icsid Jurisdiction with Regard to State-Owned Enterprises – Moving Toward an Approach Based on General International Law

In: The Law & Practice of International Courts and Tribunals
Giulio Alvaro Cortesi University of Padua and Paris I Panthéon – Sorbonne

Search for other papers by Giulio Alvaro Cortesi in
Current site
Google Scholar
Full Access

The decision of the ad hoc Committee in the case Tulip Real Estate v. Turkey of 2015 represents a new development in the methods applied by icsid tribunals to ascertain their own jurisdiction when the acts giving rise to the investor’s claim have been undertaken by a State-owned entity. Indeed, prior to this case, different views and trends had arisen over the years: some tribunals have addressed this matter by narrowing their field of analysis to the formal requirements of the claim, while others have focused on the issue of implicit designation. Lastly, there has been a growing trend, starting in 2009 with the case Toto Costruzioni v. Lebanon, where arbitrators have applied the rules on attribution at the jurisdictional stage. This last group of cases suggests that those arbitrators were developing an approach closer to that of general international law. The Tulip Real Estate v. Turkey Committee was the first annulment panel to comment on this latest development, although they commented without expressing much enthusiasm for it. The present article will focus on the issue of establishing an icsid tribunal’s jurisdiction when a State entity is involved in the proceedings, in order to cast some light on the rules that a tribunal should apply when deciding whether a dispute between the foreign investor and a State entity falls within its remit.


The decision of the ad hoc Committee in the case Tulip Real Estate v. Turkey of 2015 represents a new development in the methods applied by icsid tribunals to ascertain their own jurisdiction when the acts giving rise to the investor’s claim have been undertaken by a State-owned entity. Indeed, prior to this case, different views and trends had arisen over the years: some tribunals have addressed this matter by narrowing their field of analysis to the formal requirements of the claim, while others have focused on the issue of implicit designation. Lastly, there has been a growing trend, starting in 2009 with the case Toto Costruzioni v. Lebanon, where arbitrators have applied the rules on attribution at the jurisdictional stage. This last group of cases suggests that those arbitrators were developing an approach closer to that of general international law. The Tulip Real Estate v. Turkey Committee was the first annulment panel to comment on this latest development, although they commented without expressing much enthusiasm for it. The present article will focus on the issue of establishing an icsid tribunal’s jurisdiction when a State entity is involved in the proceedings, in order to cast some light on the rules that a tribunal should apply when deciding whether a dispute between the foreign investor and a State entity falls within its remit.

* The author would like to thank Professor Andrea Gattini and Professor Hervé Ascensio for providing guidance, inspiration and support.


The decision of the ad hoc Committee in the case Tulip Real Estate v. Turkey1 of 2015 represents a new development in the methods applied by icsid tribunals to ascertain their own jurisdiction when the acts giving rise to the investor’s claim have been undertaken by a State-owned entity. Prior to this case, indeed, different views and trends had arisen in accordance with different perceptions of the nature of investment arbitration, which, as it is often considered, lies at a juncture between public international law and commercial law. In this context, the treatment of State-owned enterprises, which possess a unique hybrid nature, is paradigmatic of the evolution of international investment law. As we will see, some tribunals have addressed this matter by narrowing their field of inquiry to the formal requirements of the claim, while others have focused on the issue of implicit designation. Finally, there has been a growing trend, which started with the case Toto Costruzioni v. Lebanon2 of 2009, whereby arbitrators apply the rules on attribution at the jurisdictional stage. This last group of cases is extremely interesting, because the approach applied therein implies that the arbitrators were paying more attention to their duty to ascertain their own jurisdiction, thereby shifting their decisions toward an approach closer to one of general international law.3 The Tulip Real Estate v. Turkey Committee was the first annulment panel to comment on this development, but the Committee did not demonstrate any enthusiasm for this new approach, thereby rendering the global picture quite blurred. Hence, the aim of the present article is to cast some light on the rules that a tribunal should apply when deciding whether a dispute between a foreign investor and a State entity might fall within its remit.

In order to offer a complete and systematic analysis of the matter at stake, the article has been divided into three main sections: (1) the first section, which is devoted to a general analysis of the wording of Article 25(1) of the icsid Convention with reference to the topic of State-owned enterprises; (2) the second section, which focuses on the different jurisprudential trends that have emerged in this field; and (3) the third section where attention will shift to the different, but related, matter of designation, to verify whether the fact that the respondent State has designated a State-owned enterprise to be a party in icsid proceedings might create a separate independent jurisdictional ground. In the conclusion, the advantages of an approach based on general international law in the field of international investment arbitration will be illustrated.

1 Requirement of Jurisdiction related to State Entities under the icsid Convention

1.1 Attribution as the Appropriate Test to Verify the Fulfilment of the Requirements of Article 25(1) of the icsid Convention

Article 25(1) of the icsid Convention, which regulates the jurisdiction of the Centre,4 is worded, apparently, in a clear manner, but many of its terms are not defined. One of them, the term: “Contracting State”, will be at the heart of the present inquiry. The approach maintained by the fathers of the Washington Convention, aimed at limiting the defining norms,5 had some merit, because it avoided involving the drafters in an endless discussion to try to identify once and for all the perimeter of such a complex entity as a State. Nevertheless, it left many potential questions unanswered, especially with respect to the treatment of State-owned enterprises, which, therefore, shall be answered in having regard to the principles arising from general international law and scholars’ writings. On this subject, after many years of debate, it has been quite broadly accepted that State-owned enterprises enjoy a separate legal personality from the State.6 As it is well known, however, there is a sort of grey area where the State and the State entity are so intertwined7 that their links may be relevant, among other hypotheses, when arbitrators are called to rule on the scope of their jurisdiction or, rather, on the issue of State responsibility with regard to the acts of these entities. These issues have attracted the attention of scholars in the fields of both international commercial arbitration and investment arbitration, creating a privileged point of view for a comparison between those two systems.

In the realm of investment arbitration, the relevant provisions are included in Article 25(1) of the icsid Convention. For the benefit of our analysis, it is first of all necessary to clarify whether this norm, by referring to a Contracting State, is establishing an objective requirement and, if it is, it is necessary to examine what is the right method to ascertain its existence.

Reputed scholars8 have observed that the jurisdictional provision in the icsid Convention addresses specifically disputes opposing a private investor to a State, thereby excluding those between private parties as well as State to State disputes.9 From this perspective, it seems that Article 25(1) of the icsid Convention carries out the same function as Article 34 of the Statute of the International Court of Justice (icj),10 thereby narrowing the categories of subjects that can be parties to an icsid proceeding, as the latter does for those before the icj, restricting this possibility (on the respondent side) to States.11 The importance of this requisite was recognized already in the jurisprudence of the Permanent Court of International Justice (pcij), where judges identified it as one of the jurisdictional requirements that they are compelled to examine ex officio.12 The same view has been upheld by scholars with reference to the icsid Convention.13

If we accept that when the acts giving rise to the investor’s claim are undertaken by a State-owned enterprise a tribunal is compelled to ascertain the fulfilment of the requirements of Article 25(1) of the icsid Convention, then a second question arises, which is related to the method that the tribunal should employ. We suggest that the most adequate tool is constituted by the rules on attribution.

In this respect, inspiration may be drawn from the different, but related, issue of the “jus standi” of State-owned companies. In this field, in fact, scholars and arbitrators have felt the need to develop a test to distinguish between government-owned companies, which may legitimately qualify as nationals of another contracting State, from those whose links with the State are so tight that they cannot be considered as distinct entities. This issue was first addressed by the father of the icsid Convention, Aron Broches, who elaborated what today is called the “Broches test”, according to which “[a] government-owned corporation should not be disqualified as a ‘National of another Contracting State’ unless it is acting as an agent for the government or is discharging an essentially governmental function”.14 Reputed scholars have observed that the Broches test has significant analogies with the rules on attribution codified in the International Law Commission’s Articles of the Responsibility of States for Internationally Wrongful Acts (arsiwa)15 and some of them have even gone further, upholding the applicability of the principles codified in those rules in order to ascertain the jus standi in judicio of State-owned companies.16 It is worth noting that this reasoning on jus standi can be easily transposed, as it was by the Maffezini Tribunal,17 to the spectacular situation existing when an arbitral tribunal is called to decide on the standing to be sued of the controlling State. Even in this situation, in fact, the rules on attribution and the related jurisprudence developed in this field can be regarded as a valuable tool for the tribunal to fulfil its task.

This idea, however, has not received unconditional approval. The Paris Court of Appeal, for example, in its ruling of 29 November 2016 on the subject of the jus standi of government-owned corporations observed18 that there is no evidence that the arsiwa, a document related to the field of State responsibility, is applicable to an inquiry as to whether a State entity is to be identified with the controlling State “afin de la priver d’un droit propre à l’arbitrage en application d’un tbi”.19

Such an observation is not fully convincing, however. For example, when the entity bringing a case is a mere shell, it is hard to accept that it possesses the full procedural capacity to institute proceedings. It is difficult to qualify it as an independent legal body capable of pursuing an autonomous interest other than the one of its “effective” controller and, consequently, as an entity endowed with a distinct power to sue.20 Hence, a test to verify whether the State company can be regarded as an effective independent entity is needed and the idea to employ the rules on attribution for such purposes seems convincing.

Attribution, although comprising a concept developed in the field of State responsibility, does not involve ex se any analysis of the legality of the act undertaken by the government-owned enterprise. Rather, it focuses only on the links existing between the enterprise and the State, being nothing else than the process “by which international law establishes whether the conduct of a […] person or other such intermediary can be considered an act of State”.21

This idea has been advanced by certain scholars who have generally referred to the “twofold relevance of attribution”,22 implicitly suggesting that attribution is relevant for both jurisdictional purposes and for the merits of a dispute. As a consequence, if the respondent State objects that it should not be a party to the proceeding since the acts of the State enterprise are not attributable to it, such objection may be characterised either as an objection ratione personae or, rather, as a defence related to the merits.

If there is a consensus, however, that attribution is the tool for ascertaining the fulfilment of the requirement set forth by Article 25(1) of the icsid Convention, then the appropriate qualification of such an objection is a challenge to jurisdiction ratione personae, which has to be settled before deciding the merits of the case. The importance of this qualification turns out to be clear at the moment in which the tribunal should deal with the objection ratione personae, which is particularly significant in cases of bifurcation.

1.2 Article 41 and the Appropriate Timing to Address the Issue of Attribution

Arbitral tribunals enjoy wide discretion under Article 41(2) of the icsid Convention to bifurcate the proceedings: to deal with preliminary objections in the first phase of the proceedings and, then, at a later stage of the proceedings, to deal with other issues related to the merits.

The prerogative to bifurcate the proceedings pertains to the class of powers encompassed in the inherent jurisdiction of the arbitral tribunal. These powers are different from the more general power to decide the case, serving only the exigency of the tribunal to be “able to function”.23 Those two categories of powers are respectively applicable to two different scopes and they should be identified not through the formal label used by the arbitrators when settling a particular matter (for example a decision to bifurcate the proceedings might take the form of either a decision24 or of a procedural order25), but in the light of the function inherent in the power exercised. Therefore, when the tribunal exercises its inherent jurisdiction to decide when it will address the issue related to its competence, it is not exercising its primary jurisdiction26 and, hence, the decision to bifurcate the proceedings or to join preliminary objections to the merits does not serve to qualify a defence as a jurisdictional one. On the contrary, based on the nature of the objection, attribution shall be regarded first of all as a jurisdictional question, and thereafter the tribunal should deal with it before proceeding to the merits. This conclusion is in line with the teachings of certain eminent jurists. For example, Judge Huber recommended that, in general terms, the joining of a preliminary objection to the merits should be “exceptional”, because in general international law, no matter how many difficulties may arise, the tribunal shall not proceed with the question of the merits without having first examined its own jurisdiction27 and, according to Anzilotti, the joining of a preliminary objection to the merits forces the State to remain before the tribunal for a longer time and thus “bears a different weight in the international field where jurisdiction is voluntary”.28

The subsequent practice of the pcij and icj also confirmed this approach and the criterion endorsed by the judges for the joinder, in a limited number of cases,29 of preliminary objections to the merits was that of “absolute necessity”.30 This theoretical approach is further justified by the consideration that the party who raises a preliminary objection may have a vested interest not only in impeding the adjudicatory body to reach the merits of the case, but also in ensuring “that the case is not discussed at all before the Court”.31 This is a view that was confirmed in the Barcelona Traction case of 1964, where the icj stated that it would not have joined the preliminary objection to the merits “except for good cause”.32

As we will see, icsid tribunals, in general, have not paid too much attention to these principles, asserting that the decision is left to the tribunal’s discretion.33 That discretion, however, should be used in compliance with the general principles of international law, set out by the above-quoted jurists and by the icj. Furthermore, joining the preliminary objections to the merits becomes a necessity only when such objections do not possess an exclusively preliminary character. Thus, generally, since attribution only relates to the subjective element of the alleged breach of international investment law, it does not entail an inquiry on the legality of the act and, therefore, on the merits. If the objection to jurisdiction is dealt with at the preliminary stage, then it will obviously prejudge the merits in this respect, but this should not be regarded as an impediment to its analysis at the jurisdictional stage.34

It is submitted that the broad reading of the discretionary powers of icsid tribunals might have been influenced by an analogy to international commercial arbitration, where the only aspects the adjudicators have to bear in mind are “cost-effective” considerations.35 In some instances, the bifurcation has been refused, for example, because partial awards are deemed potentially dangerous for the winning party as “they create a further avenue for judicial review”.36 In the icsid system, where it is not possible to request the annulment of an award before domestic jurisdictions, the reasons to deny bifurcation are not the same. Further to that, with reference to economic considerations, it is worth noticing that many cases, where the tribunal postponed the issue of attribution to the merits stage, were decided in favour of the respondent State on the ground that the alleged acts of the State-owned entity were not attributable to it.37 This represents a disservice to the respondent State which is forced to develop, in addition, a full argument on the legality of the acts and, as a rule, to bear all the consequent expenses. Lastly, following the above mentioned teaching of the icj in the Barcelona Traction case, the arbitral tribunal should give due consideration to the interest of the respondent whereby the case is not discussed beyond what is necessary to reach a definitive judgment.

2 The Inquiry on Jurisdiction when a State-owned Enterprise is Involved: Different Approaches to the Same Issue

The issue of jurisdiction has been addressed several times when a State-owned enterprise has been involved in a dispute, however often by applying different approaches, which demonstrates the deep-seated dissimilarities between tribunals.

The first approach to be taken into consideration includes those cases where tribunals have maintained that the inquiry on jurisdiction cannot encompass the analysis of any issue of attribution as long as the claims are directed against the respondent State. This first line of thought will be defined as the “formalistic approach” (2.1).

The second approach includes those cases in which tribunals have accepted the idea that a certain inquiry on the links existing between the State and the enterprise shall be carried out in order to satisfy itself that it has jurisdiction (2.2). This approach is the one followed by the largest number of tribunals, but it has been implemented through two different methods: one based on an analysis modelled on the “structural – functional test” elaborated by the Maffezini v. Spain tribunal (2.2.1), and another one, which is more nuanced, employed by the Jan de Nul v. Egypt tribunal (2.2.2).

Lastly, there is a third approach which encompasses a group of cases, starting with Toto Costruzioni v. Lebanon, where it has been expressly recognized that the issue of jurisdiction entails at the same time an inquiry regarding the issue of attribution (2.3).

2.1 The Formalistic Approach

In the first group of cases, the arbitrators were more reluctant to conduct an inquiry into the relationship existing between the State and the State enterprise. Representative of this group is the decision taken in the case Salini v. Morocco of 2001,38 which was based on a construction contract between an Italian investor and adm, a Moroccan State-owned enterprise. Morocco objected that the tribunal lacked jurisdiction ratione personae, because the dispute was between the investor and adm, an entity distinct from the State. The parties discussed at length whether adm could be regarded as a State entity, but the tribunal decided to enter into this analysis only “to satisfy the legitimate expectations of the Parties”, pointing out that this investigation was not necessary “since the claims […] are being directed against the State and are founded on the violation of the Bilateral Treaty”.39 Salini v. Morocco does not represent an isolated finding; it was upheld in the contemporary pronouncement in Consortium rfcc v. Morocco40 and was rather successful among many other tribunals, as shown by the Helnan v. Egypt41 decision. In these cases, the tribunals supported the idea that the issue of attribution pertains exclusively to the merits phase when the arbitrators are called upon to decide issues of State responsibility. Scholars have criticized, and to our understanding rather convincingly, this formalistic approach, observing that such a method allows tribunals to verify whether the respondent State is a party “to the proceeding, but not whether it is a party to the dispute”.42 It seems, in fact, that the tribunal has conflated two different issues: the referral to the tribunal (saisine) and its jurisdiction. The inquiry on jurisdiction, in fact, represents the exercise of the inherent jurisdiction (compétence incidente) of a court of law, while, on the contrary, the referral (saisine) represents only the condition for its exercise.43 In other words, such an approach would result in a hidden misapplication of Article 25(1), which, according to the interpretation suggested above, was deemed to identify the range of disputes covered by the icsid Convention and not to regulate the formal aspects of the referral of a dispute to a tribunal. This approach recently experienced a sort of “revival” in the decision of the Tulip Real Estate ad hoc Committee, which demonstrates that the topic at issue is constantly evolving.

2.2 The Prima Facie Approach

The two lines of thought grouped under the second approach may be regarded as an “intermediate” position, between the formalistic approach (2.1) and the most recent one, which will be examined in the subsequent section (2.3). Indeed, the said two lines of thought that belong to this “intermediate approach” share a common feature whereby, at the jurisdictional phase, the tribunal shall limit itself to a prima facie inquiry. This is easily verifiable based on the assertion by many arbitral tribunals that their refusal to apply the rules on attribution was due to the fact that, according to their view, the scope of application of those rules was limited to the field of State responsibility, an issue pertaining exclusively to the merits. However, within this second approach, or “intermediate area”, it is possible to make a distinction between an “early approach”, i.e., the one elaborated in the Maffezini v. Spain decision, and a more recent approach, starting with the Noble Ventures v. Romania and Jan de Nul v. Egypt decisions, in which the tribunals employed the rules on attribution, as elaborated by the International Law Commission, to decide the merits. Nevertheless, the attitude of those tribunals to refrain from applying these rules during the jurisdictional phase remained unchanged.

2.2.1 The Prima Facie Inquiry Based on Structural and Functional Assessment

The first case to be examined in this section is the Maffezini case, which opposed an Argentinian investor to Spain and arose from certain initiatives of a State company, Sociedad para el Desarrollo Industrial de Galicia (sodiga).

The tribunal recognized first of all the importance of verifying whether sodiga could be regarded as a State entity. The tribunal drew inspiration from the above-mentioned Broches test, observing that, although the test had been conceived for the different issue of the jus standi in judicio of a government-owned enterprise, it nevertheless could offer guidance for solving the issue it faced.44 The tribunal elaborated a method of analysis based on two tests: one devoted to ascertaining the existence of an agency relationship, the so-called “structural test”; and a second test, aimed at verifying whether the function carried out by sodiga might be deemed “public in nature”, stressing, however, that it was not applying the rules on attribution.

At the merits stage of the proceedings the arbitrators were called to decide on the issue of the responsibility of Spain for the acts of sodiga. The tribunal, at that point, was forced to admit that it had already settled a significant part of this issue. The tribunal stated that it had to apply “again”45 the structural and functional tests in order to ascertain the issue of responsibility, suggesting that the tribunal had, although it openly denied doing so, employed the rules on attribution already at the jurisdictional stage. The “Maffezini test” represents an important step forward in the interpretation of Article 25(1), because it is the first decision in which the arbitrators had, at least implicitly, recognized the need for an independent inquiry into the existence of their jurisdiction ratione personae and tried to develop a clear test to conduct this inquiry.46 The major drawback of the Maffezini decision, however, is that the arbitrators hid the true line of thought of their reasoning, openly denying having ruled on attribution at the jurisdictional stage.47 They thereby perpetuated the misunderstanding according to which determining attribution would be included in the review of State responsibility, while meanwhile pointing out “it is [only] the juridical attribution of a particular act by a physical person […] to a State, […] whereby it is regarded as the latter’s own act”.48 The other important aspect of the Maffezini decision, while it is not entirely persuasive, is the choice of the arbitrators to ground their analysis on a judge-made test, instead of relying directly on the rules on attribution.49 This might be regarded as a minor defect since, at that time, the 2001 arsiwa had not yet been published and, in any event, the outcome of the Maffezini case would not have been different. However, from a more theoretical point of view, this choice was probably inspired by decisions rendered in the field of international commercial arbitration, which, explicitly or implicitly, relied on the alter ego50 or veil-piercing doctrine.51 Indeed, the reliance of the arbitrators on those doctrines shows that, at the time the decision was issued, the mixed, commercial-public nature of investment arbitration was still prominent.

2.2.2 The Narrow Inquiry

This approach represents the development of a formalistic approach. It had, in fact, a prologue in the case L.E.S.I. Dipenta v. Algeria that represents a hybrid between the pure formalistic approach (2.1) and the approach that has emerged as a trend in the subsequent cases Jan de Nul v. Egypt and Saipem v. Bangladesh (2.2.2). This latter approach is somehow different from the pure formalistic inquiry of the Salini v. Morocco case, because, here, the tribunals have added an “emergency break”. Indeed, the L.E.S.I. Dipenta v. Algeria tribunal maintained that, as a general rule, jurisdiction must be upheld, as far as the claim was directed against the State,52 but added that this presumption had to be abandoned when it was clear that there was no significant State influence on the State-owned enterprise.53 The tribunal went further by upholding the test elaborated in the Maffezini v. Spain decision as the valid method to conduct such an inquiry. In subsequent cases there has been a slight movement toward a more defined approach. In both the Jan de Nul v. Egypt and Saipem v. Bangladesh decisions any reference to the “Maffezini test” was abandoned, leaving room for the application of the rules on attribution elaborated in the arsiwa. However, although more nuanced than in the pure formalistic approach, the resolute refusal to apply those rules at the jurisdictional stage remained unchanged: in fact, all those tribunals maintained that the arbitrators have to retain jurisdiction, unless it appears clear that the enterprise “has no link whatsoever with the State”.54

2.3 Application of the Rules on Attribution at the Jurisdictional Stage

After the decisions analysed above, there was a turning point that can be identified through the decisions in Toto Costruzioni v. Lebanon (decision on jurisdiction issued in 2009 – award 2012) and Gustav Hamester v. Ghana (2010). The first decision was issued in a dispute based on a construction contract signed between an Italian investor and a State entity (cegp – Conseil Executif des Grands Projets – then cdr Council for Development and Reconstruction).55 In Toto Costruzioni, the respondent State raised an objection to jurisdiction ratione personae, because the real dispute was with the State enterprise, which enjoyed a separate legal personality56 (cepgcdr) and had signed the contract containing a different clause of dispute settlement. The tribunal dealt with this objection immediately at the jurisdictional stage and concluded its analysis by stating that it had jurisdiction, because the acts of the State entity were attributable under Article 5 arsiwa.57 The tribunal was extremely clear in its finding, observing that, according to its analysis, “Lebanon may be internationally liable for the acts of the cegp” and, as a consequence, “the Tribunal has jurisdiction ratione personae”. The use of the word “may” is very important, because it traces the perimeter of the finding of the tribunal at the jurisdictional stage. The tribunal, in fact, did not rule on the matter of responsibility at the preliminary stage, which would have driven the tribunal also to assess the legality of the acts of the State entity. Rather, at the preliminary stage, it only examined the relationship between the State entity and the State. One might observe that, as a rule of general international law, State responsibility is composed of two elements: the subjective element, linked to attribution, and the objective one, relating to the breach of the international obligation. Thus, in finding that the acts of the State entity were attributable to Lebanon, the arbitrators were just verifying whether the State was involved.

A similar approach was taken in the decisions in two subsequent cases, Gustav Hamester v. Ghana and Electrabel v. Hungary, and in the award rendered in the case Tulip Real Estate v. Turkey.

The case Gustav Hamester v. Ghana arose from an investment contract signed between the investor and a Ghanaian State-owned enterprise, Cocobod, and one of the first issues for the tribunal to decide was whether the acts of Cocobod were attributable to Ghana. The tribunal, for this purpose, took two, in certain respects contradictory, steps. First, it moved toward the approach followed by the Toto Costruzioni tribunal, observing that “the question of attribution looks more like a jurisdictional question”, since “[it] does not, itself, dictate whether there has been a violation of international law” but “rather, it is only a means to ascertain whether the State is involved”.58

Then, unexpectedly, the tribunal turned towards the Jan de Nul approach, stating that it would retain its jurisdiction unless it was clear that the State-owned enterprise had no link whatsoever with the State. It refused to bifurcate the proceedings and left the full inquiry on attribution to the merits stage.

It seems that the tribunal partially conflated the issue whether attribution had to be considered a matter of jurisdiction or an issue for the merits with the appropriate timing for the tribunal to deal with it. As we have previously seen, an objection to jurisdiction keeps its preliminary character, even if the tribunal exercising its inherent jurisdiction decides to deal with it at the merits stage for its convenience. In this light the tribunal was entitled to deal with the issue of jurisdiction after having heard all the pleadings of the parties. However, the quotation of the Jan de Nul decision is misleading, because in that case the arbitrators did not examine the issue of attribution at the jurisdictional stage as they regarded it as a matter for the merits phase59 and not for convenience, as the Hamester tribunal had been willing to do.

The Hamester tribunal maintained a cryptic approach until the very end of its decision. In the dispositive part, in fact, it declares that it has jurisdiction over the dispute “as presented by the claimant”, only to conclude that the “acts of Cocobod are not attributable to Ghana”. The first finding is not convincing: although, for the prima facie inquiry, the presentation of the claim by the applicant can bear a certain weight, the duty of ascertaining jurisdiction, nevertheless, cannot be fulfilled simply by taking notice of the claimant’s presentation. This issue has been addressed by other tribunals, such as the tribunal in Pan American Energy v. Argentina, which – convincingly – affirmed that, if the tribunal was to accept the declarations of the claimant as they are, the power of the arbitrators to ascertain their own jurisdiction would “be reduced to naught”.60 In other words, to affirm to have jurisdiction over the dispute “as presented by the claimant” is equivalent to giving up the “Kompetenz – Kompetenz” principle, resulting in a misapplication of Article 41(1) of the icsid Convention.

The Hamester award was quoted in the more recent Decision on Jurisdiction, Applicable Law and Liability of 2012 in the case Electrabel v. Hungary. The tribunal faced an objection to jurisdiction of the respondent State, according to which the acts complained of were those of mvm, a State-owned enterprise, and not those of Hungary, as those acts could not be attributable to the latter.61 Apparently, the arbitrators followed the Hamester tribunal, observing that “attribution looks more like a jurisdictional”62 question, but then, after having joined this preliminary objection to the merits, concluded that “considering this issue of attribution as part of the merits of the Parties’ dispute […] the Tribunal rejects the jurisdictional objection made by Hungary as regards the acts of mvm […]”. The reasoning of the tribunal is somewhat obscure because it seems to suggest that the tribunal enjoys full freedom to qualify an objection to jurisdiction as an issue for the merits and to proceed to the analysis of the latter without having first ruled on its jurisdiction. It is worth noticing that, by proceeding in such a way, the tribunal forced Hungary to develop a full argument on the issue of legality for each claim and reached the merits only to conclude that, in any case, the acts of mvm, with one minor exception, were not attributable to Hungary.63 It seems that this decision represents a second attempt to rationalize the treatment of objections to jurisdiction ratione personae as involving the issue of attribution; but, as in the Hamester award, the arbitrators refused to follow their line of reasoning to its natural end.

Some more light on this issue might be cast by the award in the recent case Tulip Real Estate v. Turkey, of 2014, which case arose from certain acts performed by a State-owned company, Emlak. In this case, Turkey had objected that Emlak was an independent commercial entity and, therefore, that the tribunal lacked jurisdiction.

Unlike in Hamester and Electrabel, the arbitrators were more precise in identifying the object of their analysis: in fact, after having found that the acts of Emlak were not attributable to Turkey, they – correctly – stated that, as a consequence, the dispute was out of the tribunal’s jurisdiction.64 This conclusion of the tribunal shows that, in the case at hand, the arbitrators upheld a line of thought closer to the one elaborated in the Toto Costruzioni decision on jurisdiction, keeping intact the qualification of the objection to jurisdiction, even if the tribunal had joined it to its examination of the merits.

Unfortunately, due to a certain disagreement among the arbitrators, the tribunal proceeded further “for the sake of argument” to analyse whether the acts of Emlak could have been regarded as illegal and rejected the claim on the merits. The Applicant, therefore, requested the annulment of the award observing that the tribunal had manifestly exceeded its power by ruling on an issue over which it had no jurisdiction.65

As a consequence, the ad hoc Committee had occasion to discuss the whole matter of the relation between attribution and jurisdiction and its findings are highly instructive. As we already have seen, the ad hoc Committee conceded that there is a growing trend,66 which supports the necessity of examining attribution in order to rule on the issue of jurisdiction, but the Committee did not show excessive enthusiasm for this development. The ad hoc Committee seemed to prefer an approach closer to the formalistic line of thought that was supported in Salini v. Morocco and Helnan v. Egypt. We think that the Committee was mistaken in this conclusion. In the award of the tribunal, in fact, it is stated that “attribution is relevant in the present context to ascertaining whether there is a dispute with a Contracting State, here Turkey, for the purposes of the bit […]”.67 The fact that the tribunal put some emphasis on the expression “with a Contracting State, here Turkey”, suggests that the tribunal was not willing to ascertain whether there was a dispute (since the facts of the case were not seriously challengeable), but rather to ascertain whether Turkey was a party to that dispute.68 The Committee failed to grasp the true object of the tribunal’s inquiry and it moved on to verify whether, in the case at hand, there was a dispute, relying for this purpose on a previous finding of the icj on the notion of dispute, notably in the Mavrommatis case of 1924.69

This reference, however, was not pertinent since the Committee had shaped the object of its reading by leaving aside the true scope of the contention. As such, the problem before the ad hoc Committee could not be solved through the notion of dispute provided in the Mavrommatis case, but, rather, it presented some analogies to the Norwegian Loans case of 1957.70 This latter case represents, in many respects, a missed occasion for the icj to deal with the issue of State-owned enterprises,71 however, on the basis of the parties’ submissions and the opinions of the judges, it is possible to collect some interesting elements.

In one of its preliminary objections in Norwegian Loans, Norway had affirmed that it was not the real party to the dispute, at least for certain claims, because the dispute was between France and the State-owned banks, which enjoyed a separate legal personality.72 It is interesting to observe that, while for another objection, the previous exhaustion of local remedies, the French government had contested its exclusively preliminary character, with reference to the objection related to the lack of capacity to be sued, France did not contest its jurisdictional character and asked for its rejection. In its last reply, France, with reference to the third objection, identified five indicators of the existence of a relevant legal link between the banks and the respondent.73 These five criteria are clearly very similar to the “structural test” that was elaborated by the Maffezini tribunal fifty years later and that, in certain respects, might be assimilated to Article 4 of the arsiwa. On the other hand, France pointed out that the task of these banks was to grant loans to lower income citizens that, according to its analysis, was clearly a public purpose, thereby anticipating the so called “functional test” that is embodied in Article 5 of the arsiwa.74 As a consequence, the French government concluded that the two State-owned banks were two instrumentalities of the Norwegian State, which might enjoy a certain autonomy according to its internal law, but “in the international order involved directly the responsibility of the Kingdom of Norway”.75 This conclusion is very interesting because it shows that, also before the icj, the parties were conscious that they were debating the issue of attribution, and more in general, of State responsibility while discussing a preliminary objection. Judge Read, in his Dissenting Opinion, was the only member of the Court who addressed this exception and he considered it groundless, because, although the banks “had separate legal personality”, they were “acting on the basis of the advice, instruction and approval of the Minister of Justice of Norway and the Minister of Finance of Norway”.76

This last affirmation leads us to the conclusion that, in certain respects and due to the particular shape of Article 25(1) of the icsid Convention, the recent trend that consists of applying the rules on attribution in order to ascertain whether a State-owned enterprise can be identified with the State itself, is not against general international law, but rather it represents the correct tool for a tribunal to fulfil its duty of ascertaining its own jurisdiction.

3 The Possible Existence of an Implicit Designation – and its Relationship to the Application of the Rules on Attribution

3.1 Designation as Depicted by Article 25(1)

Besides the interpretative problems raised by the expression “Contracting State”, there is another separate, but related, part of Article 25(1) that must be tackled. The text of Article 25(1), in fact, not only refers to the notion of “Contracting State”, but also to a) any constituent subdivision or b) agency.

A first issue to discuss is whether the two limbs are to be read conjunctively or disjunctively. If one follows the first option, between the two parts there exists a sort of hierarchical relationship, the first part being a statement of general character that identifies the class of disputes that are encompassed within icsid jurisdiction, while the second part is to be read as a specification of the first. Therefore, while the first limb sets out an objective requirement for icsid jurisdiction to be established (i.e. the fact that one of the parties must be a “Contracting State”), the scope of the second provides only an option for procedural convenience,77 which, by itself, is not intended to enlarge the scope of application of Article 25(1).

The second possible reading is that the two parts have to be read disjunctively, as the conjunction “or” could indeed suggest. As such, the State, through designation, could expand icsid jurisdiction at its will, the tribunal being barred from inquiring whether the entity meets the objective requirement to be deemed an instrumentality for the purposes of icsid jurisdiction.

The difference between those two readings is significant when, on the respondent side, there is a State-owned enterprise. In fact, while the expression “constituent subdivision” is pretty clear in referring to the territorial sub-units of the State (such as regions, provinces etc.),78 the word “agency” is more elusive, since it may encompass not only State bodies or public authorities, but, also, in certain circumstances, State-owned enterprises.79 If it is assumed that Article 25(1) specifies an objective condition, then, although the State might have designated a State-owned enterprise as respondent party, it is always a duty of the tribunal to ascertain whether the requirements set out in Article 25(1) are met. On the other hand, if the two parts are to be read disjunctively, then the designation by the State could create, by itself, a different and sufficient ground for jurisdiction.

The significant implications from the conflict between these two approaches are clear, both from a theoretical and a practical standpoint. If we follow the first option, Article 25(1) being an objective condition, the arbitral tribunal is compelled to apply, even ex officio, the rules on attribution, to verify whether the entity designated by the State can be considered an agency for the purpose of Article 25(1); if, instead, we accept the other approach, we would assume that the arbitral tribunal should undertake a different inquiry, aimed at verifying only if the State has explicitly or (and this is the most sensitive hypothesis) implicitly designated the State-owned enterprise to be a party to the icsid proceedings. Notwithstanding the affirmed differences, one might still doubt the relevance of such a dissimilarity, at least from the point of view of the duty of the arbitrators to ascertain their jurisdiction. The most authoritative doctrine, in fact, has observed that a designation does not, ex se, create certainty about the existence of the jurisdiction ratione personae of the tribunal, but rather “a strong presumption”,80 the final word being for the tribunal to state.81 However, it is clear that when a “strong presumption” is triggered, it would be difficult for the respondent to meet the high standard of proof necessary to rebut that presumption. This conclusion is partially confirmed by the reading of the decision in a recent case, Standard Chartered Bank v. Tanzania Electric Supply Company Limited, of 2014,82 where the tribunal, in the absence of an objection by the respondent, accepted the explicit designation of a State-owned enterprise as a conclusive circumstance, but only after an extremely limited inquiry regarding the structure of the company.

In conclusion, the two different approaches reflect two possible ways of understanding the functioning of an icsid tribunal. The first one, which regards it as an adjudicative body construed and regulated by public international law, which, following the case law of the pcij83 and the teachings of authoritative scholars,84 is compelled to ascertain autonomously the existence of its own jurisdiction.85 On this point, however, it must be said that arbitrators have been generally reluctant to proceed ex officio, limiting the examination of the objection to jurisdiction only in the case of the default of a party. This fact shows, together with some examples taken from the practice, that arbitrators have not been very proactive in this respect.86

The second approach, on the contrary, pays stronger attention to the will of the parties and it is much closer to the practice in international commercial arbitration, where the parties enjoy larger freedom.87 It has often been said that investment arbitration lies at the corner of international public law and commercial law,88 even if, as it emerges from the present article, it is gradually shifting toward public international law.

3.2 Arbitral Decisions Dealing with Designation

As anticipated in the previous section (3.1), according to one of the possible interpretations of Article 25(1), the State, through designation, may create an independent ground for jurisdiction. A good way to prove the possible well-foundedness of this hypothesis is to examine some recent cases regarding implicit and explicit designation and to juxtapose them against the other cases analysed above.

The first case is a case with a decision on jurisdiction published in 2013, Niko Resources Ltd. v. Bapex and Petrobangla,89 which was based on a joint venture agreement (jva) that deferred the solution of disputes between the parties to icsid. When called to decide on the existence of its jurisdiction, the tribunal analysed the different grounds for jurisdiction alternatively proposed by the claimant. The argument put forward by the claimant was that Bapex and Petrobangla were to be deemed instrumentalities of the government and they were acting as agents of their controller, the State, therefore Bangladesh was the real party to the dispute. The tribunal conceded that, given the stringent State control over them and their function, Bapex and Petrobangla were to be considered as State instrumentalities;90 however, relying on the contract wording, which named as parties only the investor, Bapex and Petrobangla, it concluded that the jva did not evidence that those two companies were acting as agents of Bangladesh.91

The second argument of the claimant was that the acts of the State oil company and its subsidiary were attributable to the State of Bangladesh. Since the relationship between Petrobangla and the State of Bangladesh had been already considered by two other icsid tribunals, on the basis of the rules on attribution and in order to inquire whether they had jurisdiction over the case before them, the Niko tribunal intended to deal with those precedents. The findings of the tribunal, in this respect, are very interesting because the arbitrators engaged in a careful examination of the two precedents that applied the rules on attribution at a preliminary stage of the proceedings to distinguish those cases from the case before them. The first precedent that was taken into consideration was Saipem v. Bangladesh,92 but the arbitrators limited themselves to observing that, as the cited decision was based on a treaty claim, the issue of consent was different: in Saipem the State’s consent was incorporated in a bit, therefore, it was not seriously challengeable.

In the other case that was examined, Chevron v. Bangladesh,93 the dispute had arisen from a series of contracts, one of which was signed directly by the Bangladeshi government. The tribunal concluded that the two contracts “did indeed form a unified contractual scheme […]”,94 and, given that circumstance, the Chevron tribunal was at ease in finding that it had jurisdiction over the whole contract claim. The Niko tribunal seems to read the Chevron award as a tool to transpose the doctrine of implicit consent born in the field of international commercial arbitration into the realm of investment arbitration.95 This approach should be regarded with skepticism. Doubts about the opportunity to apply the same rules that have been shaped to deal with private actors to State companies have been voiced even when certain scholars have commented on international commercial awards. For example, Professor Rogers Alford criticized the ruling of the u.s. Court of Appeal of the Fifth Circuit in Bridas s.a.p.i.c. v. Government of Turkmenistan96 of 2006, in which the Court considered unjustified, and therefore invalidated, the decision of an icc arbitral tribunal to extend to the Government of Turkmenistan the arbitration clause contained in the agreement concluded between Bridas and entities owned by the Turkmen government, basing its decision on the “alter ego” and other doctrines born in the field of private commercial law. The author observed, correctly, that: “As difficult as it may be to develop a more accurate analysis of the circumstances under which a sovereign non-signatory should be bound by the signature of its instrumentality Bridas certainly underscores the hazards of freely transposing criteria established in the private sector to government contracts”.97

Further to those general considerations, what the Niko tribunal did not explain is why, if such a finding was sufficient, the Chevron tribunal should have, nevertheless, deemed it necessary to rely upon the rules on attribution. What is particularly interesting in this respect (and what the Niko tribunal omitted to consider) is that the ground for attribution identified by the Chevron tribunal was Article 4 of the arsiwa, the only article that allows for an undifferentiated attribution of both acta iure imperii and acta iure gestionis,98 thus providing the tribunal with another self-sufficient basis for jurisdiction when dealing with a contract.

What the Niko tribunal failed to recognize, at least expressly, is that in its decision the rules on attribution had already been applied when the tribunal recognized that Bapex and Petrobangla were agencies of the respondent.99 This conclusion is self-evident if ones takes into account the method applied by the tribunal to reach this deduction: not surprisingly the relation between the government of Bangladesh and these two State corporations arose from the application of the functional100 test elaborated in the Maffezini decision, that, as we have already seen, is substantially equivalent to the test set out in Article 5 of the arsiwa. Once the tribunal had accepted that Bapex and Petrobangla were in fact agencies of the Bangladeshi government, it moved further by observing that the jva, which contained the arbitration dispute settlement clause, had been approved by the competent ministry of the Bangladeshi government. The tribunal concluded that this approval could in fact be read as an implicit designation of the two State companies to be part of these proceedings. Still, it must be recalled that the tribunal was able to reach this conclusion only after having qualified Bapex and Petrobangla as instrumentalities because of their evident links with the State.

This conclusion can be underpinned by a further analysis of the above cited Hamester v. Ghana case. As already observed, the claim was based on the Germany – Ghana bit, but it must be recalled that there had been a first attempt to bring the case on the basis of the joint venture agreement between the investor and the State-owned enterprise, which contained a dispute settlement clause referring to icsid. At a certain point during the proceedings, one of the arbitrators asked why the investor had not chosen to bring its case on the basis of a contract and the answer was that “Hamester did try to submit such a claim, but the Centre declined jurisdiction on the grounds that Cocobod is not a designated constituent subdivision or agency of Ghana under Article 25(1) of the Convention”.101

This affirmation can be easily read in the sense that the Centre declined jurisdiction simply because there was no designation from the respondent State, without, however, clarifying whether the Centre had concluded that there was no other way to declare himself competent to decide the case.

It is reasonable to imagine that the Centre had not seen any sort of relationship between Cocobod and Ghana that would suggest that it might have supported the retaining of its jurisdiction, because Cocobod could be regarded as an agency or instrumentality of Ghana. If this was the implicit reasoning of the Centre, it would also explain why, not surprisingly, the attempt to repackage the contract claim as a treaty claim, on the alleged assumption that the acts of Cocobod were attributable to the respondent, was doomed to fail. The test to be applied, in order to verify whether Cocobod could qualify as an instrumentality for the purposes of Article 25(1) or whether its acts could have been attributable to Ghana, would have been substantially the same.

Elements in favour of this conclusion might be drawn from the decision in the recent case, Standard Chartered Bank (Hong Kong) Limited v. Tanzania Electric Supply Company Limited, of 2014. Also in this case the tribunal had to deal with a State-owned enterprise, but in this case the entity had been explicitly designated by the State of Tanzania to be a party to the icsid proceedings. When coming to the issue of jurisdiction, the tribunal held, correctly in our view, that even if the respondent did not raise any objection, it “must examine its jurisdiction in light of Article 25 of the icsid Convention.”102 When inquiring into the existence of its jurisdiction, the arbitrators limited themselves to observing that “tanesco is an entity wholly owned by Tanzania and designated as an agency of Tanzania pursuant to Article 25(1) of the icsid Convention”.103 This sentence might be read in favour of the thesis that is being supported here, in particular if one attaches a certain weight to the passage where the tribunal refers to the ownership by the government. In this respect, it must be said that majority ownerships have been regarded as an element of the structural test. In the Maffezini case, for instance, the arbitrators went very far in that direction, stating that “majority ownership creates a rebuttable presumption of statehood”,104 a dictum which has been reported by many distinguished scholars,105 apparently without contestation. It is true that the existence of such a presumption was squarely rejected only a month later by the Tulip Real Estate106 tribunal, but at the time of the Tanesco decision the arbitrators would have found significant judicial precedents and reputable scholars in favour of their approach. There are, therefore, good reasons to conclude that the reference to majority ownership could be regarded as an extremely synthetic way to consider whether Tanesco could be qualified as an agency and, for this purpose, the tribunal partially applied the same test that Maffezini and other tribunals had employed in order to establish attribution for jurisdictional purposes.

4 Conclusion

This article has probed the different interpretations of Article 25(1) of the iscid Convention. The impression created is that there is an on-going clash between the two different ways that can be relied on to identify the scope of icsid jurisdiction. The way that is used is related to the fact that arbitrators do not take too literally their duty to ascertain independently their own jurisdiction. This is understandable considering the fact that tribunals, although increasingly obliged to apply public international law, are still prominently composed of arbitrators coming from the international commercial arbitration world, where the will of the parties represents the cornerstone of the dispute settlement mechanism.

Such an approach, which takes into consideration the specific character of international investment arbitration, cannot help but recommend that arbitrators should be more proactive in the use of their powers. As we have demonstrated, there has been a growing trend toward a more demanding approach. This is a positive evolution. The point of view expressed by the Tulip Real Estate ad hoc Committee, although highly authoritative, is not very persuasive. Even if the definition of dispute provided by the Committee is correct, we suggest that such a definition fails to grasp the real crux of the matter, which is linked to the hybrid nature of a State-owned enterprise. We have demonstrated that, although it is true that Article 25(1) of the icsid Convention does not make any reference to the issue of attribution, it is substantially impossible for arbitrators to ascertain the existence of their own jurisdiction without – implicitly or explicitly – applying the rules on attribution. This is the case, as we have seen, when the tribunal, instead of verifying directly whether the acts of the State-owned enterprise are attributable to the respondent, shifts the object of its inquiry to whether those enterprises can qualify as an instrumentality for the purposes of Article 25(1). In this instance, while referring to two apparently distinct requirements, the tribunal is substantially conducting the same inquiry, and applying the same method for the same purpose. We have seen that, when the same problem was brought before the icj, even if the issue was not definitively settled by it, both the parties and the judges who discussed this matter found that it was indeed reasonable to raise further jurisdictional inquiries, even if they implicitly touched the issue of State responsibility.

With regard to the link between the two constitutive elements of the latter, attribution and legality, we think that, in certain, very complex cases, it might be difficult to differentiate one from the other; however, at least from a theoretical point of view, they can always be distinguished, as the Toto Costruzioni tribunal exemplarily illustrated. As a consequence, arbitrators should always keep in mind both the opinion of Judge Huber, according to which the joining of the objection to jurisdiction to the merits should be limited to exceptional circumstances in order not to compel a State to remain in proceedings for longer than necessary, and the practice of the icj, according to which the interest of the respondent that the case not be discussed before the tribunal shall be duly considered. We are aware that our observations touch upon the realm of discretion of a tribunal with reference to the use of its inherent jurisdiction. However, we submit that the exigency to rule on jurisdiction before moving on to the merits would render arbitrators more aware of the differences existing between the subjective and objective elements in the field of State responsibility and of the advantages, in terms of clarity and proper conduct of the proceedings, when choosing to rule, first, on the subjective element at a preliminary phase and, then, on the objective element at the merits stage.

1 Tulip Real Estate Investment and Development Netherlands B.V. v. Republic of Turkey, icsid Case No. arb/11/28, Decision on Annulment, 30 December 2015.

2 Toto Costruzioni Generali S.p.A. v. The Republic of Lebanon, icsid Case No. arb/07/12, Decision on Jurisdiction, 11 September 2009.

3 On the existence of such a duty in the icsid system, see Monique Sasson, “Investment Arbitration Procedure”, in M. Bungenberg, J. Griebel, S. Hobe, A. Reinisch (eds.), International Investment Law, A Handbook (2015), 1288, 1345. See also Bernardus Henricus Funnekotter and others v. Zimbabwe, icsid Case No. arb/05/6, Award, 15 April 2009.

4 Convention on the Settlement of Investment Disputes between States and Nationals of Other States (opened for signature 18 March 1965, entered into force 14 October 1966), Article 25 (1).

5 See regarding this aspect Historique de le Convention cirdi, Documents Relatifs à l’Origine et à l’Elaboration de la Convention pour le Règlement des Différends Relatifs aux Investissements entre États et Ressortissants d’Autres États, Vol. iii, Washington d.c., cirdi (1968), 381.

6 Bernard Audit, Transnational Arbitration and State Contracts: Findings and Perspectives (1988), 89: “The starting principle is therefore that the creation of a separate entity gives rise to a presumption of effective separation”. See also Fouad A.M. Riad, “L’entreprise Publique et Semi-publique en droit international privé”, 108 r.c.a.d.i. (1963), 566, 586–587. Ignaz Seidl-Hohenveldern, Corporations in and Under International Law (1987), 55–56.

7 For an extensive analysis of the links existing between States and State-owned enterprises, see, for example, Benjamin L. Liebman, Curtis J. Milhaupt (eds.), Regulating the Visible Hand, The Institutional Implications of State Capitalism (2016).

8 Gerold Zeiler, “Jurisdiction, Competence, and Admissibility of Claims in icsid Arbitration Proceedings”, in C. Binder, U. Kriebaum, A. Reinisch and S. Wittich (eds.), International Investment Law for the 21st Century – Essays in Honour of Christoph Schreuer (2009), 80: “It must be remembered that icsid is an arbitration institution. […]. Somewhere in the rules of such an institution, there needs to be a provision which stipulates the prerequisites under which the services of the institution are offered.”

9 Mario Amadio, Le Contentieux International de l’investissement privé (1967), 101: “Ratione personae, la compétence du Centre est spécialisée à un type précis de différend international. Le Centre n’a vocation à connaitre ni les différends interétatiques ni entre les personnes privées. Il ne peut connaitre que des différends de nature mixte entre ‘un Etat contractant […] de lui et le ressortissant d’un autre Etat contractant’ ”. Aron Broches, “The Convention on the Settlement of Investment Disputes: Some Observations on Jurisdiction”, 5 Columbia Journal of International Law (1966), 263, 265.

10 Legality of Use of Force (Serbia and Montenegro v. Belgium), Preliminary Objections, Judgment, i.c.j. Reports 2004, p. 279, para. 46: “The Court can exercise its judicial function only in respect of those States which have access to it under Article 35 of the Statute. […]. It is the view of the Court that it is incumbent upon it to examine first of all the question whether the Applicant meets the conditions laid down in Articles 34 and 35 of the Statute and whether the Court is thus open to it.” See also Mohieddine Mabrouk, Les Exceptions de procédure devant les jurisdictions internationales (1966), 78.

11 Chittharanjan Felix Amerasinghe, Jurisdiction of Specific International Tribunals (2009), 437: “The provisions of the Convention specify in addition to consent broadly three other requirements which must be satisfied, if an arbitral tribunal under the Convention is to have jurisdiction. These are reflected in Article 25 of the Convention […]. The third requirement concerns the nature of the parties. One must be a State”.

12 Appeal from a Judgment of the Hungaro/Czechoslovak Mixed Arbitral Tribunal, p.c.i.j., Series A/B, No. 61, Judgment of December 15th, 1933, p. 221.

13 Christoph H. Schreuer, et al. (eds.), The icsid Convention, a Commentary (2013), 530: “But it must be remembered that not all the Convention’s jurisdictional requirements are subject to the parties’ disposition […]. Therefore, the Tribunal […] cannot rely on the parties’ understanding when it come to the Convention’s objective requirement”. Eduardo Savarese, La nozione di Giurisdizione nel sistema icsid (2012), 24: “La disposizione dunque identifica gli elementi indefettibili per l’esercizio della giurisdizione da parte dei tribunali icsid i quali devono verificare d’ufficio la sussistenza di questi requisiti in base al principio cd. Kompetenz – Kompetenz espressamente riconosciuto dall’Articolo 41. Comma 1 – The tribunal shall be judge of its own competence”.

14 Aron Broches, “The Convention on the Settlement of Investment Disputes Between States and Nationals of other States”, 136 r.c.a.d.i. (1972), 331, 355.

15 Paul Blyschak, State Owned Enterprises and International Investment Treaties, 6 Journal of International Law and International Relations (2011), 1, 35.

16 Mark Feldman, “State Owned Enterprises as Claimant in International Investment Arbitration”, 31 icsid Review (2016), 27–28.

17 Emilio Agustín Maffezini v. The Kingdom of Spain, icsid Case No. arb/97/7, Decision on Jurisdiction, 25 January 2000, para. 79.

18 The Court faced a request submitted by the Ukrainian State to set aside an ad hoc uncitral award, rendered in a case which had opposed a Russian State-owned company, tatneft, and the Republic of Ukraine on the basis of the Russia-Ukraine bit of 1998. The Applicant contended that the arbitral tribunal had exceeded its jurisdiction, ruling over a dispute that was in fact between the two States, as, according to its analysis based on Article 8 arsiwa, tatneft was a mere shell company pervasively controlled by the Russian State. See Cour d’Appel de Paris, Sentence du 29 novembre 2016, n° 14/17964, p. 3.

19 Ibid., p. 4. The wording of the judgment of the Court of Appeal is hard to construe. In particular, the words “propre droit à l’arbitrage” could encompass many procedural aspects, from the issue related to the existence of an autonomous cause of action (droit à agir) to the related matters of the jus standi in judicio (qualité pour agir) and the interest (intérêt à agir). See, regarding this aspect Charles de Visscher, Aspects récents du droit procédural de la Cour International de Justice (1966), 107.

20 According to the old adage “pas d’intérêt, pas d’action”, in fact, it is not possible to identify an autonomous power to sue, when an interest is lacking. The tribunal should recognise that when the interest pursued is not that of the State company, but, rather, that of the State, the company would lack autonomous jus standi (qualité pour agir). On the tight relationship between interest and jus standi before the icj, see supra, note 19, at 75. See also Roger Perrot, Cours de Droit Judiciaire privé (1972–1973), 75. We maintain that the above-mentioned reasoning is in compliance with the finding of the most recent literature in this field, which has identified the unity of interest between the State and the State company as a condition that, together with other indicators, empowers the tribunal to apply the veil-piercing doctrine, that the author considers to be “a special form of attribution”. We do not fully share the author’s view on this point, but we consider that in some circumstances, by referring to veil-piercing, the arbitrators have implicitly applied the test for attribution. On this point, see Albert Badia, Piercing the Veil of State Enterprises in International Arbitration (2014), 201–202. On the possibility of employing the control test in order to examine the absence of autonomy at the jurisdictional stage, see also Bernard Hanotiau, Complex Arbitrations: Multiparty, Multicontract, Multi-issue and Class Actions (2006), 84.

21 James Crawford, State Responsibility: The General Part (2013), 113.

22 Michael Feit, “Responsibility of the State under International Law for the Breach of Contract Committed by a State-Owned Entity”, 28 Berkeley Journal of International Law (2010), 142, 145.

23 Separate Opinion of Judge Sir Gerald Fitzmaurice, in Case concerning the Northern Cameroons (Cameroon v. United Kingdom), Preliminary Objections, Judgment of 2 December 1963: i.c.j. Reports 1963, p. 15, at 92.

24 Emmis International Holding, b.v. Emmis Radio Operating, b.v. Mem Magyar Electronic Media Kereskedelmi És Szolgáltató Kft v. Hungary, icsid Case No. arb/12/2, Decision on Respondent’s Application for Bifurcation, 13 June 2013.

25 Churchill Mining plc and Planet Mining Pty Ltd v. Republic of Indonesia, icsid Case No. arb/12/14 and 12/40, Procedural Order No. 8, 22 April 2014, para. 12.

26 Georges Abi-Saab, Les exceptions préliminaires dans la procédure de la Cour Internationale (1967), 86: “Les pouvoirs qui constituent cette compétence incidente sont prévus dans plusieurs des articles du Statut et du Règlement de la Cour: ainsi le pouvoir de statuer sur sa compétence (article 36, para. 6 du Statut) sur les exceptions préliminaires en général (article 62 du Règlement) […] et en général, tous les pouvoirs nécessaires pour le contrôle de la direction du procès et de la bonne marche de la procédure (article 48 du Statut)”.

27 p.c.i.j., Series D, Addendum No. 2, Revision of the Rules of Court of June 26th, 1926, p. 89: “Moreover, for that very reason, it did not seem possible, except in particular cases, for instance under a general or special agreement between the Parties, to join the question of jurisdiction and the merits. Before an international court, which had to apply international law, questions of jurisdiction should be regarded as independent and preliminary. Whatever difficulties might result – they must bow before the fundamental principles of international law.”

28 p.c.i.j., Series D, Elaboration of the Rules of Court of March 11th, 1936, Third Addendum to No. 2, 1936, 647.

29 Ibrahim F.I. Shihata, The power of the International Court of Justice to determine its own jurisdiction (1965), 87.

30 Ambatielos case (jurisdiction), Judgment of July 1st, 1952: i.c.j. Reports 1952, p. 28, Individual Opinion of Judge Levi Carneiro, p. 48 (“I consider that such a joinder should only be made when it is absolutely necessary”).

31 Supra note 19, at 107.

32 Barcelona Traction, Light and Power Company, Limited, Preliminary Objections, Judgment, i.c.j. Reports 1964, p. 4, at 44.

33 See, for instance, Emmis International Holding, b.v. Emmis Radio Operating, b.v. Mem Magyar Electronic Media Kereskedelmi És Szolgáltató Kft v. Hungary, icsid Case No. arb/12/2, Decision on Respondent’s Application for Bifurcation, 13 June 2013, para. 37: “The Tribunal has, by virtue of Article 41 of the icsid Convention and Arbitration Rule 41, discretion as to whether to suspend the proceedings on the merits and determine a jurisdictional objection as a preliminary issue or whether to join it to the merits”.

34 Shabtai Rosenne, “The International Court of Justice: Revision of Articles 79 and 80 of the Rules of Court”, 14 Leiden Journal of International Law (2006), 78, 79.

35 Jay E. Grenig, International Commercial Arbitration (2015), 206.

36 Nigel Blackaby, Constantine Partasides qc, Alan Redfern, and Martin Hunter, Redfern and Hunter on International Arbitration (ed. 2009), § 9:25.

37 See, for example, Gustav F W Hamester GmbH & Co kg v. Republic of Ghana, icsid Case No. arb/07/24, Award, 18 June 2010.

38 It is important to observe that this decision was dispatched to the parties just one year after the Maffezini case, but it followed a completely different logic.

39 Salini Costruttori S.p.A. and Italstrade S.p.A. v. Kingdom of Morocco, icsid Case No. arb/00/4, Decision on Jurisdiction, 16 July 2001.

40 Consortium rfcc v. Kingdom of Morocco, icsid Case No. arb/00/6, Decision on Jurisdiction, 16 July 2001, para. 34.

41 Helnan International Hotels A/S v. The Arab Republic of Egypt, icsid Case No. arb/05/19, Decision of the Tribunal on Objection to Jurisdiction, 17 October 2006, para. 91: “The Arbitral Tribunal does not need to decide on the status of egoth in order to assess its jurisdiction in this case. It has already found that the Claimant has established the existence of a prima facie dispute directly arising out of an investment and involving egypt […]. However, since the parties have thoroughly discussed this point, it considers ex abundanti cautela that it is its duty to solve this disputed issue in this Decision”.

42 Yves Nouvel, “Les entités paraétatiques dans la jurisprudence du cirdi”, in C. Leben (ed.), Le contentieux arbitral transnational relatif à l’investissement (2006), 31: “Sur ce point précis, la solution retenue dans l’affaire Salini paraît très formaliste. Certes, elle démontre que l’Etat est partie à l’instance, mais elle n’établit pas qu’il est partie au différend. Il ne suffit pas qu’un investisseur introduise une requête contre un Etat sur la base de la convention bilatérale pour qu’un réel conflit d’intérêts juridiques existe entre les parties”.

43 Supra note 29, at 87; see also supra note 26, at 86.

44 Supra note 17, at para. 79.

45 Ibid., paras. 51–52. The fact that the tribunal ruled on attribution at the preliminary stage has been a step confirmed by certain scholars as well. See, for example, Kay Hober, “State Responsibility and Attribution”, in C. Schreuer, P. Muchlinski, F. Ortino (eds.), The Oxford Handbook of International Investment Law (2008), 558: “It follows […] that the Tribunal did in fact rule on the question of attribution already in the jurisdictional decision, notwithstanding its statement that it would deal with that issue together with the merits”. On the same issue, see Simon Olleson, “Attribution in Investment Treaty Arbitration”, 31 icsid Review (2016), 457, 467. See also Enrico Milano, “The Investment Arbitration between Italy and Cuba: The Application of Customary International Law under Scrutiny”, 11 The Law and Practice of International Courts and Tribunals (2012), 499, 511–520. The author holds that the two operations [of ascertaining State responsibility and verifying which are the true parties to the dispute for jurisdictional purposes] should be kept conceptually distinct as attribution on the merits: “[…] is a mere application of general international law secondary rules to a factual situation in order to draw legal consequences”; whereas in a jurisdictional analysis “[…] the secondary rules are extrapolated from their original context to apply the terms of a convention”. The Maffezini test was employed, albeit exclusively at the merits stage, by the arbitral tribunal in the Italy v. Cuba award of 2008.

46 The Maffezini approach was later employed by the Chevron v. Bangladesh tribunal, which quoted directly the Maffezini decision as a point of departure for its inquiry. See Chevron Bangladesh Block Twelve, Ltd. and Chevron Bangladesh Blocks Thirteen and Fourteen, Ltd. v. People’s Republic of Bangladesh, icsid Case No. arb/06/10, Award of 17 May 2010 (Buergenthal, Beechey and Nariman), icsid Review (2011): “In the Maffezini case, the Tribunal concluded that the entity’s conduct could be attributed to the Spanish government, because it had been created by government decree; it had been created to carry out governmental functions in the field of regional development and it was 88% State owned”.

47 Supra note 17, at para. 89.

48 Bin Cheng, General Principles of Law as applied by International Courts and Tribunals (1953).

49 Luca Schico, “Attribution and State Entities: Diverging Approaches in Investment Arbitration”, 12 The Journal of World Investment & Trade (2011), 289.

50 Eduardo Silva Romero, “Are States Liable for the Conduct of Their Instrumentalities?”, in E. Gaillard, J. Younan (eds.), State Entities in International Arbitration (2008), 49: “To support such holdings some icc arbitral tribunals have relied on the theory of the alter ego. In their awards these tribunals concluded that, although the signatory to the agreement was a legally separate entity distinct from the State, it had acted for the State, which was the signatory’s disclosed or undisclosed principal”. The alter ego test is a doctrine born in the U.S. courts in 1921 and then imported into the field of international commercial arbitration. Its main features are quite similar to the test applied by the Maffezini tribunal. See on this point Karen Vandekerckhove, Piercing the Corporate Veil (2007), 83. See also William W. Park, “An Arbitrator’s Dilemma: Consent, Corporate Veil and Non-Signatories”, in Permanent Court of Arbitration (ed.), Multiple Party Actions in International Arbitration (2009), 1–4. For practical applications, see, for example, Bridas v. Turkmenistan, icc Case No. 9151, Interim Award, 8 June 1999, paras. 564–569. See also, more recently, Capital India Power Mauritius I and Energy Enterprises (Mauritius) Company v. India, icc Case No. 12913/ms, International Court of Arbitration, Award, 27 April 2005: “The state had the authority to control the mseb board by appointing its members, removing them at pleasure, setting and directing its policies, and totally controlling its funding. The commitments made by mseb in the ppa were, therefore, entirely dependent on the willingness of som.”

51 See, for example, icc Case No. 8385, 1995, in Jean-Jacques Arnaldez, Yves Derains, Dominique Hascher, Collection of icc Arbitral Awards (1997), 479.

52 Consorzio Groupement l.e.s.i. – Dipenta v. People’s Democratic Republic of Algeria, icsid Case No. arb/03/08, Award, 10 January 2005, para. 19 (i).

53 Ibid., para. 19 (ii).

54 Jan de Nul n.v. Dredging International n.v. v. Arab Republic of Egypt, icsid Case No. arb/04/13, Decision on Jurisdiction, 16 June 2006, para. 85: “An exception is made in the event that if it is manifest that the entity involved has no link whatsoever with the State”. See also Saipem S.p.A. v. The People’s Republic of Bangladesh, icsid Case No. arb/05/07, Decision on Jurisdiction and Recommendation on Provisional Measures, 21 March 2007, para. 144.

55 Agreement between the Italian Republic and the Lebanese Republic on the Promotion and Reciprocal Protection of Investments, signed on 7 November 1997. Entered into force on 9 February 2000.

56 Supra note 2, at para. 38.

57 Supra note 2, at para. 60.

58 Supra note 37, at para. 143.

59 Supra note 54, at para. 89.

60 Pan American Energy, llc et al. v. The Argentina Republic, icsid Case No. arb/03/13, Decision on Preliminary Objections, 27 July 2006, para. 50.

61 According to the respondent, although 99% of its capital was owned by the Hungarian State, it was not an organ of the Hungarian State, nor was it performing public functions or under the State’s effective control. See Electrabel s.a. v. Hungary, icsid Case No. arb/07/19, Decision on Jurisdiction, Applicable Law and Liability, 30 November 2012, paras. 7.24–7.26.

62 Ibid., para. 7.61.

63 This way of proceeding is confirmed by the arbitrators who stated that “accordingly, the Tribunal does not consider mvm’s conduct […] as attributable to Hungary” and that the subsequent analysis on the legality of these acts was aimed at clarifying “what would have been the outcome […], had some of the impugned conduct been considered as attributable to Hungary”. Those two sentences show that the tribunal proceeded toward such an analysis only to complete a double check, but the decision would already have been complete by observing that the acts were not attributable to the State. This way of proceeding has been followed by other tribunals, such as for example the tribunal in Tulip Real Estate.

64 Tulip Real Estate Investment and Development Netherlands B.V. v. Republic of Turkey, icsid Case No. arb/11/28, Award, 10 March 2014, para. 327: “In conclusion, the Tribunal […] determines by majority that Emlak’s conduct […] is not attributable to the Turkish State and is, on that basis, outside the remit of the Tribunal.” This statement was interpreted by the ad hoc annulment Committee as equivalent to saying that the Tribunal had no jurisdiction over the matter at stake. On this point, see supra note 1, at para. 192.

65 Tulip v. Turkey, icsid Case No. arb/11/28, Award, supra note 64, para. 134.

66 Supra note 1, para. 174.

67 Tulip v. Turkey, icsid Case No. arb/11/28, Award, supra note 64, para. 276.

68 About the possibility to verify the existence of a dispute “ratione personae”, see supra note 19, at 36.

69 Supra note 1, at para. 180: “There is no specific definition of the term ‘dispute’ either in the bit or in the icsid Convention. When interpreting that term, in accordance with Article 31(3)(c) of the vclt, ‘there shall be taken into account […] [a]ny relevant rules of international law applicable in the relations between the Parties.’ As the International Court of Justice (‘icj’) stated, there is: […] established case law on that matter, beginning with the frequently quoted Statement by the Permanent Court of International Justice in the Mavrommatis Palestine Concessions case in 1924: ‘A dispute is a disagreement on a point of law or fact, a conflict of legal views or of interests between two persons’ ”.

70 Case of Certain Norwegian Loans, Judgment of July 6th 1957: i.c.j. Reports 1957, p. 9.

71 See Badia, supra note 20, at 43.

72 Supra note 70, Preliminary Objections submitted by the Government of the Kingdom of Norway, 20 April 1956, para. 31.

73 In particular France observed that: 1) the banks were created by the law; 2) their capital was provided by the government; 3) the Storthing (the Norwegian Parliament) decided on the privileges of the banks and 4) was in charge of the designation of two out of three of their administrators (while the other one was appointed by the king) and 5) for the appointment of their five members of the audit board, supra, note 70, Reply of the Government of the French Republic, p. 407.

74 This attempt to compare the structural test and the functional test was carried out by the tribunal in the case edf v. Romania of 2009. Although the assimilation is far from being entirely persuasive, it grasps the common purpose of the test proposed by the Maffezini decisions and the rules set out by the arsiwa. See edf (Services) Limited v. Romania, icsid Case No. arb/05/13, Award, 8 October 2009, para. 187.

75 Supra note 70, Reply of the Government of the French Republic, at 407.

76 Supra note 70, Dissenting Opinion of Judge Read, at 91.

77 Supra note 13, at 152.

78 Ibid., p. 153: “Therefore it might be concluded that ‘constituent subdivision or agency’ covers every territorial entity below the State”.

79 Ibid.

80 Ibid., p. 154.

81 Supra note 11, at 461.

82 Standard Chartered Bank (Hong Kong) Limited v. Tanzania Electric Supply Company Limited, icsid Case No. arb/10/20, Decision on Jurisdiction and Liability, 12 February 2014.

83 Factory at Chorzow (Germany v. Poland), (Claim for Indemnity) (Jurisdiction), Judgment of July 26th 1927, p.c.i.j., Series A, No. 9, p. 32.

84 Supra note 11, at 461; Gerald Fitzmaurice, “The Law and the Procedure of the International Court of Justice”, 33 British Yearbook of International Law (1958), 28–29: “The failure of the respondent to object to jurisdiction is normally taken as an acquiescence in the validity of the alleged basis of jurisdiction, and is, in itself, sufficient to confer jurisdiction on the basis of forum prorogatum. However in such circumstances the power [to ascertain its own jurisdiction] can still be exercised in regard to questions beyond the issue of consent such as the ability to appear before the Court and the propriety of exercising the jurisdiction conferred”. For a more nuanced view, see supra note 29, at 301.

85 Supra note 13, at 529. It must be recalled that the most eminent scholars have upheld that the same approach shall be applied because “not all of the Convention’s jurisdictional requirements are subject to the parties’ disposition”.

86 See, for example, Maritime International Nominees Establishment v. Republic of Guinea, icsid Case No. arb/84/4, Award, 6 January 1988; Mobil Oil Corporation and others v. New Zealand, icsid Case No. arb/87/2, Findings on Liability, Interpretation and Allied Issues, 4 May 1989.

87 Jérôme Ortscheidt, Christophe Seraglini, Droit de l’arbitrage interne et international (2013), 595: “A priori, on pourrait considérer qu’il convient de présumer qu’en demeurant silencieuses sur la question de la compétence arbitrale, les parties ont entendu couvrir les éventuelles causes d’incompétence et qu’en conséquence quand bien même l’arbitre identifierait de lui-même une difficulté éventuelle à cet égard, il n’aurait pas à l’invoquer de son propre fait”.

88 Giuditta Cordero Moss, Commercial Arbitration and Investment Arbitration Fertile Soil for False Friends?, in C. Binder, et al. (eds.), supra note 8, 782–797.

89 Niko Resources Ltd. v. Bangladesh Petroleum Exploration & Production Company Limited (“Bapex”) and Bangladesh Oil Gas and Mineral Corporation (“Petrobangla”), icsid Case No. arb/10/18, Decision on Jurisdiction, 19 August 2013.

90 Ibid., para. 228.

91 Ibid., para. 241.

92 Saipem v. Bangladesh, supra note 54.

93 Supra note 46, at 256–294.

94 Ibid., p. 270.

95 See, for example, icc Case No. 8910, 1998, at 574. See also Bernard Hanotiau, supra note 20, at 261: “In the field of commercial arbitration this doctrine has been applied in icc Case No. 8910, where the arbitral tribunal identified a group of contracts of which only one contained an arbitration clause, as a valid ground to retain its jurisdiction, because those contracts together formed a unified contractual scheme”.

96 Bridas s.a.p.i.c. v. Government of Turkmenistan, United States Court of Appeals for the Fifth Circuit, No. 04-20842, 21 April 2006.

97 Roger P. Alford, “Binding Sovereign Non-Signatories”, 19 Mealey’s International Arbitration Reports (2004).

98 “It is irrelevant for the purposes of attribution that the conduct of a State organ may be classified as ‘commercial’ or as ‘acta iure gestionis’ ”. International Law Commission, Draft Articles on Responsibility of States for Internationally Wrongful Acts, November 2001, Supplement No. 10 (A/56/10), Commentary, p. 41.

99 Supra note 89, at para. 228: “The Tribunal concludes that both entities are agencies or instrumentalities of the government of Bangladesh in the sense of Article 25 of the icsid Convention”.

100 Ibid., para. 228: “These factors indicate that Petrobangla and bapex exercise functions which are vested in the Government and which the Government has delegated to them”.

101 Supra note 37, at para. 27.

102 Supra note 82, at para. 109.

103 Ibid., para. 110.

104 Supra note 17, para. 77: “The question whether or not sodiga is a State entity must be examined first from a formal or structural point of view. Here a finding that the entity is owned by the State, directly or indirectly, gives rise to a rebuttable presumption that it is a State entity”.

105 See, for example, Jan Ole Voss, The Impact of Investment Treaties on Contracts between Host States and Foreign Investors (2011), 149. Rudolf Dolzer and Christoph Schreuer, Principles of International Investment Law (2nd ed., 2012), 223.

106 Tulip v. Turkey, icsid Case No. arb/11/28, Award, supra note 64, at para. 289. See also Debevoise & Plimpton llp, “Suing State-Controlled Entities in Investment Arbitration: The Demise of the ‘Presumption of Statehood’?”, 5 Arbitration Quarterly (2014), 1, 7.

  • 19

    Ibid., p. 4. The wording of the judgment of the Court of Appeal is hard to construe. In particular, the words “propre droit à l’arbitrage” could encompass many procedural aspects, from the issue related to the existence of an autonomous cause of action (droit à agir) to the related matters of the jus standi in judicio (qualité pour agir) and the interest (intérêt à agir). See, regarding this aspect Charles de Visscher, Aspects récents du droit procédural de la Cour International de Justice (1966), 107.

  • 71

    See Badia, supra note 20, at 43.

  • 80

    Ibid., p. 154.

  • 94

    Ibid., p. 270.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 435 0 0
Full Text Views 1089 202 8
PDF Views & Downloads 3592 1025 92