John Beechey's presidency of the ICC International Court of Arbitration will be remembered as a period of revision and expansion of the ICC arbitration system, which he led with intelligence, open-mindedness and a truly international vision. Happily, he will continue to be active in the field, returning to more direct involvement in arbitration practice and the resolution of salient issues in international transactions. It is a pleasure to devote these lines to him, although I may be treading on dangerous ground as some of the arbitrators who have issued dissenting opinions enjoy a distinguished status in the arbitration world, and he may be one of them.

1. Introduction

The silence of the leading arbitration institutions (with the notable exception of ICSID) and most national arbitration laws on the issuing of dissenting opinions has left the field open to a legal debate, characterized for many years by its liveliness. The admissibility of dissenting opinions, however, is uncontested worldwide. They are normally communicated to the parties along with the award when they take the form of a separate document and, to the extent they criticize 'ordinary issues of fact and law',2 they do not imperil the authority of the award or its enforceability. The debate becomes of interest when it pertains to the underlying reasons of dissent in international arbitration proceedings or to the actual use of the dissenting opinions.

We are concerned here only with dissenting opinions stricto sensu. By this we mean either a reasoned opinion incorporated in an award expressing a position different from that of the majority on one or more issues, or a fully reasoned document issued after the award has been finalized, disagreeing with and criticizing certain decisions reached by the majority in the award. Let it be clear: a document of dissent that is [Page138:] issued by an arbitrator at the end of the deliberations and submitted to the other members of the tribunal before the award is finalized is not a dissenting opinion stricto sensu; it may ultimately become one, but equally, if clear, well-founded and impartial, it can lead to new deliberations and possibly even to a unanimous decision. What we are interested in here are genuine dissenting opinions, i.e. the sincere and serious expression of a view that differs from that of the majority notwithstanding the fact that, as is almost always the case, it is issued by the arbitrator appointed by the losing party. We exclude as being devoid of interest for the present reflection dissenting opinions that are impolite, openly biased, contrary to proven facts or in defiance of clear rules.

2. A sincerity balancing exercise

The sincerity and seriousness of a dissenting opinion should be tested through an inner balancing exercise that takes account of the particularities of international arbitration. Such an exercise should follow several steps:

First, realize that parties need a just and enforceable award, not an opinion. If the award includes and discusses the arguments of the losing party indicating to it why they have been rejected, there is no room for a dissenting opinion in the name of proper dispensation of justice. The legitimacy of the arbitral proceedings is apparent in such a case and the losing party's confidence in them normally acquired.

Second, ponder the seriousness and accuracy of the dissent and exclude the desire - be it conscious (for whatever reason) or not thought-through (due to lack of professionalism) - to show the appointing party zealous support by analyzing factual and legal issues in a different manner, in order to demonstrate how the dissenting arbitrator would decide the dispute.

Third, assess whether or not all efforts have indeed been made, during the deliberations, to convince the other arbitrators of the accuracy of the opinion. This presupposes, of course, not only that the deliberations strictly followed the collegiality principle but also that a thorough exchange of views among the arbitrators actually took place on all issues. In other words, the minority arbitrator should not see the issuing of a dissenting opinion as an easy way out of the deliberations, and the majority arbitrators should have thoroughly considered the minority arguments. Further, when, despite such meaningful dialogue, the other arbitrators remain unconvinced, attempt to re-examine in good faith the majority's reasoning and weigh the importance of the controversial issue and its impact on the dispositive section of the award.

Fourth, consider whether it is better to issue a dissent, at the risk of appearing biased, or to accept the majority decision as a matter of common sense and fair play. Indeed, the minority arbitrator must not forget (i) that today the generally accepted rule in international arbitration is that all arbitrators - including those appointed by parties - [Page139:] must be and remain independent and impartial and (ii) that the picture that consistently emerges from statistics is that dissenting opinions almost always support the position of the party that appointed the dissenting arbitrator. It is thus advisable to consider seriously whether the issuing of a dissenting opinion could give the impression of partiality - even though the dissenting arbitrator sincerely believes in the dissent - and thus bring its author into disrepute given the rightly austere postures on this question in the arbitration community today, but also the antagonism in it. And, in the words of Pierre Mayer,3 consider further that 'there should be a mixture of modesty and fair play. There was a discussion, a debate, and the arbitrator was in a minority. There is an award, a majority award, and the minority should admit defeat.'

Fifth, comprehend that there is no duty to issue a dissent: no duty is owed to anyone and nobody is entitled to criticize the fact that a dissenting opinion has not been issued. The only possible exception where issuing a dissenting opinion may be regarded as a duty is when it is motivated by a violation of due process, i.e. the right of the parties to be heard or the collegial principle applicable to deliberations. It is however the reaction to such extreme situations and their disclosure that constitute the 'duty' and not the dissenting opinion on substance resulting therefrom. If justified, such a disclosure should in principle take place before the finalization of the award and will thus very probably lead to the replacement of the presiding arbitrator; in any event, an ex post facto disclosure is a ground for setting aside the award and does not necessitate the drafting of a dissenting opinion on substance. Finally, remember that, as has been repeatedly pointed out,4 the fact of not issuing a dissenting opinion and of signing an award does not mean that all arbitrators consent to it. If that were the case, dissenting opinions would be much more frequent in practice.

It is arguable that, after such a complete inner balancing exercise, would an arbitrator decide to issue a dissenting opinion in very few cases. This may nonetheless occur when the dissenting arbitrator remains convinced of the seriousness of his/her intellectual disagreement, and the dissent responds to a real need where cultural idiosyncrasies and egos have no part to play. Such a need - in essence an expression of individual integrity - may in that case be objectively justified if the dissenting opinion is proved to be of actual or potential use.


3. No real use in commercial arbitration

As far as international commercial arbitration is concerned, we do not believe that a dissenting opinion may serve any useful purpose, given that awards are not generally published; on this there is widespread agreement. The only situation where it may serve a purpose - and which immediately comes to mind in the context of this tribute - is the scrutiny of awards by the ICC Court under Article 33 of the ICC Rules of Arbitration. It is well known that the Court may refer to the dissent in its comments to the tribunal and invite the latter to strengthen its reasoning or reconsider points of substance in the draft award in light of the dissent, if the Court has found the dissenting opinion particularly persuasive. We are not aware of any recent examples; we only know that the percentage of dissenting opinions in ICC arbitrations is low, all the more so when they are expressed in a separate document. From past experience of Court practice, we believe, however, that the flaws in a draft award, to the extent they may lead to comments to the tribunal on points of substance, would normally be apprehended by the Secretariat, the rapporteur assigned to the case or the Court members, regardless of the dissenting opinion.

4. Focusing on investment arbitration

Only in investment arbitration would dissenting opinions appear to have some real use. It is argued that they contribute to the development of investment law, as awards and dissenting opinions are for the most part published.

In recent years, the contribution of dissenting opinions to the development of investment law has been the subject of specific - and maybe unfinished - debate between most distinguished members of the arbitration community, generated by Jan Paulsson's and Albert Jan van den Berg's mistrust and criticism of the system of party-appointed arbitrators.5 Charles N. Brower and Charles B. Rosenberg, on the other side, have commented6 on van den Berg's disapproval of dissenting opinions issued by party-appointed arbitrators, and striven to explain that dissenting opinions play an important role in fostering the legitimacy of investment arbitration; one of the authors' arguments was precisely the contribution they make to the development of investment law.

Tangible proof of such contribution would undoubtedly be awards that clearly follow dissenting opinions in earlier cases on certain questions. After meticulous research, Brower and Rosenberg relied on four ICSID [Page141:] awards that referred to earlier dissenting opinions, two of which to the same opinion. Leaving aside the particularly small number of awards, the references cannot be considered determinative: they are just occasional, limited remarks made in passing on a previously adopted position, akin to references in scholarly literature.

It is difficult to form with certainty an opinion on the existence of examples of awards that have actually followed dissenting opinions in earlier cases because, as yet, not all awards in investment arbitration are published. For the moment, research is necessarily limited to ICSID and NAFTA awards, so whatever conclusions are drawn remain relative. It seems improbable, however, that such examples actually exist, at least in significant numbers, given the systematic reliance by arbitrators on previous decisions by arbitral tribunals, notwithstanding the absence of a formal doctrine of precedent in investment arbitration. Such reliance is strongly supported by scholars - it is even characterized as a 'moral obligation' incumbent upon arbitrators - for the sake of consistency and predictability in this area, where the law is not yet well developed.7 In any event, irrespective of the arbitrators' real motives for referring to earlier decisions, this practice is understandable, given that similar or identical issues of investment law recur in different cases, and it has the effect of marginalizing the 'use' of dissenting opinions in this context. Conversely, the dissenting opinions are in practice sometimes used by the parties in their pleadings, with or without specific reference to them as such.

We have not carried out exhaustive research into published investment awards, but just looked at the ICSID awards issued from 2013 onwards, i.e. after the article by Brower and Rosenberg. Our impression is of a significant number of ICSID awards accompanied by a dissenting opinion.8 One may seriously wonder whether publicity induces arbitrators to dissent, be it under the effect of 'arbitrator shopping' or a quasi-obligation to render public the full debate on a novel or [Page142:] particular issue. Dissenting opinions are thus almost always separate documents, sometimes extending to hundreds of pages, and resemble pieces of academic literature.

Below we refer to three ICSID cases, which are pertinent to the present reflection on the 'use' of dissenting opinions in investment arbitration.

The intense discussions generated by the decision on jurisdiction in Abaclat and Others v. Argentine Republic9 and by the long and thorough dissenting opinion of Prof. Georges Abi-Saab, notably criticizing the majority's acceptance of jurisdiction over sovereign bonds and admission of mass proceedings in a treaty-based arbitration, are well known. We focus here on the first issue, which arose later in two other cases giving rise to differing decisions: Ambiente Ufficio SpA and others v. Argentine Republic,10 also accompanied by a detailed dissenting opinion, and, more recently, Poštová banka, a.s. and Istrokapital SE v. Hellenic Republic.11Abaclat was the first ICSID case involving sovereign debt, so it was only to be expected that the question of jurisdiction would be at issue and would give rise to the expression of strongly held views. It is not our intention to take a position on those views, but merely to see how subsequent tribunals dealt with the same issue and how they reacted to the much discussed dissenting opinion expressed in that case.

Interestingly, the Ambiente decision, which retained ratione materiae jurisdiction on the basis of a 'double-barreled' test (i.e. after examining its competence with respect to the security entitlements held by the claimants under Art. 25(1) of the ICSID Convention and Art. 1 of the Argentina-Italy BIT), contains in both the body of the decision and the footnotes repeated references to the Abaclat dissenting opinion on this and other issues, although it followed the reasoning of the majority in Abaclat. In particular, the Ambiente tribunal followed Abaclat on the following points (i) the non-significance of a distinction between bonds and security entitlements since they are part of one and the same economic operation ('general unity of an investment operation'); (ii) non-endorsement of a narrow reading of the term 'investment' in Art. 25(1) of the Convention, which thus covers bonds and security entitlements; (iii) fulfilment of criteria generally ascribed to the Salini test, although in Ambiente this test was considered to be of a non-jurisdictional nature; (iv) the understanding that Art. 1 of the Argentina-Italy BIT specifically addresses financial instruments and that the term 'obligaciones' undoubtedly includes bonds and security entitlements; and (v) the territoriality requirement of the BIT regarding investments of a purely financial nature, which should be considered fulfilled if the [Page143:] state benefits from the investment. Prof. Abi-Saab had vigorously criticized the Abaclat majority on all of the above, except for the construction of Art. 1 of the Argentina-Italy BIT on the definition of 'investment', regarding which he agreed that 'its language is large enough to encompass the security entitlements in the Argentinean bonds' (para 68).

Given the heated scholarly debate to which Abaclat had given rise, the fact that both parties in Ambiente had apparently discussed the dissenting opinion of Prof. Abi-Saab in their pleadings and that one of the arbitrators was endorsing that opinion - and finally issued his own dissenting opinion - it was natural and intellectually fair and appropriate that the Ambiente tribunal refer to it, albeit disapprovingly. As far as we are aware, it is the first time that a dissenting opinion displayed such tangible endurance in investment arbitration. It is however doubtful whether that endurance may actually be perceived as a genuine contribution to the development of investment law.

The Poštová case did not have the same close relationship with Abaclat, since it was based on a different BIT (the Slovakia-Greece BIT),12 but it too concerned sovereign debt instruments (the Greek Government Bonds (GGB)) held by Poštová banka, and again raised the question of whether the holding of such bonds constituted a 'protected investment' allowing the tribunal to accept jurisdiction ratione materiae. The tribunal's analysis seemed to support a 'subjective' definition of 'investment' based on the will of the state parties as expressed in the BIT, although it finally also examined GGB interests in light of objective criteria.

On the basis of a detailed comparison and discussion of the Argentina-Italy and the Slovakia-Greece BITs, the tribunal in Poštová found that there was clear difference of wording in the definition of investment, namely the Slovakia-Greece BIT does not contain any reference to 'obligations' or to 'securities', much less to public titles or public obligations and sovereign debt. The numerous references by the parties in their pleadings to the decisions in Abaclat, in particular, and Ambiente were therefore to be expected and necessary. The tribunal unanimously decided that it had no jurisdiction over the dispute. Although this outcome coincided with the positions taken by the dissenting opinions in Abaclat and Ambiente, the Poštová award contained no reference at all to those dissenting opinions or to any part of their reasoning.


5. Concluding remarks

With respect to investment arbitration, it is difficult to conclude with a clear-cut response to the question of whether genuine dissenting opinions are of real use. We strongly doubt, however, that a general proposition that dissenting opinions contribute to the development of investment law can be substantiated. Our impression is rather that this is an excuse for what an arbitrator may perceive as an intellectual need to express a different opinion. So, the same inner balancing exercise as discussed above should take place before issuing a dissenting opinion in investment arbitration, and to the five steps listed above we would add a sixth: seriously consider whether the dissent concerns an issue that arises for the first time, is quite important and decisive per se for the resolution of the dispute and is likely to reappear in future cases. If this balancing exercise comes out in favour of issuing a dissenting opinion, it will very probably give rise to a legal debate within the arbitration community, which is always constructive. Also, it cannot logically be ruled out that, departing from the general practice of following previous awards, future tribunals might exceptionally follow the dissenting opinion due to its persuasiveness on the novel issue. Nor can it be ruled out, however, that the flawed or questionable reasoning of the award on the issue in question would alone be sufficient to give rise not only to scholarly debate but also to the same ultimate result (the adoption of the dissent's reasoning by future tribunals), without the guidance of the dissenting opinion. Lastly, the drafting of an article by the dissenting arbitrator on the issue in question, rather than a dissenting opinion, might sometimes be a more appropriate channel to satisfy the imperative and genuine need of expression and, precisely, it could be of real and easier use.

Founding Partner, A. Dimolitsa & Associates, Athens, Greece; Vice-Chair, ICC Institute of World Business Law.

P.J. Rees QC & P. Rohn, 'Dissenting Opinions: Can they Fulfil a Beneficial Role?' (2009) 25 Arbitration International 329.

P. Mayer, 'Dealing with Dissenting Opinions' in B. Berger & M.E. Schneider, eds., Inside the Black Box: How Arbitral Tribunals Operate and Reach Their Decisions, ASA Special Series No. 42 (Juris, 2014) 67 at 69.

See e.g. P. Lalive, J.-F. Poudret, C. Reymond, Le droit de l'arbitrage interne et international en Suisse (Lausanne: Payot, 1989) at 415 (Art. 189, § 16); P. Mayer, supra note 3 at 72; M. Arroyo, 'Dealing with Dissenting Opinions in the Award: Some Options for the Tribunal'(2008) 26 ASA Bulletin 437 at 463.

J. Paulsson, 'Moral Hazard in International Dispute Resolution' (2010) 25 ICSID Review 339; A. J. van den Berg, 'Dissenting Opinions by Party-Appointed Arbitrators in Investment Arbitration' in M. Arsanjani et al., eds., Looking to the Future: Essays on International Law in Honor of W. Michael Reisman (2011) 821.

C.N. Brower & C.B. Rosenberg, 'The Death of the Two-Headed Nightingale: Why the Paulsson-van den Berg Presumption that Party-Appointed Arbitrators are Untrustworthy is Wrongheaded' (2013) 29:1 Arbitration International 7.

G. Kaufmann-Kohler, 'Arbitral Precedent: Dream, Necessity or Excuse?' (2007) 23 Arbitration International 357 and 'Is Consistency a Myth?' in E. Gaillard & Y. Banifatemi, eds., Precedent in International Arbitration (2008) 137; C. Schreuer & M. Weiniger, 'A Doctrine of Precedent?' in P. Muchlinski, F. Ortino, Ch. Schreuer, eds., The Oxford Handbook of International Investment Law (2008) 1188.

See Ambiente Ufficio SpA and others v. Argentine Republic, ICSID Case No. ARB/08/9; Garanti Koza LLP v. Turkmenistan, ICSID Case No. ARB/11/20; Opic Karimum Corporation v. Bolivarian Republic of Venezuela, ICSID Case No. ARB/10/14; Kiliç Insaat Ithalat Ihracat Sanayi ve Ticaret Anonim Sirketi v. Turkmenistan, ICSID Case No. ARB/10/1; RSM Production Corporation v. Saint Lucia, ICSID Case No. ARB/12/10; Cervin Investissements SA and Rhone Investissements SA v. Republic of Costa Rica, ICSID Case No. ARB/13/2; Renée Rose Levy de Levi v. Republic of Peru, ICSID Case No. ARB/10/17; ConocoPhillips Petrozuata BV, ConocoPhillips Hamaca BV and ConocoPhillips Gulf of Paria BV v. Bolivarian Republic of Venezuela, ICSID Case No. ARB/07/30; Mamidoil Jetoil Greek Petroleum Products Societe Anonyme SA v. Republic of Albania, ICSID Case No. ARB/11/24; Venoklim Holding BV v. Bolivarian Republic of Venezuela, ICSID Case No. ARB/12/22; Iberdrola Energía SA v. Republic of Guatemala, ICSID Case No. ARB/09/5.

Abaclat and Others v. Argentine Republic, ICSID Case No. ARB/07/5, award of 4 Aug. 2011

Ambiente Ufficio SpA and others v. Argentine Republic, ICSID Case No. ARB/08/9, decision of 8 Feb. 2013

Poštová banka, a.s. and Istrokapital SE v. Hellenic Republic, ICSID Case No. ARB/13/8, award of 9 Apr. 2015

It was also based on the Cyprus-Greece BIT with regard to Istrokapital, but this is irrelevant in the present context.