitrability is apparent in some opinions by the CJEU Advocate General ("AG"). In CDC, ⁹ AG Jaaskinen argued for restricting arbitrability of matters relating to competition law, because arbitration, that is not subject to the competence of the CJEU, does not ensure a uniform application of EU law. Similarly, in Genentech,¹⁰ AG Wathelet argued against the minimalist approach to court control. As will be discussed in section V. below, according to the minimalist approach, courts owe deference to the tribunal’s evaluation and may not independently verify the compatibility of the award with public policy. In Achmea,¹¹ AG Wathelet accepted the arbitrability of disputes relating to EU law, but on the assumption that courts exercise control on the awards according to the maximalist theory.

In its final judgments on the former two cases, the CJEU ignored the matter. The latter mentioned case, Achmea, was based on a referral by the German Supreme Court (BGH) ¹² relating to the annulment proceeding of an investment award.¹³ One of the invoked annulment grounds was that the award was null because the dispute was not arbitrable: as arbitral tribunals are not bound by the EU duty to apply EU law in a uniform way, the effective application of EU law would be endangered if the dispute had been arbitrable. This line of thought resembled the situation prior to the seminal US case Mitsubishi.¹⁴ Prior to Mitsubishi, US courts excluded arbitrability whenever the issues in dispute assumed the accurate application of norms reflecting important policies, such as competition law. In Mitsubishi, the court affirmed that disputes regarding competition law may be subject to arbitration, as long as the court has the possibility to exercise control on the award (the so-called "second look doctrine"). The Mitsubishi doctrine has informed an era of ever increasing room for arbitrability, both in the USA and in Europe. In Achmea, the BGH requested the CJEU to confirm that there is no basis to restrict the scope of arbitrability, as long as the courts may control the award’s compatibility with fundamental principles of the forum. Thus, the BGH embraced the second look doctrine introduced by Mitsubishi and endorsed the maximalist theory. AG Wathelet¹⁵ concurred with this line of thought. The CJEU,¹⁶ however, did not accept this approach, and concluded that investment disputes are not arbitrable.¹⁷ However, the CJEU distinguished between investment disputes and commercial disputes, and specified that its conclusion did not apply to commercial disputes. Also this time, therefore, for commercial arbitration the CJEU did not clarify the extent of court control that it expects for it to permit arbitrability of matters related to EU law. The CJEU, however, seemed to indirectly endorse, as an obiter dictum, the AG’s assumption that, in controlling commercial arbitral awards, courts should follow the maximalist approach. As I have explained elsewhere,¹⁸ however, the distinction made by the CJEU between investment and commercial arbitration is not convincing. There may be, therefore, reasons to fear that the CJEU turns towards a restrictive approach to arbitrability also in respect of commercial arbitration.

This would confirm a disturbing trend that can be detected in national courts. Courts in EU states such as Austria,¹⁹ Belgium,²⁰ Germany²¹ and England,²² denied the arbitrability²³ of disputes regarding contracts of commercial agency. EU agency law is deemed to be necessary for the achievement of the internal market. Hence, some courts have affirmed that disputes concerning commercial agency should be decided by courts of EU member states: choosing a court outside the EU, or choosing arbitration, may endanger the effective enforcement of EU law. This shows that domestic courts in a variety of EU member states are assuming a restrictive approach towards arbitrability. The link between court control and arbitrability is evident.

The observations made in respect of the correlation between the possibility to exercise court control and the scope of arbitrability may be extended to the matter that is relevant here, i.e. awards without reasons. Without touching the cornerstone principle that arbitral awards may not be reviewed on the merits by the controlling courts, it does not seem to be advantageous to the credibility of arbitration that awards are rendered without any explanation of the elements and reasoning that led to the decision. Quite to the contrary, the calls for transparency that have recently turned round the world of investment arbitration²⁴ seem to be affecting also commercial arbitration.²⁵ Commercial awards are being published to a larger extent than earlier,²⁶ institutions are being more transparent regarding the appointment and removal of arbitral tribunals,²⁷ and initiatives flourish to increase transparency in commercial arbitration.²⁸ This suggests that the general expectations towards arbitration are not going in the direction of releasing arbitral tribunals from any obligation to state the reasons for their decisions. The intention, when permitting that an award does not state reasons, is to enhance arbitration and increase its effectiveness. The risk is that exactly the opposite result is obtained, i.e. that arbitration loses its credibility and, instead of being promoted, is restricted.

IV. REASONING AND COURT CONTROL: JURISDICTION

One of the grounds for setting aside an award or refusing its enforcement is that the arbitral tribunal did not have jurisdiction. In connection with the arbitral tribunal’s jurisdiction, particularly the existence and validity of the arbitration agreement, the doctrine of Kompetenz-Kompetenz was developed. According to this doctrine, an arbitral tribunal has the competence to decide on its own competence.²⁹ The main implication of this doctrine is that a tribunal does not have to suspend the proceeding in case the validity of the arbitration agreement is questioned. The arbitral tribunal has the power to make a decision on the existence and validity of the arbitration agreement and, if the decision is in the affirmative, the tribunal may proceed with the substantial aspects of the dispute.

This principle can be found in article 16 of the UNCITRAL Model Law for instance. A system that even more clearly gives priority to the arbitral tribunal’s evaluation of its competence is France, where the so-called effet négatif de la compétence-compétence was developed.³⁰ According to this theory, courts must refer the dispute to arbitration whenever they are seized with a dispute which is subject to an arbitration agreement. Also under the UNCITRAL Model Law a court must refer the dispute to arbitration if there is an arbitration agreement. However, in the wording of article 8 of the Model Law, the court refers to arbitration "unless it finds that the agreement is null and void, inoperative or incapable of being performed." This wording opens for a thorough examination by the court of the existence, validity and effectiveness of the arbitration agreement. The French Civil Code of Procedure goes further and restricts the court’s examination.³¹ The only possibility courts have at this stage, is to make a cursory review of the arbitration agreement. If the court is prima facie satisfied that the arbitration agreement exists and is valid, it shall refer the dispute to arbitration. The underlying idea is that it is for the arbitral tribunal to make a deeper evaluation of its competence.

What do the Model Law and the French effet négatif de la compétencecompétence provide as to the issue raised in section II. above, i.e. the effects for the court of the tribunal’s decision on its own competence?

The same issue of competence that was decided by the tribunal may be put forward for the purpose of challenging the validity of the award or of preventing its enforcement. The Model Law says in articles 34(2)(a)(i) that the court may set aside an award if it finds that the arbitration agreement did not exist or was invalid, or that a party was under some incapacity. The same can be said for French law: article 1492 No 1 of the Civil Procedure Code gives the court the power to set aside the award if the arbitral tribunal did not have jurisdiction. The same may be said regarding article 36(1)(a)(i) of the Model Law and article V(1)(a) of the New York Convention, in respect of enforcement. Neither the Model Law, nor the French Civil Procedure Code or the New York Convention, however, explain the relationship between the tribunal’s competence to decide on its own competence, and the court’s power to evaluate the validity of the arbitration agreement or the legal capacity of the parties.

According to the prevailing doctrine, the court retains its power to determine the existence and validity of the arbitration agreement, or the parties’ capacity to enter into it, even if the tribunal already has evaluated the matter.³² This may result in a different outcome from the one to which the tribunal came and may lead to setting aside the award or refusing its enforcement. In practice, this means that the award has no preclusive effect and the mentioned issues ultimately are subject to the court’s evaluation. This approach is supported also in France.³³ The theories of Kompetenz- Kompetenz and of l’effet négatif de la compétence-compétence, which were developed to enhance the autonomy and thus the efficiency of arbitration, do not go as far as to affirm that the tribunal’s determination of the existence and validity of the arbitration agreement are final and the court owes deference to the tribunal’s determination.

The reasons stated in the award, therefore, are important to permit the court to verify on what basis the arbitral tribunal assumed jurisdiction. However, the court is not bound by the tribunal’s reasoning. This does not mean that reasons are useless and may be dispensed of: by explaining the reasons for its decision on jurisdiction, the arbitral tribunal supplies the factual and legal basis upon which the decision was made, the arguments that were examined and the reasons for discarding the opposed arguments. While not bound by the tribunal’s reasoning, a court will have a more complete basis for developing its own reasoning than if it had not had the benefit of the tribunal’s reasons.

V. REASONING AND COURT CONTROL: PUBLIC POLICY

One of the other grounds for setting aside an award or refusing its enforcement is that the award infringes public policy. In the Model Law and in the New York Convention this ground is dealt with in the same articles as the ground relating to the arbitral tribunal’s jurisdiction, that was discussed in section IV. above: challenge to the validity is regulated in article 34(2)(b) (ii) of the Model Law and refusal of enforcement in articles 36(1)(b)(ii) of the Model Law and V(2)(b) of the New York Convention.

The only difference between the provision on jurisdiction and that on public policy suggests that conflict with public policy is considered to be a more serious defect of the award than the wrong determination by the tribunal of its jurisdiction. While the ground relating to jurisdiction has to be raised by one party, the ground relating to public policy can be raised ex officio by the court. By giving the court the power to consider the matter of public policy ex officio, the applicable sources show that it is not possible to delegate to the parties the decision of whether the issue of conformity with public policy shall be considered. Moreover, it should not be possible to delegate to the tribunal the determination of whether or not public policy was infringed. The logical consequence is that the court’s power to exercise its control notwithstanding the arbitral tribunal’s determination of the same issue is at least equally preserved in respect of both public policy and the tribunal’s jurisdiction. As was seen in section IV. above, the award does not have preclusive effects in respect of the tribunal’s determination of its own competence. Similarly, there should be no preclusive effects in respect of the tribunal’s determination of conformity with public policy.

However, in connection with public policy there is no unitary approach to the effects for the court of the tribunal’s determination. Two opposed doctrines were developed to define the degree of control that courts may exercise on the award’s conformity with public policy.³⁴ The Paris Court of Appeal³⁵ developed the minimalist doctrine, according to which courts owe deference to the tribunal’s evaluation. The Dutch Court of Appeal³⁶ developed the maximalist doctrine, according to which courts may independently evaluate whether ordre public is infringed. The maximalist approach has effects that are comparable with the effects recognised by the doctrine of compétence-compétence and of l’effet négatif de la compétence-compétence; according to this approach, the court may carry out its independent evaluation of the issue. The minimalist approach goes further in affirming the finality of arbitral awards, and assumes that the tribunal’s determination of the public policy issue has preclusive effects for the court.

For both approaches, the starting point is that court control is not meant to re-open the dispute that was decided by the award. In particular, the court may not review the arbitral tribunal’s assessment of facts, evaluation of evidence and application of law. When the court has to determine whether the award is compatible with public policy, however, a question arises as to how far the court may go in its evaluation.

According to the maximalist approach, the court may independently evaluate whether the award leads to a result that violates public policy, irrespective of whether the arbitral tribunal already has considered the same matter. This means that the court may independently evaluate the evidence that was already evaluated by the arbitral tribunal and may form its own opinion of the disputed facts. Furthermore, the court may independently evaluate how the law shall be applied. All this is done solely for the purpose of ascertaining whether the award violates public policy. This is not made for the purpose of reviewing whether the tribunal correctly interpreted the evidence or applied the law.

The minimalist approach assumes that the court shall limit itself to verifying whether the arbitral tribunal has considered the matter. If the arbitral tribunal has concluded that public policy was not violated, the court has to accept this conclusion. Hence, according to the minimalist approach, the arbitral tribunal’s evaluation of whether the award is compatible with public policy is binding on the court. The award has, therefore, preclusive effect.

The minimalist approach is particularly represented in France.³⁷ However, French case law recently seems to have embraced the maximalist approach, at least in areas such as corruption and money laundering.³⁸

As I argue in section III. above, the maximalist approach is to be preferred, as it is more compatible with the rationale of court control. By postulating that the court owes deference to the evaluation that the tribunal made of the conformity of the award with public policy, the minimalist theory effectively delegates to the arbitral tribunal the assessment of this ground for annulment and for refusing enforcement. Thus, the minimalist approach effectively deprives court control of any possibility to exercise its functions. In my opinion, this seriously undermines the credibility of arbitration as dispute resolution mechanism, and creates a significant risk that the scope of arbitrability is restricted. As discussed in section III. above, the first signs may be seen in connection with disputes regarding EU law.

What role does this inconsistent approach play for the award’s reasons?

In the minimalist approach, reasons are crucial: if the only role for the court is to determine whether the arbitral tribunal considered the matter, it is necessary that the award states reasons reflecting that the matter has been considered.

Also in the maximalist approach reasons are important. They will not prevent the court’s independent consideration of the matter, but they will give an important starting point for the court’s evaluation. It is generally recognised that the court’s power to set aside an award or refuse its enforcement on the ground of public policy shall be exercised restrictively: ³⁹ only fundamental principles qualify as principles of public policy, and only serious infringements of these principles justify setting aside an award or refusing its enforcement. If it is necessary to initiate a full-fledged inquiry to ascertain whether an award infringes, for example, competition law,⁴⁰ it could be argued that the infringement is not so serious as to justify applying the public policy ground.

It has been suggested⁴¹ that it should be possible to exercise court control by examining, in some detail, the reasoning of the award. Only in exceptional cases, such as when the award has no reasons, or the award did not consider the applicability of public policy rules, should the court be allowed to go further and examine the parties’ pleadings or the evidence produced in the arbitral proceedings or, in extreme cases, to launch a full-fledged investigation. This reflects the minimalist approach.

I can subscribe to this scale of court control’s intensity, with one addition: in order to safeguard the efficacy of the public policy rule, I would add that the court may go further and examine the pleadings and the evidence also when the court does not find the award’s reasoning convincing. With this addition, the intensity of court control corresponds to the criteria laid down by the maximalist theory. If the examination of the award or the examination of the evidence and documentation upon which the award relies gives the court reason to conclude that competition law was infringed, and that this infringement is so serious that it affects public policy, the award may be set aside or refused enforcement.

The maximalist approach is not meant to give the court the power to review the tribunal’s decision. As was seen above, public policy is not violated simply because the award has wrongly applied the governing law. Even when the allegedly incorrectly applied provisions are mandatory, there is no automatic effect on public policy. Public policy is affected only if the result of the award seriously infringes fundamental values in the socio-economic system. It is, in other words, not the technical content of a legal rule that may constitute public policy, but the underlying principles. The narrow scope of the public policy rule, therefore, prevents that court control becomes a review of the merits: an award may not be set aside simply because the tribunal did not accurately apply certain rules of law.

In summary, the reasons of the award are indispensable to court control in respect of public policy if the court applies the minimalist theory. The simple circumstance that the reasons make reference to the arbitral tribunal’s public policy evaluation is, under the minimalist approach, sufficient to exclude the court’s authority to consider that issue. If the court applies the maximalist theory, similarly to what was said in respect of jurisdiction, the reasons are not indispensable, because the court develops its own independent evaluation of the issue. However, reasons are important because they supply the full set of arguments and evaluations upon which the award is based. Thus, the award’s reasons have an important function in that they permit the court to make an informed assessment of the public policy issue, taking into due consideration the elements that were considered by the arbitral tribunal.

VI. REASONING AND COURT CONTROL: EXCESS OF POWER

Another situation in which the award’s reasons are crucial, when the award is challenged or enforcement is resisted, is when the losing party alleges that the arbitral tribunal went beyond the scope of its power. The main purpose of the rule permitting annulment or refusing enforcement on the ground of excess of power is to ensure that the award is rendered within the borders set by the arbitration agreement and by the parties’ pleadings. An award that goes beyond those borders is an award not based on the parties’ consent to arbitrate, and is therefore invalid and unenforceable. Generally, this ground is considered to apply when the arbitral tribunal decides on matters that were not covered by the arbitration agreement or were not raised by the parties, or when it otherwise goes beyond the facts that were referred to by the parties. As an illustration may be mentioned an award ordering one party to set off its claims against claims that the other party has under a separate contract that was not subject to the arbitration agreement under which the arbitral tribunal was appointed.

This ground for annulment and for refusing enforcement applies first of all to excess in the strict sense, meaning that the tribunal decided matters beyond the borders of the dispute as was set by the parties (ultra petita partium). This ground, however, can apply also to the opposite situation, in which the tribunal does not consider issues presented by the parties (infra petita partium). While some arbitration laws, such as the UK Arbitration Act⁴² and the Swiss Private International Law Act⁴³ expressly mention the invalidity of an award that is infra petita, the Model Law and the New York Convention are silent on the matter. However, case law and literature extend the ground of excess of power also to this scenario.⁴⁴

Such an excess of power may occur in connection with the factual circumstances that the tribunal relied on, in connection with the remedies that the tribunal ordered, or in connection with the legal reasoning made by the tribunal. In all these situations, the reasons stated in the award shed light on the basis for the decision and on the relationship between the basis for the decision and the parties’ pleadings. Without a reasoned award, it would be impossible for the court to understand whether the tribunal has considered the parties’ pleadings and on what basis the tribunal founds its decision.

The importance of the award’s reasons may be seen particularly in respect of the legal basis on which the award relies. Often, the defence of excess of power is considered to be irrelevant to the questions of law, as this is deemed to be a question relating to the merits of the dispute (and therefore immune from court control), rather than to the scope of the arbitral tribunal’s power.⁴⁵ In my opinion, to the contrary, the applicable law is certainly relevant to the scope of the dispute: the scope of the dispute is mainly determined by the disputed contract, but the contract does not have legal effects simply in force of itself—it receives its legal effects from the governing law. As I have discussed elsewhere, ⁴⁶ the same contract wording may have dramatically different effects depending on which law governs it. It may be argued, therefore, that the scope of the dispute is determined by the contract as construed under its applicable law. An arbitral tribunal that applies legal sources different from the law chosen by the parties in the contract may be equal to a tribunal that applies a contract wording different from the wording contained in the disputed contract. Hence, the tribunal may be deemed to have gone beyond the powers that were conferred to it under the contract.⁴⁷

It should be pointed out, however, that in many legal systems, the arbitral tribunal is not expected to restrict its legal reasoning to the sources that were pleaded by the parties.⁴⁸ Therefore, the simple circumstance that a source was applied on the tribunal’s own initiative does not constitute an excess of power. However, the right to be heard may be infringed, as will be seen in section VII. below.

Furthermore, courts do not have any power to control the tribunal’s application of the law—with some notable exceptions, such as England and the United States. This becomes relevant particularly when it comes to the interpretation and application of the law in relation to the contract terms. A contract may contain terms that contradict the law that was chosen by the parties to govern it. For example, the contract may provide for payment of penalties in case of breach by a party of its obligations, and it may be governed by a law that does not consider contractual penalties to be enforceable. Whether the tribunal applies the law strictly even when this may contradict the terms of the contract, or whether it acts more flexibly and deems that contract terms must prevail even when they contradict the applicable law, is a question of the sensitivity of the particular tribunal.⁴⁹ The courts have no say, as long as this remains a question of merits, i.e. a question of the tribunal’s interpretation and construction of the contract and of its application of the law. The courts would have jurisdiction, however, if the arbitration agreement or the parties’ common instructions contained restrictions as to the tribunal’s power to apply the law. However, the parties usually do not attempt to restrict the tribunal’s power with regard to the impact that the governing law may have on the interpretation and construction of the contract.⁵⁰

However, the review of the tribunal’s application of the law, which courts are not allowed to make, should be distinguished from the review of whether the tribunal had the power to apply a certain law. The courts have the power to make this latter review.

In particular, this is important in situations where the dispute arises out of a contract that contains a clause choosing a certain governing law, but the arbitral award takes into consideration provisions from other laws as well. The typical example would be a contract between parties belonging to two EU States and having effects on the EU market, but containing a clause choosing as governing law the law of a non-EU State. If the contract violates the EU competition law, for example, the arbitral tribunal may be forced to consider EU competition law in order to avoid rendering an award that infringes public policy (see section V. above). Usually, competition law would be invoked as a defence by the party who did not comply with the contract. The argument would be that breach of the contractual obligation may not be sanctioned, because the obligation was invalid as it infringed EU-competition law. However, the arbitral tribunal may also raise the question of infringement of EU-competition law ex officio. If the arbitral tribunal applies a law different from the law chosen by the parties, however, it may be deemed to have exceeded its power. It is, therefore, necessary to verify whether the tribunal has the power to apply a law different from the law that was chosen by the parties, or whether this goes beyond the powers that the parties conferred on the tribunal.

Whether the arbitral tribunal exceeds its power or not, depends on the reasoning that led to disregarding the choice of law made in the contract. If the tribunal does not justify its application of a non-chosen law, there is a basis for considering the award to be in excess of the power that the parties conferred on the tribunal. However, where the legal reasoning gives a basis for applying a law different from the law chosen by the parties, there is no excess of power.⁵¹

As I have discussed elsewhere, there are at least two approaches: the traditional one, that is often frowned upon as being old fashioned and unfriendly to arbitration, but that in my opinion is the most arbitration friendly approach;⁵² and the modern one, that is often deemed to be the most progressive and arbitration-friendly, but that in my opinion creates unpredictable results and endangers the creditworthiness of arbitration as a dispute resolution method.⁵³ The traditional approach uses the private international law to explain the arbitral tribunal’s power to consider laws different from the law chosen by the parties. The modern approach considers that the tribunal has full discretion to consider any law, without the necessity to justify the selection of the law (the so-called voie directe).

According to the traditional approach, the parties’ or the arbitral tribunal’s choice of law is the exercise of a power regulated by that branch of the law that goes under the name of private international law (also known as conflict of laws, or choice of law rules).

In the field of international arbitration, some specific private international law rules are codified in international conventions,⁵⁴ whereas others are regulated in national arbitration law. Of particular interest in the trans national context may be the 2015 Principles of Choice of Law in International Contracts,⁵⁵ published by the Hague Conference and meant as a restatement of generally recognised principles of private international law.

Regarding the law applicable to the merits of the dispute, which is the relevant issue here, national arbitration law is harmonised by the UNCITRAL Model Law and contains a choice of law rule giving the parties the possibility to choose the applicable law.⁵⁶ In case the parties have not made use of their party autonomy, the UNCITRAL Model law directs the arbitral tribunal to apply the conflict rules that the tribunal "considers applicable".⁵⁷ The Model Law, thus, follows the traditional approach and applies the private international law mechanism. However, in the revision of 2006, the Model Law received a more flexible version and was emancipated from the automatic application of the private international law of the state where the tribunal has its venue. This was meant to cater for situations where the venue of arbitration has no connection with the dispute, and where it may be appropriate to apply the conflict rules of states with closer connection. In practice, in many situations, arbitral tribunals may still consider the venue of the arbitration as a significant connecting factor and thus apply its conflict rules to determine the law applicable to the merits of the dispute.⁵⁸

However, not all jurisdictions have adopted the Model Law’s reference to private international law (of the venue or otherwise): some jurisdictions provide a specific conflict rule for arbitration,⁵⁹ some contain no guidelines at all,⁶⁰ and some regulate the so-called voie directe,⁶¹ that does not rely on private international law and gives no criteria for the selection of the applicable law. It cannot be excluded that arbitral tribunals will make use of private international law even when the arbitration is subject to an arbitration law that does not regulate the selection of law or provides for the voie directe: although not required to apply conflict rules, arbitral tribunals may find that the well-known mechanism of private international law permits to select the law more objectively and predictably than mere discretion without guidelines. Even though conflict rules may be complicated and sometimes they may leave room for discretion, the abundant literature and case law on the area contribute to rendering its application quite predictable.

With the exception of the latter mentioned approach, the voie directe, (that is the starting point for the modern approach discussed below), this short overview shows that the private international law has still a quite important role in arbitration.⁶² Even those who deem it undesirable to apply a national system of private international law, can find acceptable the mechanism underlying private international law in generally acknowledged principles such as the Hague Principles on Choice of Law.⁶³

It is therefore justified to look at the private international law as the framework for determining the scope of the choice of law made by the parties and selected by the arbitral tribunal. As the choice of law is the exercise of a power based on private international law, any restrictions or exceptions that rely on principles or rules of private international law do not violate the parties’ choice of law; they simply implement the parties’ choice of law pursuant to its scope as determined in the legal sources upon which it is based.

One of the restrictions that are relevant to the arbitral tribunal’s power to consider a law different from the law chosen by the parties, is that the parties’ choice in private international law generally applies only within the scope of contract law (and to a certain extent also tort law). If a dispute has implications of company law or property law, or if it is on areas where states exercise their regulatory powers such as competition law (also known as overriding mandatory rules or lois de police), the court will apply the law chosen by the parties to the contractual issues of the dispute, and to the other aspects it will apply the law selected according to the appropriate conflict rules (so-called dépeçage).

It has been suggested that party autonomy in arbitration has a wider scope than party autonomy in private international law.⁶⁴ In arbitration law, often the parties are permitted to choose the law applicable not only to the contract, but more generally to the merits of the dispute—as an example, the UNCITRAL Model Law says that the parties may choose the law applicable to the substance of the dispute. This is interpreted as giving party autonomy a wider scope than the one party autonomy has in private international law as described above: if the dispute has implications that go beyond the mere contract law, according to this opinion party autonomy would also cover these aspects. According to this logic, therefore, the private international law would not be a sufficient basis to justify the arbitral tribunal’s power to override the parties’ choice of law: the parties’ choice of law would not be restricted to the mere contract matters, but would extend to any issues within the scope of the dispute.

However, although the wording of arbitration law may seem wider than, for example, the wording of the Rome I Regulation—according to which the parties may choose the law applicable not to the merits of the dispute, but only to contractual obligations—it may be questioned whether this can be taken as a basis for assuming that the parties in arbitration have the power to choose the law applicable to matters of company law, property law or regulatory matters. This is because the choice of law possible in arbitration relates to the merits of the dispute. Therefore, it must be interpreted within the scope of the dispute that may be decided by the arbitral tribunal. Generally, arbitration may decide disputes between the parties on rights and obligations that the parties may dispose of, and decides the dispute with effects for the parties. An arbitral tribunal may not render an award with effects for third parties: therefore, an arbitral award will not be empowered to decide that the resolution of a company body is invalid, or that a certain asset of the insolvent debtor is not available to the generality of the creditors. These aspects, being outside of the dispute, cannot be covered by the broad language of the conflict rule for arbitration. A dispute may have implications of, for example, company law, property law or competition law, but these issues will not be the direct object of the claim. They will be preliminary issues that the tribunal may consider in the course of its determination of the claim. For example, a claim for reimbursement of damages due to breach of contract may require that the tribunal examines whether the board resolution upon which the contract was signed was valid, or whether the contract was valid under competition law. The dispute, however, will not become a dispute of company law or of competition law: it will be a dispute of contract law, for the decision of which the tribunal must consider some preliminary issues. The scope of the dispute, and therefore its merits, are defined by the claim, and the claim is, in our example, a contractual claim. The wording of the choice of law rule contained in the Model Law, therefore, does not seem to extend the scope of party autonomy in a significant manner.

If the choice made by the parties is not unlimited, it becomes necessary to find a source regulating the scope of party autonomy. Article 28 (2) of the Model Law refers to private international law for the eventuality that the parties have not chosen the law. However, it is systematically correct to use the same reference also for the purpose of specifying the scope of party autonomy in article 28 (1). Where else should the source regulating the details of party autonomy be found? The Model Law itself does not give crucial elements such as the parameters for a valid exercise of party autonomy, the specification of its scope, or its effects. All these elements are to be found in the private international law, and it is only natural to refer to the private international law that is referred to in article 28 (2) also for the purpose of article 28 (1).

In summary, a systematic interpretation of the sources applicable to arbitration, including also the private international law, seems to give justification for the arbitral tribunal’s power to consider rules of competition law not belonging to the law that was chosen by the parties.⁶⁵

The relevance of private international law mechanisms in arbitration, however, is not uncontroversial. Private international law has been, in my opinion, unjustly considered an excessively rigid and old-fashioned mechanism detrimental to the effectiveness of arbitration. On this basis, it has been discarded in various jurisdictions in favour of more flexible approaches meant to enhance the prominence of party autonomy—such as the voie directe, contained, inter alia, in article 1511 of the French Code of Civil Procedure, in article 27a of the Swedish Arbitration Act, as well as in numerous Arbitration Rules;—including also the UNCITRAL Rules. Indeed, giving the arbitral tribunal full discretion to apply whatever sources the tribunal deems appropriate, provides the tribunal with unlimited flexibility. This, however, is not necessarily positive.

A first reason why a flexible approach is not necessarily desirable is that it may lead to unpredictable results. If the arbitral tribunal is not guided by objective criteria when it determines the applicable law, the parties are not in a position to assess in advance their respective legal positions and thus evaluate the advisability of initiating the arbitration. In some cases, determining the applicable law in advance is essential to the evaluation of whether to start an arbitral proceeding or not. For example, whether the applicable law will be Italian or Norwegian will decide whether a four years old claim is time barred or not, as the period of limitations is 10 years under Italian law and 3 years under Norwegian law. If the criteria for determining the applicable law are not objective, the parties are put in the paradoxical situation of having to initiate an arbitration to be able to assess whether there is at all a basis for initiating arbitration.

A second reason why a flexible approach to determining the applicable law is not necessarily desirable is that, without conflict of laws rules, the tribunal has no legal basis upon which it may restrict the choice of law made by the parties. This becomes relevant when the choice of law made by the parties leads to disregarding overriding mandatory rules or principles of public policy in the law(s) that would be applicable if the parties had not made a choice of law. While the traditional approach founds the arbitral tribunal’s selection of the applicable law on the private international law, the voie directe does not provide a legal basis. This may seem to suggest that party autonomy is not restricted by the private international law. Hence, this seems to suggest that party autonomy is unlimited. Therefore, not having a basis to restrict the parties’ choice, the tribunal may fear that the award will be invalid or unenforceable for excess of power if it does not follow the parties’ choice, for example by considering competition law. To justify the arbitral tribunal’s power to consider laws different from the law chosen by the parties, it has been suggested that the arbitral tribunal has an ethical duty to consider applicable overriding mandatory rules, also taking into account the necessity to preserve the credibility of arbitration.⁶⁶ The specific content of this ethical duty is not readily determinable, therefore it has been suggested that arbitral tribunals should exercise their discretion with a certain pragmatism and shall look at the criteria that would be applied by courts. As a result, it seems that this approach does not differ significantly from the traditional approach based on private international law.⁶⁷

However, if the tribunal has no legal framework to contain the parties’ choice, the result may be that arbitration lends itself to practices that violate fundamental principles in the international community. This is not desirable, at least for two reasons. First, it may lead to rendering an award that is invalid or unenforceable because it violates the court’s public policy. Second, as I argue in section III. above, it undermines the credibility of arbitration as a method for settling disputes—because it makes arbitration a means to evade applicable mandatory regulation of fundamental character.

The reasons of the award, in summary, are crucial to the court’s evaluation of the tribunal’s authority to apply a law that was not chosen by the parties. If the arbitral tribunal bluntly disregards the parties’ choice, it exceeds its power. If it exercises its power to select the applicable law, it does not exceed its authority.

It may not always be easy to determine the borderline between the review of the tribunal’s application of the law and the decision as to whether the tribunal had the authority to apply that law. The former is not within the scope of the jurisdiction of the court. The latter may be evaluated by the courts when determining whether the arbitral tribunal exceeded the power that it was granted by the parties. This judicial control has to be based on a careful analysis of the reasons for the award to verify that the proper criteria for the exercise of the defence are met.

VII. REASONING AND COURT CONTROL: RIGHT TO BE HEARD

Another ground for invalidity and unenforceability of awards is violation of a party’s right to be heard. For what interests us here, the allegation that the right to be heard was violated overlaps often with the ground of excess of power that was discussed in section VI. above. In particular, it might overlap with the allegation that one party was taken by surprise by the basis for the tribunal’s decision, and was thus deprived of the possibility to present its case from the point of view that was assumed by the arbitral tribunal.

Also the ground relating to the right to be heard can be applied to a situation in which the tribunal based its decision on a legal reasoning that was not pleaded or did not consider one or more issues pleaded by a party (an award infra petita partium), as it is generally assumed that the right to be heard implies a duty by the tribunal to listen.⁶⁸

A party may have been taken by surprise if the tribunal relied on facts or sources that were not pleaded by the parties, or if the tribunal applied the sources in a manner that did not correspond to the parties’ pleadings. Not having expected that the tribunal would base its decision on those elements or on that reasoning, the parties did not make their comments thereon and were deprived of the possibility to present new arguments or evidence that they might have presented if they had known that a certain fact would be considered or that a certain source would be applied by the tribunal.

It must be pointed out that, in some systems, arbitral tribunals have the power to develop their own legal arguments and are not expected to request the parties to comment on the tribunal’s legal reasoning.⁶⁹ However, even in those systems, the tribunal’s room for independent legal reasoning may be restricted if the parties may be deemed to have been taken by surprise by the application of a certain source.⁷⁰

The award’s reasons can contribute to verifying whether the parties were given the possibility to present their respective cases, whether they had the possibility to expect the application of the sources upon which the decision relied, and whether the parties’ pleadings were taken into due consideration.

VIII. REASONING AND COURT CONTROL: DUE PROCESS

Regarding the grounds for setting aside an award or refusing its enforcement on the basis that procedural errors were made, the award’s reasons are not as crucial as they are in respect of the grounds discussed in the previous sections. However, reasons may be very helpful, because they may constitute a record of the procedural history that can be used to evaluate whether errors were committed during the proceedings.

It is possible that a party challenges the validity of an award or resists its enforcement—for example, alleging that it was discriminated against because it was given less time than the other party to present its evidence; or alleging that an arbitrator was not independent; or objecting to the decision on costs in the award.⁷¹ These allegations may, depending on the circumstances, be groundless. For example, one party may have been given less time than the other to present its evidence because that party’s evidence was produced in the language of the proceedings, while the other party’s evidence needed to be translated. In such a situation, the different time frame granted to the parties does not reflect a discrimination but reinstates equality of treatment between the parties. The challenge to the independence of the arbitrator may have been made too late, for example because the challenging party was asked, in the course of the proceedings, to comment on a certain circumstance regarding the independence of the arbitrator, and failed to object at that time. If the party waited to challenge the arbitrator until after the award had been rendered, it may have lost the right to invoke that circumstance. The fees of the arbitrators may be very high compared to the value of the dispute, but this may be due to excessive work that the arbitral tribunal was forced to do because of numerous and groundless requests made by the party who challenged the fees.

In all these situations, and many others, it is possible to provide evidence that the invoked circumstances are not grounds for invalidity and unenforceability. However, this requires a considerable work for the other party and the court, who are expected to collect evidence that may be scattered among hundreds of e-mails exchanged during the proceedings, and to deduce from them what were the reasons for the arbitral tribunal’s conduct. If the award contains reasons, and the reasons have a section explaining the procedural history, it becomes much simpler for the controlling court to follow how the proceedings were conducted and to take a decision on whether the circumstances justified the tribunal’s conduct or whether the award shall be set aside or refused enforcement.

IX. ARE THE REASONS OF AN AWARD DISPENSABLE?

It was seen in section I. above that arbitration is under pressure to become a more efficient process. Recent developments such as the creation of the role of the emergency arbitrator testify to the desire to ensure that arbitration is capable of speedy responses to the needs of the parties. Many arbitration institutions have added rules for expedited proceedings to their rules for ordinary proceedings.⁷² Expedited proceedings are meant to be completed in a shorter time frame than ordinary proceedings. A number of measures permits a more expedited conduct of the proceedings: for example, the number of exchanges of submissions that are to take place between the parties may be reduced, the tribunal may be given the discretion to decide the case on the basis of written submissions only, the deadlines for writing the award may be reduced, etc.

The measures to enhance efficiency may also affect the reasons of the award. For example, under the Stockholm Rules for Expedited Arbitration, the award is to state reasons only if one of the parties so requests.⁷³ This means that, under these rules, the base case is that an award needs not state reasons.

Most rules for expedited arbitration have not gone so far, and awards are usually expected to state reasons even in expedited proceedings. The ICC Expedited Procedure Rules,⁷⁴ for example, do not permit to derogate from article 32 (2) of the Arbitration Rules, according to which awards must state the reasons upon which they are based.⁷⁵ As a matter of fact, the ICC Notes to the Parties and Arbitral Tribunals state that: "Any award under the Expedited Procedure Provisions shall be reasoned. Arbitral tribunals may limit the factual and/or procedural sections of the award to what they consider to be necessary to the understanding of the award, and state the reasons of the award in as concise a fashion as possible." In my experience as member of the ICC Court of Arbitration, the scrutiny of the draft awards rendered under the Expedited Procedures takes into consideration the need for an expedited processing, but is no less thorough in substance than for ordinary arbitration.

The UNCITRAL Working Group II on dispute resolution, which is currently working on matters relating to expedited arbitration, discussed at its 69th session in February 2019 whether it is desirable to eliminate the requirement that the award states the reasons for the decision.⁷⁶ The outcome of the discussion was not favourable to eliminating the requirement, but the simple fact that the matter was discussed, shows that measures are contemplated for the purpose of enhancing efficiency.

In a similar vein, for ordinary (non-expedited) arbitration proceedings the UNCITRAL Model Law on International Commercial Arbitration permits the parties to agree that the award shall not state any reasons.⁷⁷ This is a less drastic solution than the described approach in the Stockholm Rules for Expedited Arbitration. While the Stockholm rules require that one party actively requests that the award states reasons, the Model Law requires that both parties actively agree to exclude the reasons. This should be seen as a hommage to the primacy of party autonomy, rather than as a measure to enhance efficiency. In the quest for efficiency that inspires contemporary arbitration, however, the possibility for the parties to agree on discarding the reasons may become an appealing option.

As was discussed in the previous sections, however, the reasons of the award may be necessary for the purpose of rendering court control meaningful. Given that the award’s reasons are necessary or important to permit court control, how can this be reconciled with the systems that permit the parties to agree that the award shall not state any reasons, or even that assume that awards shall not have reasons unless one party requests so? As an award that does not state reasons prevents or at least hampers court control, it must be assumed that an award without reasons will not be deemed acceptable by a court. Permitting the parties to exclude reasons may create considerable problems if the applicable arbitration law does not permit to exclude court control, and may create difficulties in case of enforcement.

In some systems, such as under Swedish law, awards do not need to state reasons,⁷⁸ and the parties furthermore are allowed to waive the right to challenge the validity of the award.⁷⁹ Therefore, the Stockholm rules are not inconsistent with the applicable arbitration law, when they assume that awards shall not have reasons unless one party requires so. However, under the Model Law, the parties may not waive their right to challenge the validity of the award. How can this be reconciled with the possibility that the Model Law gives to the parties to agree on having an award without reasons? When the reasons are necessary to evaluate the validity of the award, how to interpret the parties’ agreement that the award shall not state reasons? Is it an indirect way of waiving the right to challenge which, however, may not be waived under the applicable arbitration law? Or is it a reason to consider the award invalid, because it prevents the court from exercising control? The inconsistency in the system is evident. ⁸⁰

The same inconsistency applies to enforcement of the award: the New York Convention does not permit the parties to waive their right to resist enforcement.⁸¹ The losing party may, of course, elect not to resist enforcement, thus achieving the same result as if it had waived the right to resist. However, an agreement in which a party commits itself in advance not to resist enforcement in case it loses, is not contemplated in the New York Convention. If an award is without reasons, therefore, there is a possibility that its enforceability is at risk.

The foregoing seems to suggest that it is not advisable for the parties to agree on an award without reasons. While it may be tempting to expedite the award by doing away with its reasons, the final result may be far less efficient than if the tribunal had been given some additional weeks to draft reasons—as the award may run the risk of being set aside or refused enforcement. If the award may not be carried out, it does not help much that it was rendered within a short timeframe.

X. WHAT STANDARD SHOULD BE MET WHEN STATING THE REASONS OF THE AWARD?

The foregoing shows what is necessary or advisable to address in the award, for the sake of preserving its validity and enforceability. However, there does not seem to be a general standard applicable to international arbitration,⁸² and in any case the level of detail will depend on the characteristics of the dispute. A simple dispute based mainly on the evaluation of evidence will not require as extensive legal analysis as a dispute presenting various legal issues; a dispute in which the parties have a high level of litigiousness will require a more detailed procedural history than a dispute in which both parties acted seriously and professionally, etc.

It is advisable to write reasons that not only meet the requirement of court control, but also explain to the parties, in a convincing manner, the basis for the decision. This, however, does not mean that the award shall be written with the aim of developing the law and becoming a precedent for future disputes. Considering that in commercial arbitration the priority is still efficiency, it does not seem advisable to significantly exceed the standard that is necessary to satisfy court control. The style of investment awards, therefore, is not necessarily to be followed for commercial awards. The arbitral tribunal should strive to strike a reasonable balance between the two extremes: an award that does nothing more than ordering one party to pay a certain amount, on one hand, and an award that constitutes a treatise on the matters at issue, on the other hand.

A reasonable balance suggests that an award contains:

(i) Reference to the arbitration agreement, as well as an explanation of the basis upon which the tribunal assumed jurisdiction on the dispute. If the jurisdiction was contested, the award should refer to the applicable sources of law that constitute a basis for jurisdiction. As was seen in section IV. above, the controlling courts have an independent power to determine whether the tribunal have jurisdiction. However, the reasoning of the award will be the starting point for the court’s considerations. For example, the existence of the arbitration agreement will depend on the formal requirements in the arbitration law at the place of arbitration, as well as on the evidence examined by the arbitral tribunal. If the award gives the relevant details, the annulment or the enforcement court will have a full basis to evaluate the matter. Furthermore, a party may be deemed to have lost its right to invoke that the arbitral tribunal did not have jurisdiction, if that party had the possibility to object to jurisdiction during the proceedings, but elected not to do so within the given term. A detailed reference in the award to the communication that has been made on this matter during the proceedings, constitutes an important basis for the court’s evaluation of the parties’ conduct and its decision on whether a party lost its right to challenge jurisdiction or not;

(ii) A description of all steps that led to the appointment of the arbitral tribunal. Similarly as to what was mentioned in respect of the loss of the right to invoke that the arbitral tribunal lacked jurisdiction, also in respect of the right to challenge the constitution of the arbitral tribunal, a party may lose its right if it does not present its objections before the term set for doing so during the proceedings. If the award refers to any communication that was made during the proceedings on, for example, the qualifications or the independence of the arbitrators, the court will be able to verify whether later challenges are timely or not;

(iii) A description of the procedural history. As was seen in section VIII. above, a thorough description of the proceedings may contribute to the effectiveness of the award as it provides the court with a basis to verify the regularity of the procedure. It happens, sometimes, that a party is particularly litigious, possibly giving rise to the suspicion that it is trying to delay the proceedings—for example, it presents numerous unfounded requests to extend the terms, it objects to all procedural decisions taken by the tribunal, it requests discovery of voluminous evidence that is not relevant, etc. The award should record all the parties’ requests and objections, and it should explain the reasons for its decision on each of them. The award should refrain from characterising or commenting on the parties’ conduct beyond giving objective explanations of the reasons for admitting or rejecting the requests and objections. Tempting as it may be under some circumstances, any non-objective comment may expose the award to being challenged on the basis that the arbitral tribunal did not act in an impartial way. This does not mean that the tribunal shall succumb to the so-called due process paranoia and accept, out of fear for infringing that party’s right to due process, any abusive conduct by the parties. It means that any decision taken by the arbitral tribunal to prevent dilatory tactics should be motivated with objective reasons;

(iv) A description of the facts upon which the decision is made. This description has the double purpose of making it possible to follow the reasoning upon which the decision is based, and to show that the arbitral tribunal has taken into due consideration the facts invoked by each of the parties. To be useful, therefore, the description should be functional and provide a basis for the legal issues that are decided in the award. Facts that are not relevant to the legal issues, or details that are not necessary for illuminating the legal issues, should not be referred to. If the parties have made reference to irrelevant facts or to an excessively high degree of details, the award may make reference to that when it comments on the parties’ respective positions (item (v) below), explaining why those facts are deemed irrelevant or the level of detail excessive;

(v) A description of the parties’ position and the remedies requested by each of the parties. Also this description has the double purpose of making it possible to follow the reasoning upon which the decision is based, and to show that the arbitral tribunal has taken into due consideration the arguments presented by each of the parties. Often, this part of the award incorporates verbatim the parties’ written statements. This is made in part to save time, because copying and pasting the parties’ pleadings is quicker than reformulating the parties’ arguments. In part, however, this is made also to avoid being criticised for having overlooked some of the parties’ arguments. In my opinion, copying and pasting the parties’ arguments is not an effective measure against the possible criticism that certain arguments were not considered. This only achieves to push the criticism forward: while the part of the award that incorporates the parties’ pleadings is exhaustive, the part of the award containing the reasoning will reveal if one argument has not been considered. In my opinion, it is better to summarise the parties’ arguments in the award, rather than copying them. Certainly, the arbitral tribunal will thoroughly study the parties’ arguments, possibly taking notes for its own understanding and to structure the case. In my experience, these notes are quite easy to be transformed into a summary of the parties’ main pleadings. A summary like this would render the award more readable than the copy-paste exercise, and does not require much more work than what the arbitrator anyway has to do for the purpose of understanding the case. If a tribunal is concerned with ensuring that the description captures the main arguments presented by each of the parties, it may circulate this description and ask the parties to comment thereon;

(vi) A description of the tribunal’s reasoning. This part should address the parties’ main factual and legal arguments, making specific reference to the invoked evidence and the arguments of each of the parties, and explaining the tribunal’s opinion on each of these arguments. The award should refrain from making non-objective or even sarcastic comments on the arguments of the parties, and even arguments that are not convincing at all should be given due consideration. If an argument is manifestly irrelevant, illogical or unfounded, it will not be necessary to spend many sentences to dismiss it; but the dismissal should be objective and respectful. This part of the award should also explain the tribunal’s legal reasoning, identifying the legal sources that it applies, explaining how it selected them and how they are applied to the facts. The award should be precise in making reference to the applicable law. Arbitral tribunals are often under the impression of being empowered to apply the law more flexibly than national courts of law. It is correct that an award remains valid and enforceable even if it has applied the law wrongly. However, as seen in sections VI. and VII. above, an award runs the risk of being in excess of power or infringing a party’s right to be heard if it applied a law different from the law chosen by the parties without explaining the legal basis for this application;

(vii) The determination of the arbitration costs, containing also an evaluation of the reasonableness of the parties’ costs if they are to be allocated between them, the indication of the criteria for determining the fees of the arbitral tribunal and of the institution, if any, as well as an explanation of how the costs have been allocated between the parties;

(viii) The dispositive. The dispositive should be clear and self-sufficient: it should be possible to enforce the award on the simple basis of the dispositive. The only purpose of the dispositive should be to constitute the basis for enforcement. Therefore, the dispositive should have no evaluations, arguments or explanations, but only orders or determinations (including also the determination that the tribunal has or does not have jurisdiction). Furthermore, all parameters for determining the amount of payments or for enforcing other ordered remedies should be identified or identifiable in the dispositive. The enforcing court should not be expected to interpret the contract or to evaluate the quality of a delivery or other performance. All evaluations should be made in the body of the award, and the dispositive should make the specific orders and determinations that follow from the considerations made in the body of the award. The only open issues in the dispositive should be capable of being determined in a mechanical way—for example, the amount due as interest should be determined or determinable on the basis of the criteria contained in the dispositive: a specific date for when the calculation starts, the final date for the calculation (usually, the date of actual payment), and the rate quoted in a specific source on that date. It is also useful to specify, in the dispositive, that all claims that are not reflected in the dispositive are to be deemed rejected. This prevents that new arbitral proceedings be initiated on the basis of the same claims that were the basis for the dispute decided in the award.

XI. CONCLUSION

The foregoing shows that the award’s reasons are necessary to permit the court to exercise control over the compatibility of the award with public policy (if one adheres to the minimalist doctrine), as well as over the arbitral tribunal’s exercise of its power (particularly in case the tribunal considered sources different from the law chosen by the parties or the sources pleaded by the parties).

If the court adheres to the maximalist theory, the award’s reasons permit the court to develop its own evaluation of the public policy issue considering all elements that were considered by the arbitral tribunal. This applies also to the court’s determination of the arbitral tribunal’s jurisdiction.

Furthermore, the award’s reasons are important to permit the court to exercise control over the proceedings’ compatibility with the right to be heard and other principles of due process. The reasons of the award, therefore, should not be sacrificed in the name of more efficient arbitration proceedings.

Efficiency of the proceedings, however, is an important concern in commercial arbitration. Therefore, it does not seem appropriate to adopt the same style of drafting reasons that is customary for investment awards. Having to strike the balance between the desire to write extensive reasons capable of contributing to the development of the law and the need to ensure efficient proceedings, the reasons for commercial awards should be sufficient to permit court control, but need not exceed the requirements set by the necessity to permit court control.

---

---