Jump to Content Jump to Main Navigation
The Law of International Responsibility edited by Crawford, James; Pellet, Alain; Olleson, Simon; Parlett, Kate

Part IV The Content of International Responsibility, Ch.54.1 Other Specific Regimes of Responsibility: Investment Treaty Arbitration and ICSID

Zachary Douglas

From: The Law of International Responsibility

Edited By: James Crawford, Alain Pellet, Simon Olleson, Kate Parlett

From: Oxford Public International Law (http://opil.ouplaw.com). (c) Oxford University Press, 2015. All Rights Reserved. Subscriber: null; date: 24 January 2020

Responsibility of states — Diplomatic protection — Wrongful acts — Attribution — Circumstances precluding wrongfulness — Treaties, interpretation — Vienna Convention on the Law of Treaties — Immunity from jurisdiction, ratione materiae

(p. 815) Chapter 54.1  Other Specific Regimes of Responsibility: Investment Treaty Arbitration and ICSID

The International Centre for the Settlement of Investment Disputes

The International Centre for Settlement of Investment Disputes (ICSID) was established under the 1965 Convention on the Settlement of Investment Disputes between States and Nationals of Other States.1 Provision for ICSID arbitration of foreign investment disputes is now frequently included in bilateral investment treaties, foreign investment laws, and investment agreements. In addition to its regular arbitration procedures, ICSID has also established an Additional Facility for the administration of arbitrations in which the parties do not fulfil the normal jurisdictional requirements under the ICSID Convention. This chapter examines the arbitration mechanism established by the ICSID Convention and governed by the ICSID Arbitration Rules within the framework of the law on the responsibility of States for international wrongs. Arbitration under the Additional Facility Rules is not considered separately.

(p. 816) The State/State and investor/State regimes distinguished

The ICSID Convention embodies two distinct spheres of rights and obligations: one applicable to the legal relationships between the Contracting States, the other applicable to the legal relationships between an investor with the nationality of one Contracting State and the Contracting State that is host to the investment.2 Each of these spheres of rights and obligations gives rise to different regimes of State responsibility, which will be referred to as the ‘State/State’ and the ‘investor/State’ regimes.

(a)  The State/State regime of international responsibility

Within the State/State regime, the Contracting States have committed themselves both to the general obligation to adhere to the law of treaties in the interpretation and application of the ICSID Convention (whether by virtue of the Vienna Convention on the Law of Treaties or by customary international law) and to the specific primary obligations set out in the ICSID Convention itself which are properly classified as treaty obligations opposable by one Contracting State to another. These include the obligations upon a Contracting State:

  1. (a)  to accord various immunities and privileges to the property, assets and communications of the ICSID and to its officers and staff and to persons appearing in ICSID proceedings (articles 18–24);3

  2. (b)  to desist from giving diplomatic protection or bringing an international claim with respect of a dispute involving one of its nationals that has been submitted to ICSID arbitration, save in the case of non-compliance by the respondent Contracting State with the award (article 27);4

  3. (c)  to desist from appealing against ICSID awards or pursuing any other remedy with respect thereto save those remedies envisaged by the Convention itself (article 53);5

  4. (d)  to recognize an ICSID award as binding and enforce the pecuniary obligations imposed by the award within its territory as if it were a final court judgment but subject to the applicable rules on sovereign immunity from execution (articles 54 and 55);6

  5. (e)  to adopt such legislative or other measures as may be necessary to make the provisions of the ICSID Convention effective in its territory (article 69).7

In addition to these specific inter-State obligations arising under the ICSID Convention, the structure and purpose of the Convention necessitates the implication of a further general obligation upon Contracting States:

  1. (f)  to desist from invoking procedures in their municipal courts or passing laws or adopting executive acts that have the effect of frustrating an investor’s recourse to ICSID arbitration.

The general rules of State responsibility for international wrongs govern the consequences of any violation of these inter-State obligations. Although there is no precedent to date, the most likely scenario would be the invocation of the international responsibility (p. 817) of the Contracting State which is host to the investment in question by the Contracting State of the national investor. This bilateral dispute would be subject to the jurisdiction of the International Court of Justice in accordance with article 64 of the ICSID Convention. The waiver of diplomatic protection or other remedies in article 27 of the Convention should not be interpreted to prevent recourse by one Contracting State against another with respect to obligations (a) to (e) as described above, even if the dispute touches upon an existing claim by an investor. This should also be the case with respect to obligation (f) so long as the purpose of the Contracting State’s application to the ICJ is to protect the integrity of the ICSID system rather than to achieve compensation on behalf of its national investor. This purpose could be readily discerned from the remedy requested by the claimant Contracting State (ie a request for a declaratory judgment rather than damages assessed by reference to the prejudice suffered by the investor).

The wording of article 64 of the Convention also leaves upon the possibility that a plurality of injured Contracting States might bring proceedings against another Contracting State in a single claim before the International Court in the sense of article 46 of the ILC Articles. Article 64 is not as unequivocal as other treaty provisions about the possibility of a plurality of interest in the invocation of responsibility. For instance, the Treaty of Versailles allowed ‘any interested Power’ to apply in the event of a violation of the provisions concerning transit through the Kiel Canal, and four States brought proceedings before the Permanent Court of International Justice on this basis in The SS Wimbledon.8 Nevertheless, it might be envisaged that where, for example, a Contracting State declares that its own courts will henceforth exercise an appellate jurisdiction with respect to existing ICSID awards, a plurality of other Contracting States would have the requisite interest under ILC article 42 to invoke the responsibility of that State for a violation of articles 53, 54, and 69 of the Convention, whether or not their nationals were currently award creditors.

(b)  The investor/State regime of international responsibility

Within the realm of investor/State disputes, an ICSID tribunal can have jurisdiction to decide claims founded upon:

  1. (i)  international standards of investment protection in an investment treaty;9

  2. (ii)  contractual undertakings in an investment agreement;10 and

  3. (iii)  relevant municipal law enactments of the host State.11

There is no particular symmetry between the juridical character of the instrument conferring jurisdiction on an ICSID tribunal and its ratione materiae jurisdiction. An ICSID tribunal can, for instance, have jurisdiction by virtue of an investment treaty but nevertheless be competent to decide claims founded upon all three sources of rights and obligations. (p. 818) The important point is that the legal quality of an ICSID arbitration in investor/State cases is the same regardless of the substantive law applied by the tribunal to the claims or the juridical nature of the instrument conferring jurisdiction. An ICSID award is an ICSID award, whether it settles a contractual claim or a claim for expropriation. Likewise, the status of the procedural rules governing ICSID arbitrations does not fluctuate depending on the legal nature of the dispute. Paradoxically perhaps, an ICSID tribunal with jurisdiction founded upon an arbitral clause in an investment agreement to determine contractual claims can be more ‘internationalized’ or ‘self-contained’ vis-à-vis municipal legal systems than an arbitral tribunal with jurisdiction by virtue of Chapter 11 of NAFTA but operating under the UNCITRAL Arbitration Rules or the ICSID Additional Facility Rules.

The sui generis character of the ICSID regime12 for investor/State disputes does generate certain complexities for the assimilation of the law of State responsibility into that regime.13 What is the trigger or the connecting factor for the rules and principles of State responsibility to apply in the context of an ICSID arbitration? Is it the application of international law as the substantive law governing the determination of a particular claim? That would appear to be the consensus revealed in ICSID awards that have decided claims based on investment treaty obligations. But suppose an investment agreement between the investor and host State contains a governing law clause selecting international law or the tribunal decides to apply international law in accordance with article 42. The investor brings claims for breach of the agreement and for expropriation of its assets invested pursuant to the agreement. Is the investor required to exhaust local remedies for the expropriation claim? Surely not, because article 26 is interpreted so as to reverse the presumption that the exhaustion rule applies; viz it does not apply unless there is a specific provision in the instrument conferring jurisdiction which says that it does apply. But article 26 of the Convention concerns the jurisdiction of the tribunal. Article 44(b) of the ILC Articles, which recognizes that the exhaustion rule applies unless is excluded, concerns the invocation of State responsibility.

We have previously assumed that the law of State responsibility applies when the substantive law governing the claim is international law. So does a rule of the applicable substantive law (article 44(b) of the ILC Articles) trump the applicable procedural rule (article 26 of the ICSID Convention) or vice versa? Or, contrary to our original assumption, is article 44(b) of the ILC Articles properly classified as a procedural rule that must give way to the lex specialis in article 26 of the ICSID Convention? Suppose further that the national State of the investor has formally waived its right to demand reparation for the losses caused to its nationals by the expropriation in question. Article 45 of the ILC Articles provides that ‘the responsibility of a State may not be invoked if … the injured State has validly waived the claim’. Is this waiver by the injured State effective in relation to claims governed by international law that are brought by its nationals? Suppose finally that the ICSID tribunal upholds both the contractual claim and the expropriation claim. Does the finding of liability (international responsibility) with respect to the expropriation claim entail that countermeasures may be employed by the national State of the investor against the host State in accordance with article 49 of the ILC Articles?

Some of these questions might not appear to be critical in practice having regard to the existing ICSID jurisprudence. But there is evidence that the difficulties that plague any (p. 819) rationalization of the relationship between the secondary rules of State responsibility and ICSID arbitration are producing controversial outcomes. For instance, investors have attempted to bypass the principle of privity of contract to sue the central government, rather than a local agency which was party to the investment agreement, simply by pleading a claim based on an investment treaty and relying upon international rules of attribution.14

The notion of a ‘sub-system’ of State responsibility

The first step in resolving these conceptual difficulties is to differentiate the State/State regime of international responsibility from the investor/State regime. It has been proposed by the present writer that the investor/State regime should be conceptualized as a ‘subsystem’ of State responsibility.15

International law does not contain a single body of secondary rules of State responsibility for all wrongful acts committed by a State.16 This is particularly evident in the case of international treaties that confer rights directly upon non-State actors, such as the European Convention on Human Rights, the Algiers Accords establishing the Iran/US Claims Tribunal, bilateral investment treaties, NAFTA, the Energy Charter Treaty, and the ICSID Convention. These treaties create mechanisms for non-State actors to invoke the international responsibility of contracting States which transcend the traditional dichotomy between public and private international law. The secondary obligations generated by the implementation of State responsibility in these cases are different in juridical character from secondary obligations that arise on the inter-State plane. It is thus appropriate to consider them as sub-systems of State responsibility that share a distinctive feature in that the new legal relationship which arises upon the commission of the wrongful act is between a State and a non-State actor rather than between two or more States. Unlike in the traditional domains of public international law, the obligations created in special regimes involving non-State actors, such as in the investor/State sphere of the ICSID Convention, ‘are not simply based on the separation of States, and consequently not focused on the anti-parallel exercise of sovereignty by interference of one State in the sovereignty of another State …’.17

The ILC Articles recognise the existence of sub-systems of State responsibility by incorporating an important lex specialis reservation in article 55:

These articles do not apply where and to the extent that the conditions for the existence of an internationally wrongful act or the content or implementation of the international responsibility of a State are governed by special rules of international law.

(p. 820) The commentary to article 55 refers to the examples of the World Trade Organization Dispute Settlement Understanding and the European Convention on Human Rights as regimes that, in varying degrees, displace the rules contained in the Articles on State Responsibility.18 ICSID is another such regime.

(a)  Content of the sub-system of responsibility under the ICSID Convention

It is useful to commence our investigation into the actual content of the secondary rights and obligations established by the ICSID sub-system of State responsibility by determining the extent to which the general rules for the mise-en-oeuvre of responsibility as between States can be transplanted into it. For the purposes of this discussion, those general rules will be taken to be accurately reflected in the ILC Articles.

The ILC Articles carve out the institution of secondary obligations owed to non-State actors in the form of a reservation in article 33 to the scope of obligations set out in Part 2 to the Articles:

1. The obligations of the responsible State set out in this Part may be owed to another State, to several States, or to the international community as a whole, depending in particular on the character and content of the international obligation and on the circumstances of the breach.

2. This Part is without prejudice to any right, arising from the international responsibility of a State, which may accrue directly to any person or entity other than a State.

Investment treaties are mentioned explicitly in the commentary to article 33(2) as giving rise to a situation where a ‘primary obligation is owed to a non-State entity’ and such entity has the possibility of invoking State responsibility ‘on its own account and without the intermediation of any State’.19 The ICSID Convention, although not conferring substantive primary rights upon non-State entities, does nevertheless provide non-State entities with a procedural mechanism for invoking State responsibility. With respect to any violation of this mechanism by a respondent Contracting State, the secondary consequences are not, therefore, governed by Part 2 of the ILC Articles on the ‘Content of the International Responsibility of a State’ by virtue of article 33.20 Furthermore, Part Three on ‘The Implementation of the International Responsibility of the State’ is also inapplicable by its own terms, insofar as it relates exclusively to the invocation of responsibility by an injured State rather than any non-State actor.

The non-applicability of Parts 2 and 3 of the ILC Articles to the ICSID sub-system of State responsibility and the general lex specialis reservation in article 55 have several important consequences.21 First, the rules for the invocation of responsibility in Chapter I of Part Three, including the admissibility of claims, cannot be uncritically transplanted into the ICSID sub-system. Secondly, it cannot be assumed that ICSID tribunals are competent to order the different forms of reparation set out in Chapter II of Part Two. Thirdly, an ICSID award does not create a truly ‘international’ liability at the inter-State level of responsibility such as would be the case, for example, with a judgment of the International Court of Justice. If this were otherwise, then a respondent Contracting State might, for instance, (p. 821) resist the enforcement of an ICSID award in the territory of a non-Contracting States by appealing to sovereign immunity from jurisdiction. The liability created by this subsystem of international responsibility is perhaps more adequately described as having a transnational commercial nature in view of the commercial interests at the heart of the dispute, although in itself this label is in practice unlikely to be of great utility. Fourthly, it is arguable that the lex specialis reservation in article 55 might have the effect of rendering various provisions of Part One of the ILC Articles inapplicable to the ICSID regime. For instance, a measure taken by the host State that causes prejudice to a foreign State might not be internationally wrongful vis-à-vis the national State of the investor because it is a lawful countermeasure directed against a breach of an international obligation by the national State of the investor. The investor might nevertheless argue that the prejudice caused to its private interests by the countermeasures is both justiciable before an ICSID tribunal and liable to attract a remedy in damages. The investor would argue that an investment treaty obligation is owed to the investor directly and any rule precluding wrongfulness as between the host State and the national State of the investor is res inter alios acta.22

(b)  Non-applicability of inter-State rules for invocation of responsibility

The preconditions for the mise-en-oeuvre of responsibility in the inter-State system are set out in Part Three of the ILC Articles and include article 44 on the ‘Admissibility of Claims’:

The responsibility of a State may not be invoked if:

(a)  the claim is not brought in accordance with any applicable rule relating to the nationality of claims;

(b)  the claim is one to which the rule of exhaustion of local remedies applies and any available and effective local remedy has not been exhausted.

The ILC’s commentary to article 44 makes it clear that the rules on the nationality of claims and the exhaustion of local remedies are not merely relevant to the ‘jurisdiction or admissibility of claims before judicial bodies’ but are of a ‘more fundamental character’ insofar as ‘[t]hey are conditions for invoking the responsibility of a State in the first place’.23

The relationship between these rules in article 44 of the ILC Articles and the investor/ State sub-system will be explored in more detail with reference to the natio n ality of claims. The ICSID Convention contains a specific provision regulating the ratione personae jurisdiction of ICSID tribunals in article 25:

(p. 822) (1)  The jurisdiction of the Centre shall extend to any legal dispute arising directly out of an investment, between a Contracting State … and a national of another Contracting State …

(2)  ‘National of another Contracting State’ means:

(a)  any natural person who had the nationality of a Contracting State other than the State party to the dispute [on the relevant date] […]; and

(b)  any juridical person which had the nationality of a Contracting State other than the State party to the dispute [on the relevant date] […] and any juridical person which had the nationality of the Contracting State party to the dispute on that date and which, because of foreign control, the parties have agreed should be treated as a national of another Contracting State for the purposes of this Convention.’

Suppose that the investor asserts claims based on a contractual breach of its investment agreement with the host State and a violation of an investment treaty obligation. The host State files an objection to the ratione personae jurisdiction of the ICSID tribunal and/or objects to the admissibility of the claims on the basis of the tenuous link between the investor and the Contracting State whose nationality is invoked.24 Should the tribunal defer to general international law on the invocation of State responsibility to supplement article 25 of the Convention with respect to both claims? Or only with respect to the claim alleging a violation of an investment treaty obligation? In other words, is the connecting factor to the general international law on the admissibility of claims (i) the submission of a claim governed by international law, or (ii) the status of article 25 of the Convention as a rule of international treaty law? Or (iii) is there no connecting factor that would justify an ICSID tribunal’s reference to general international law on the admissibility of claims?

If the general international law on the admissibility of claims were to supplement article 25 of the Convention by reason of the investor’s reliance on a cause of action grounded in international law, this would produce an asymmetry between the ICSID tribunal’s ratione personae jurisdiction in relation to contractual and investment treaty claims. This cannot have been the intention of the drafters of the autonomous test of nationality in article 25. The second possibility, that the status of article 25 as a provision of an international treaty attracts the supplementary application of other international rules on the nationality of claims, is no more appealing. The experience of the Iran-US Claims Tribunal is informative in this respect.

In the Dual Nationality Case25 Iran had contended that arbitrations before the Tribunal were an instance of diplomatic protection so that a solution to the admissibility of claims by dual nationals ‘must be found in public international law and not disputes between one State and nationals of the other, which could be resolved by the application of private international law’.26 The Tribunal rejected this contention because the object and purpose of the Algiers Accords was not to ‘extend diplomatic protection in the normal sense’.27 The rules of customary international law did not, therefore, prevent the Tribunal from exercising jurisdiction ratione personae over United States nationals that simultaneously held Iranian citizenship.28

ICSID tribunals have often been sensitive to the sui generis role of the nationality test in article 25 for the ICSID regime, which is not comparable to the object and purpose of (p. 823) the nationality of claims rule of diplomatic protection. In Československá Obchodni Banka, AS v Slovak Republic,29 an ICSID Tribunal was confronted with a jurisdictional challenge by the Respondent to the effect that the Claimant was no longer the real party in interest because it had assigned the beneficial interest of its claims to its national State, the Czech Republic, after the arbitral proceedings had commenced.30 The Tribunal did not rely upon the rule of customary international law that an alien must have beneficial ownership over the contractual claim that provides the factual basis of a diplomatic protection claim by its national State.31 Instead it held that:

absence of beneficial ownership by a claimant in a claim or the transfer of the economic risk in the outcome of a dispute should not and has not been deemed to affect that standing of a claimant in an ICSID proceeding, regardless whether or not the beneficial owner is a State Party or a private party.32

It is informative to consider the approach taken by ICSID tribunals to the nationality requirement for individuals and juridical persons set out in article 25(2)(a) and (b) respectively. Have tribunals interpreted the nationality requirements in the ICSID Convention against the background of the principles underpinning the nationality of claims rules in diplomatic protection? Or have tribunals proceeded on the basis that the ICSID nationality requirements serve an autonomous function in a Convention that confers rights directly upon non-State actors? The analysis that follows reveals a large measure of contradiction and uncertainty to the extent that, for the time being, no clear trend in favour of either approach can be discerned.

The diplomatic protection rule with respect to natural persons was propounded by the International Court of Justice in the Nottebohm case.33 The Court imposed a requirement for the admissibility of diplomatic protection claims that there must be an ‘effective’ or ‘genuine’ link between the individual who has suffered the injury and the national State espousing the claim and thereby rejected the conferral of nationality under municipal law as definitive for this purpose. The Court was concerned to ensure that only one State could have standing to bring a diplomatic protection claim on the basis that the individual ‘is in fact more closely connected with the population of the State conferring nationality than with that of any other State’.34 This is a primary concern of the nationality of claims rule in diplomatic protection because of the risk that individuals and corporate entities may shift their allegiances to powerful States to take up their grievances at the international level.35 Resort to such contrivances will inevitably cause friction between States.

In Soufraki v United Arab Emirates,36 the claimant, Soufraki, had lost his Italian nationality automatically in 1991 by operation of Italian law upon his acquisition of Canadian nationality. Thereafter, he could have reacquired his Italian nationality automatically either by making an application or taking up residence in Italy for one year. He maintained that he (p. 824) fulfilled the latter residency requirement37 and thus was an Italian national for the purposes of article 25(2)(a) and the corresponding definition of an investor in the Italy/UAE BIT. The Italian authorities confirmed by a series of certificates that they regarded him as an Italian national at the relevant times. The Tribunal did not import a ‘genuine link’ requirement from the law of diplomatic protection and then seek to identify his dominant or effective nationality (the UAE had claimed that his dominant nationality was Canadian).38 Instead, the Tribunal decided that the pronouncements of the Italian authorities could not be treated as dispositive of the question; in particular because there was no evidence that they had conducted an investigation into his assertion of residency before issuing the confirmation.39 The Tribunal in essence approached the issue as a question of fact rather than a question of law by undertaking its own investigation of the evidence pertaining to Soufraki’s alleged residence in Italy for a year following his acquisition of Canadian nationality.40 If the issue had been approached as a question of law, then the nationality test in article 25(2)(a) and the BIT would have had to be resolved by renvoi to Italian law. As the Italian certificates and Soufraki’s possession of a valid Italian passport confirmed his status in the Italian legal order, these instruments would have had to be treated as dispositive of the question of law, unless they had been procured by fraud, which was not addressed by the Tribunal.

It was not open to the Tribunal to treat the nationality requirement in article 25(2)(a) and the BIT as a question of fact and thereby to consider Soufraki’s status in the Italian legal order as an evidential matter that could be suppressed by reference to other evidence. For the Tribunal to decline jurisdiction in spite of Soufraki’s status in the Italian legal order, it had to rule that the nationality requirement in article 25(2)(a) and the BIT is an autonomous test so that factors in addition to Soufraki’s possession of Italian nationality should be taken into account. This was the approach taken by the International Court in the Nottebohn case when it imported a ‘genuine link’ requirement into the nationality of claims rule in diplomatic protection. In other words, it was not permissible for the tribunal to question Soufraki’s status in the Italian legal order, but it was open to the tribunal to decide whether his Italian nationality could be invoked against the UAE for the purposes of an international claim under the ICSID Convention. But although this route was open, it would not have been a welcome precedent either. There is little evidence that the drafters of the ICSID Convention had envisaged that the nationality requirements for natural or juridical persons should be interpreted against the background of diplomatic protection rules or any other notion of ‘dominant’ or ‘effective’ nationality. As will be discussed with respect to the nationality of juridical persons, there are compelling reasons not to investigate the quality of the factual connections between the corporate entity and the Contracting State of incorporation. Indeed, one of the disquieting ramifications of the decision in Soufraki is that it widens the gulf between the position of a natural person and a juridical person vis-à-vis the ICSID arbitration mechanism. The Tribunal was alive to this problem but was content to articulate it rather than confront it:

[The Tribunal] appreciates that, had Mr Soufraki contracted with the United Arab Emirates through a corporate vehicle incorporated in Italy, rather than contracting in his personal capacity, no problem of jurisdiction would now arise.41

(p. 825) Annulment proceedings have upheld this award.42 In particular, the Annulment Committee in Soufraki addressed the issue of whether the ICSID tribunal ‘could make an independent determination of the nationality of the Claimant or whether it was bound by the determination made by the Italian municipal and consular authorities through the different documents, such as passports and certificates of nationality, issued to the Claimant’.43 The ad hoc Committee first distinguished the act of granting nationality at domestic level from the act of recognizing nationality on an international level. The Committee concluded that:

Summarizing, the Tribunal had the power to determine whether it had jurisdiction to hear the dispute. In determining whether the jurisdictional requirements of the ICSID Convention and the BIT have been satisfied, the Tribunal is empowered to make its own investigation into the nationality of parties regardless of the presence of official government nationality documents. Certificates of nationality constitute prima facie—not conclusive—evidence, and are subject to rebuttal. In fine, the Tribunal did not manifestly exceed its powers in deciding that it had to determine for itself Mr. Soufraki’s nationality.44

Turning now to juridical persons, the International Court in Barcelona Traction did not follow Nottebohm by adopting the same ‘genuine link’ criterion for identifying the national State eligible to bring a diplomatic protection claim on behalf of the corporation. But the Court achieved the same objective of channelling the interests in the subject matter of the claim into a single rubric of nationality by determining the nationality of the claim as belonging to the State where the aggrieved corporate entity is incorporated and rejecting the separate jus standi of the State whose nationals comprise the majority of shareholders.45 Although declining to follow Nottebohm, the Court emphasised that a ‘close and permanent connection’ had been established between the company Barcelona Traction and Canada as the State of incorporation insofar as Barcelona Traction had its registered office there with its accounts, share registers and listing with the Canadian tax authorities and board meetings had been held in Canada for many years.46 (Indeed, Brownlie’s review of Barcelona Traction and State practice generally led him to the conclusion that ‘a doctrine of real or genuine link had been adopted, and, as a matter of principle, the considerations advanced in connection with the Nottebohm case apply to corporations’.47) In this way, the Court was able to identify both a single nationality for the claim and a real connection between the national State and the claim. The result was that Canada, as the State of incorporation, would have had standing against Spain for the latter’s alleged expropriatory acts vis-à-vis the company; however, a claim by Belgium, whose nationals owned 88 per cent of the shares in Barcelona Traction, was held to be inadmissible.

The ICSID Convention does not identify a threshold link between the Contracting State and its national beyond the connection of nationality. The question is whether the conception of nationality for juridical persons employed by article 25(2)(b) should be understood to encompass the same objectives as for the nationality of claims rule in diplomatic (p. 826) protection. Is article 25(2)(b) concerned with the identification of a single nationality for the claim and a real connection between the Contracting State and the investor’s claim? This is the question that divided the Tribunal in Tokios Tokelės v Ukraine.48 The majority’s reasoning can be summarized as follows:

  1. (1)  Article 25(2)(b) does not define corporate nationality but instead the outer limits within which disputes may be submitted to ICSID arbitration. The Contracting States are permitted to define corporate nationality in the instrument recording consent to ICSID arbitration subject only to those outer limits.49

  2. (2)  The Ukraine/Lithuania BIT defines corporate nationality as ‘any entity established in the territory of the Republic of Lithuania in conformity with its laws and regulations’. The investor company, Tokios Tokelės, was a lawfully registered company in Lithuania. That is dispositive in satisfying the definition of corporate nationality in the BIT and such definition is within the outer limits of the requirements in article 25(2)(b) of the ICSID Convention.50

  3. (3)  The fact that 99 per cent of the shares of Tokios Tokelės were owned by Ukrainian nationals (ie nationals of the respondent Contracting State) is irrelevant to the definition of corporate nationality in the BIT (being focused exclusively on the municipal legal act of incorporation) and does not take the jurisdiction of the Tribunal outside the limits of article 25(2)(b) of the ICSID Convention.51

The dissenting judgment of President Weil made the following points:

  1. (1)  Whilst article 25(2)(b) does not define corporate nationality, it does not leave this matter to the discretion of the Contracting States either. Any definition agreed upon by the Contracting States in bilateral instruments cannot offend the object and purpose of the ICSID Convention.52

  2. (2)  The object and purpose of the ICSID Convention is to regulate the settlement of international investment disputes. Article 25(2)(b) cannot be interpreted to allow nationals of a Contracting State to invoke an international dispute resolution mechanism against that State through the subterfuge of a company incorporated in another Contracting State, thereby evading the jurisdiction of municipal courts and tribunals.53

  3. (3)  The fact that 99 per cent of the shares of Tokios Tokelės were owned by Ukrainian nationals is therefore a relevant factor for the requirement of nationality in article 25(2)(b)—a criterion of public international law which is concerned with the economic reality of the investment structure rather than treating the municipal legal acts of the investor as conclusive.54

Which of these approaches to the conception of nationality for juridical entities in article 25(2)(b) is correct? As the dissenter noted, much turns upon the perception of the ‘philosophy’ underlying the ICSID Convention.55 The interest of the Contracting States (p. 827) in maintaining the jurisdiction of their courts and tribunals over their own nationals was repeatedly emphasized by the dissenter.56 In contrast, the majority purported to do no more than faithfully adopt the ordinary meaning of the applicable texts.57 Although not articulated by the majority, perhaps the principle favouring the ‘autonomous’ approach is the affirmation of a broad discretion vested in Contracting States to define nationality and extend the benefits of the Convention to the greatest number of potential investors, thereby increasing the likelihood that the ICSID Convention will achieve its objective of stimulating the inflow of capital. The majority did say:

The investment would not have occurred but for the decision by the Claimant to establish an enterprise in Ukraine and to dedicate to this enterprise financial resources under the Claimant’s control. In doing so, the Claimant caused the expenditure of money and effort from which it expected a return or profit in Ukraine.58

There is no doubt that the majority’s approach in Tokios signals a departure from a diplomatic protection rationalization of the ICSID dispute resolution mechanism and permits a significant dilution in the bond between the Contracting State and the investor claiming its nationality. Indeed the ease with which the formal requirement of incorporation can be discharged has led to the growing practice of establishing investment vehicles in a jurisdiction which is ‘covered’ by an investment treaty with the host State of the investment. These investment vehicles may be corporate shells in a tax-friendly jurisdiction that are bound to transfer any commercial returns from the investment enterprise to the parent company in a different jurisdiction.

The outer limit of the flexibility endorsed by the majority in Tokios is likely to be breached in situations where the investor is found to have engaged in forum shopping by restructuring its investment in order to gain the requisite nationality for recourse to ICSID arbitration. The crucial question is the timing of the restructuring. If it is after the dispute has arisen then the claim is clearly inadmissible. But what if the investment is structured in the first place so that the investor can avail itself of the ICSID arbitration procedure in the event that a dispute arises in the future? The majority seemed to indicate that this might render the claim to be inadmissible as the structure would be for an ‘improper purpose’:

The Claimant manifestly did not create Tokios Tokelės for the purpose of gaining access to ICSID arbitration under the BIT against Ukraine, as the enterprise was founded six years before the BIT between Ukraine and Lithuania entered into force. Indeed, there is no evidence in the record that the Claimant used its formal legal nationality for any improper purpose.59

If the primary objective of the ICSID Convention is to ‘stimulat[e] a larger flow of private international capital into those countries which wish to attract it’,60 then an investor’s decision to structure its investment to attract the protection of the ICSID Convention is an (p. 828) example, par excellence, of the Convention achieving its objective. Suppose the Ukrainian investors behind Tokios Tokelės would not have invested their capital in Ukraine but for the availability of ICSID arbitration due to the level of sovereign risk in that country. Would their Lithuanian investment vehicle be denied recourse to ICSID arbitration because of their decision to structure their investment in Ukraine to meet the nationality requirements of article 25(2)(b) of the ICSID Convention? If the essence of the ICSID Convention is simply a procedural adjustment to the mechanism of diplomatic protection, allowing nationals to step into the shoes of their sovereign State, there would be grounds for insisting upon the compatibility of the ICSID nationality requirement with the objectives of the nationality of claims rule. The approach taken by the Tribunal in Loewen v United States of America with respect to the nationality requirement in Chapter 11 of NAFTA would be generally endorsed for the ICSID Convention as well. The Tribunal stated that:

It is that silence in the Treaty that requires the application of customary international law to resolve the question of the need for continuous national identity.61

If the ICSID Convention were to be interpreted against this background, then investors who have structured their investment to benefit from the ICSID Convention and thereby reduce sovereign risk to an acceptable level would probably be denied access to the ICSID dispute resolution mechanism insofar as the tribunal would be justified in examining the quality of the links between the corporate entity and the Contracting State whose nationality is relied upon. The interest of the Contracting States in avoiding international litigation for the ultimate benefit of its own citizens would also be paramount. But if the investor/State regime established by the ICSID Convention is autonomous in purpose and design to reduce sovereign risk for putative investors in developing economies, then it should be more resilient, if not impervious, to the influence of diplomatic protection. If this be the ‘philosophy’ of the ICSID Convention, then access to the ICSID dispute resolution mechanism would be open to an investor who structures its corporate affairs to ensure that its investment vehicle qualifies as a national of a Contracting State, whether or not, as an individual, the investor has the same nationality of the Contracting State that is host to the investment.

A broad interpretation of the nationality requirements does potentially cause difficulties with respect to diplomatic protection claims. If, for example, Latvia were to bring a claim against Ukraine for the failure to enforce an ICSID award rendered in favour of Tokios Tokelės, there would no doubt be objections to Latvia’s locus standi before an international tribunal as the real beneficiaries of the claim would be Ukrainian nationals. Latvia, however, as a party to the ICSID Convention has its own interest in Ukraine’s compliance with articles 53 and 54 of the Convention. Another problem would arise where the investor has more substantial links to a non-Contracting State than the Contracting State whose nationality has been invoked. If the host State obtains an award in its favour, then it might be compelled to enforce the award in the courts of the non-Contracting State (where the investor has its primary assets), which is not bound by the enforcement obligation in article 54(1). These difficulties have not yet arisen in practice, but their theoretical possibility is a factor that admittedly goes against the broad interpretation favoured by the majority in Tokios Tokelės v Ukraine.

(p. 829) (c)  Partial applicability of inter-State forms of reparation for injury

Article 34 of the ILC Articles specify restitution, compensation and satisfaction as the forms of reparation for an injury available under the general law of State responsibility. An investor is unlikely to petition an ICSID tribunal for satisfaction from the host State in the form of ‘an expression of regret’ or be any more tempted by the other modalities for satisfaction listed in article 37(2) of the ILC Articles. That satisfaction appears to be so foreign to the remedial priorities of an investor does, nonetheless, provide an important insight relevant to the question of whether the other forms of reparation are appropriate. In truly international cases, the declaratory judgment is the most frequently requested remedy for the reasons articulated by Judge Hudson in Diversion of Water from the Meuse:62

In international jurisprudence, however, sanctions are of a different nature and they play a different role, with the result that a declaratory judgment will frequently have the same compulsive force as a mandatory judgment; States are disposed to respect the one not less than the other.

Unlike diplomatic protection in customary international law, ICSID proceedings are concerned with the vindication of private interests and the principal advantage of ICSID arbitration for investors is that the fate of their claims is not dependent upon the vicissitudes of the diplomatic relationship between States. The corollary of this essential feature of the ICSID investor/State regime is that forms of reparation that have evolved in inter-State cases cannot be assumed to be part of the remedial arsenal of ICSID tribunals.

The most important question in this context is whether an ICSID tribunal is competent to grant restitution. (Restitution should not be confused with specific performance, the latter being confined to the enforcement of contractual obligations. There does not appear to be a single instance of an international tribunal ordering specific performance of a treaty obligation.63) The preferable view is that an ICSID tribunal is not competent to order restitution. First, even in general international law, the status of this remedy is very doubtful. The ubiquitous references to Chorzów Factory in ICSID awards do not acknowledge the existence of a specific provision for restitution in the treaty conferring jurisdiction upon the Permanent Court of International Justice in that case, nor the fact that restitution was not actually claimed by Germany. The statement about the primacy of restitution as a remedy for an international wrong was strictly obiter and the validity of this statement is certainly not confirmed by the paucity of instances when restitution has been awarded by international tribunals. Secondly, there are acute difficulties with such a remedy that an ICSID tribunal is ill-equipped to resolve. Juridical restitution requires specific legislative, executive or judicial acts on the part of the host State to restore the antecedent legal position of the investor under its municipal law where such acts might contravene constitutional norms or affect the rights of third parties. Material restitution is also problematic due to the limited ability of ad hoc tribunals to supervise and enforce transfers of property between the parties.

Consistent with the observation about the difficulty associated with the award of restitution in the ICSID regime is the fact that the Contracting States are only obliged to enforce the pecuniary obligations arising out of an ICSID award (article 54(1)). This (p. 830) implies that non-pecuniary obligations are not enforceable.64 In some Contracting States this principle features in local legislation.65

The discussion of non-pecuniary remedies in Enron v Argentina66 obscures rather than clarifies the issue. Several Argentine provinces had made tax assessments on various operations of Enron’s investment in Argentina. Enron considered that the tax assessments were unlawful and expropriatory, but payment of the taxes had been suspended by the Argentine Supreme Court to await a final judicial decision. Enron requested that the Tribunal declare the taxes assessed to be expropriatory (and therefore in breach of the US/Argentina BIT) and ‘that they be annulled and their collection permanently enjoined.’67 The tribunal upheld its jurisdiction to award ‘specific performance and an injunction’ should it determine on the merits that Enron’s request was justified.68 Turning first to the remedy of specific performance, it is impossible to fathom which obligation the tribunal (or Enron) had in contemplation. The assessment of taxes, if found to be contrary to an international obligation on the merits, could be declared by the tribunal to be a nullity. That is not specifi c performance of the international obligation; that is rather the secondary consequence of a breach of a primary obligation of international law. The usefulness of such a declaration would be limited because there would be no sanction in the event that Argentina did not adopt the relevant executive or legislative acts to annul the tax assessments, assuming that there would be no constitutional impediment for the relevant State organ to do so. But if the tax assessments were raised by Argentina in a counterclaim, then a declaration of nullity would mean that a damages award in favour of Enron (on the basis of a different primary claim) would not be set off against that amount. Alternatively, a declaration of nullity might be useful to Enron in resisting the collection of the taxes in the courts of a different jurisdiction, although most municipal legal systems would refuse to enforce a claim for foreign revenues in any case under domestic conflict of laws rules. So much for a declaration of nullity. Next, if Enron had in fact paid over the amounts due under the tax assessment, then it might claim restitution. That would be a form of juridical restitution, requiring Argentina to adopt specific acts to restore the antecedent status quo (again by annulling the tax assessments) and returning the sums paid by Enron. If this is what the tribunal meant by ‘specific performance’, then (laxity in relation to the correct terminology aside) the justificatory reasoning is sparse indeed. First, no mention is made of article 54(1), which requires Contracting States only to enforce pecuniary obligations arising out of an award. As previously stated, this can only mean that non-pecuniary obligations awarded by ICSID tribunals are not enforceable. Second, if Enron does ultimately pay the taxes assessed by the Argentine provinces, and such tax assessments are later adjudged to be internationally unlawful by the tribunal on the merits, then what interest does Enron have in claiming restitution of the antecedent status quo when it can simply claim damages for breach of an international obligation? Finally, with respect to the requested ‘injunction’ to ‘permanently enjoin’ Argentina from enforcing the tax assessments, there is again (p. 831) no reference to article 54(1) or mention of the futility of such a remedy in the tribunal’s decision. Given that an ICSID tribunal only has the power to recommend temporary injunctions under article 47, then, regardless of article 54(1), it would be surprising if it had the authority to order a permanent injunction in the final award. Again, no reference is made to article 47 or Rule 39 dealing with provisional measures in the tribunal’s decision.

In relation to the remedy of compensation, some international authorities suggest that a punitive element might feature in the assessment of damages in recognition of the detriment caused to the relationship between the State litigants. This might well explain the notorious distinction between lawful and unlawful takings in Chorzów Factory, which appears to rest on the premise that unlawful takings attract a higher level of reparation on the basis that an international delict has been committed. In Amco v Indonesia,69 the investor suffered a denial of justice before an administrative body of the host State. But for that denial of justice, however, the substantive decision of the administrative body would not have been different because there was a legitimate and compelling basis for the decision to revoke the investor’s licence despite the violation of the investor’s procedural rights. The ICSID Tribunal nevertheless awarded compensation to the investor for the detriment produced by the revocation; the implicit principle being that any breach of an international obligation demands compensation regardless of a causal link between the violation and the damage.70 If the investor/State arbitration mechanism created by the ICSID Convention is really directed to the vindication of private rights, then the award of damages beyond the realm of compensation for actual loss caused to the investor is beyond the jurisdiction of the tribunal. Unless the decision in Amco represents a camouflaged award of punitive damages in the absence of a causal link to a compensable loss suffered by the investor, which is barely conceivable, then the damages award can only be justified by reference to distinct prejudice caused to Amco’s national State; perhaps rationalized as the general concern of a Contracting State that the minimum standards of treatment of aliens found in customary international law are upheld with respect to its own nationals. An investor engaged in a singular battle for compensation with respect to private economic activities is an unlikely and inappropriate champion of this wider public interest, especially given the often tenuous connection between the investor and its national Contracting State. To take a hypothetical example, it would be untenable for an investment treaty tribunal to increase the amount of damages to account for the fact that the host State had breached its obligations under a BIT on several occasions in relation to different investors of the same nationality. Damages in an investment treaty claim are assessed purely on the basis of the harm caused to the economic interests of the investor by the host State, without regard for any factors in the relationship between the host State and the national State of the investor.

This distinction between the distinct forms of reparation owed to the investor and the national State is actually endorsed in a much overlooked passage in the Permanent Court’s decision in Factory at Chorzów:

The reparation due by one State to another does not however change its character by reason of the fact that it takes the form of an indemnity for the calculation of which the damage suffered by a private person is taken as the measure. The rules of law governing the reparation are the rules of international law in force between the two States concerned, and not the law governing relations (p. 832) between the State which has committed a wrongful act and the individual who has suffered the damage. Rights or interests of an individual the violation of which rights causes damage are always in a different plane to rights belonging to a State, which rights may also be infringed by the same act. The damage suffered by an individual is never therefore identical in kind with that which will be suffered by a State; it can only afford a convenient scale for the calculation of the reparation due to the State.71

This passage highlights that there is a substantive difference between the reparation for wrongs done to individuals and to States and thus compels a measure of caution in approaching the Court’s classic statement on restitution as the primary remedy in international law and the measure of damages in lieu in the context of the investor/State regime.

(d)  The law applicable to the substance of the investment dispute

The investor/State regime of responsibility created by the ICSID Convention implicates a plethora of legal relationships that gives rise to a diverse range of applicable laws. The investor is often a corporate entity established under a municipal law of one Contracting State, whereas its investment is generally a bundle of rights acquired pursuant to the municipal law of a different Contracting State. The public law or international regulatory obligations of the Contracting State that is host to the investment may have an impact upon this bundle of private law rights, and the public acts of the Contracting State might attract its international responsibility upon a breach of the minimum standards of treatment in investment treaties or customary international law. The investor/State regime thus summons the metaphorical image of a kaleidoscope of applicable laws, unlike the State/State regime where public international law is destined to play an exclusive role, and questions of municipal law are treated as questions of fact.72

The approach to the choice of law for investment disputes submitted to ICSID has given rise to a great deal of controversy, not in the least due to textual ambiguities in article 42(1).73 The problem can be resolved by returning to first principles relating to choice of law in general and adapting such principles to the sui generis nature of the investment treaty regime. The basic rule must be that an ICSID tribunal should decide a dispute in accordance with the proper laws of the issues raised by the dispute. If the investor’s claim rests upon a particular interpretation of a clause in an investment agreement, then the proper law of this issue is the law governing the contract. An express choice of law clause in the investment agreement will naturally be dispositive of the proper law in this instance. If the host State defends the claim by reference to an environmental regulation that purports to modify the investor’s contractual rights then the tribunal will interpret the regulation in accordance with its law and assess its impact on the contractual rights of the parties by reference to the law governing the contract. If the investor anticipates this defence with an argument to the effect that the regulation is incompatible with international law and thus should be considered a nullity, the tribunal will apply international law to this issue. The proper laws of the issues raised by the dispute must in each case be determined by the ICSID tribunal through an objective analysis of their juridical foundation and by reference to appropriate connecting factors.

(p. 833) An example of the utility and necessity of characterizing the issue in dispute is the English Court of Appeal’s judgment in Macmillan Inc v Bishopsgate Investment Trust Plc (No 3).74 The claimant, Macmillan, was a publicly listed company in which Robert Maxwell and his family had an interest. Macmillan in turn had a majority shareholding in the New York company Berlitz and these shares were registered in Macmillan’s name. Upon the instructions of Maxwell, Macmillan’s shares in Berlitz were transferred to Bishopsgate, a company owned and controlled by Maxwell, to be held as nominee for the account and for the benefit of Macmillan. The Berlitz shares were then fraudulently pledged to secure debts of companies privately owned by Maxwell and his family. After the collapse of the Maxwell empire, the shares were held as security by three banks, which were co-defendants in the case. Macmillan claimed that the banks had been unjustly enriched by receipt of the shares as security in breach of the trust relationship between Macmillan and Bishopsgate. The banks defended the claim by asserting that they were bona fide purchasers for value without notice of the breach of trust. (If this were the case, then the banks’ title to the shares would defeat Macmillan’s claim in unjust enrichment.) Macmillan argued that insofar as its claim was in unjust enrichment, any defence raised by the defendant banks should be governed by the law applicable to that claim. That law would be English law. The Court of Appeal rejected this approach. It was not the claim that required characterization, but the particular issue concerning the banks’ defence; namely, whether they had priority of title over the interest asserted by Macmillan. This issue related to property: had the banks acquired good title over the Berlitz shares? Berlitz was a New York company and thus the issue of title to shares in Berlitz was governed by the law of the place of its incorporation—the law of New York. According to this law, the defendant banks had acquired good title to the Berlitz shares.

The ‘proper law of the issue’ approach to the applicable law in ICSID arbitrations is preferable to the inference of a single choice of law from the nature of the legal instrument which embodies the consent of the parties to ICSID jurisdiction. According to this approach, if consent is recorded in an ICSID arbitration clause in an investment agreement, then any dispute submitted to ICSID arbitration on the basis of this arbitration clause will be governed by the proper law of the investment agreement in its entirety. But there is no reason in principle to adhere to such an inflexible choice of law rule. For instance, consider an example where that the parties have selected the UNIDROIT Principles of International Commercial Contracts as the rules of law to govern their investment agreement pursuant to article 42(1) and a dispute arises about a clause in that agreement exempting the foreign investor from liability to pay VAT. The interpretation of the text of that clause is governed by the proper law of the investment agreement, viz the UNIDROIT Principles. But the State might raise a defence based upon the application of VAT legislation in force at the time the investment agreement was concluded. The proper law of the taxation issue is clearly the municipal law of the host State and it is inconceivable that the UNIDROIT Principles could apply to this issue. Likewise, the investor might rely upon a double taxation treaty to bolster its claim to the VAT exemption. In this case it will be international law, and most certainly not the UNIDROIT Principles, that determines whether the State parties intended to confer rights directly upon non-State actors by concluding a double tax treaty and whether that treaty has the effect of exempting the investor from VAT liability in the host State.

(p. 834) The same situation arises when the consent to ICSID arbitration is embodied in an investment treaty. The investor might claim a breach of the national treatment obligation with respect to the host State’s refusal to accord the foreign investor a VAT exemption where it has done so for all the national investors in the same industry. That issue is governed by international law. But if the State defends by relying upon the contractual bargain, which accorded the investor other benefits on the understanding that it would be liable for VAT, then it is the proper law of the investment agreement that applies. The national treatment obligation under international law would not override the contractual treatment specifically negotiated by the investor with the host State in this context.

Consistent with this approach is the statement of principle from the very first decision of an ICSID tribunal with jurisdiction founded upon a bilateral investment treaty. In Asian Agricultural Products Ltd v Sri Lanka, the complex nature of the choice of law approach to investment disputes was identified with great insight:

… the Bilateral Investment Treaty is not a self-contained closed legal system limited to provide for substantive material rules of direct applicability, but it has to be envisaged within a wider juridical context in which rules from other sources are integrated through implied incorporation methods, or by direct reference to certain supplementary rules, whether of international law character or of domestic law nature …75

Several tribunals have since recognized that disputes submitted to ICSID arbitration concerning investment treaty obligations give rise to issues governed by a diverse range of laws. Thus, for instance, in CMS Gas Transmission Company v Argentina, the Tribunal remarked that, with respect to choice of law in ICSID arbitrations:

a more pragmatic and less doctrinaire approach has emerged, allowing for the application of both domestic law and international law if the specific facts of the dispute so justifies. It is no longer the case of one prevailing over the other and excluding it altogether. Rather, both sources have a role to play.76

The next stage in the evolution of the choice of law methodology in investment disputes will be for tribunals to develop a set of choice of law rules to determine the proper law of different types of issues that arise. The basic framework for this methodology can be simply stated. First, threshold questions about the existence, scope or nature of an investment frequently arise in investment disputes within the jurisdictional, liability and quantum phases of the proceedings. The proper law of the rights or interests relating to property that comprise the investment is the law of the host State. Second, the proper law of the issue of whether the rights and interests qualify as an investment is the treaty itself. Third, the proper law of whether the conduct of the host State is violative of the minimum standards of treatment encapsulated in the investment treaty is the treaty and general principles of international law.77

The judicial development of an objective set of choice of law rules to determine the proper law of the contentious issues in an investment dispute will demand greater (p. 835) transparency from tribunals in their reasoning as to a particular choice of law in any given situation. This in turn will prevent injustices caused by an unprincipled selection of rules by tribunals from different sources of law to fit a particular preordained result in the particular case.

Reisman has identified one paradigm of injustice that could be avoided by the approach advocated here: the case where the tribunal searches in vain for a remedy in national law and so appeals to the more amorphous principles of international law to fashion a remedy regardless:

The question is whether or not the law of the host State addresses the issue at hand. If it does and, as part of its law, has decided not to grant remedies in such matters then there is no remedy, as none is provided in the law that must be applied … If an ICSID tribunal takes the claimant’s demand for a remedy as the framework of inquiry and assumes that if that remedy is not provided by the host State’s law, the Tribunal must then proceed to search for it in international law, the Tribunal will subvert the propose of the dispositive choice of law in Article 42(1) and create a new regime: national law is applied insofar as it provides a particular remedy, but if it does not, international law is then searched for the remedy.78

The present writer would ask a slightly different threshold question than the one posited by Reisman. Rather than determine ‘whether or not the law of the host State addresses the issue at hand’, the approach advocated here would question whether the law of the host State is the proper law of the issue at hand. If the investor advances its case on the basis of a contractual breach, then the proper law of the contract applies. If the proper law so determined provides no remedy, then, as Reisman intimates, it is impermissible for the tribunal to search for one in a different law. On the other hand, if the investor maintains that the host State abused its executive power to frustrate the performance of the contract and thereby violated an international treatment obligation in an applicable investment treaty, then the investor may be entitled to a remedy even where the proper law of the contract would not provide one, but only if the host State’s conduct is found to have breached its international treatment obligation. Such an approach does not subvert the purpose of article 42(1), which surely is concerned with ensuring that the law that objectively governs an issue is applied. Article 42(1) repels the fallacy of a singular governing law.

The impermissible approach to choice of law is aptly demonstrated in Wena v Egypt.79 Wena alleged in ICSID proceedings that Egypt breached several provisions of the UK/Egypt BIT when a State-owned company, the Egyptian Hotel Company (‘EHC’), seized two hotels (the ‘Luxor Hotel’ and the ‘Nile Hotel’) which were the subject of separate lease agreements between Wena and EHC. The lease agreements between Wena and EHC stipulated that disputes between the parties must be submitted to ad hoc arbitration in Cairo.80 Following the seizure, Wena had brought a contractual arbitration against EHC for breach of the Nile Hotel lease on 2 December 1993.81 Wena was awarded EGP 1.5 million in damages as compensation for the seizure of the Nile Hotel; however, this ad hoc tribunal simultaneously ordered that Wena surrender the hotel to EHC due to its own breaches of the lease agreement.82 Wena continued to operate the Nile hotel until 1995 when it was evicted pursuant to the tribunal’s decision. Wena brought similar contractual (p. 836) arbitration proceedings against EHC with respect to the Luxor Hotel lease on 12 January 1994. The second ad hoc tribunal also found in favour of Wena and awarded EGP 9.06 million in damages and also ordered Wena to surrender the hotel to EHC.83 The award was subsequently annulled by the Cairo Court of Appeal.84 Wena remained in occupancy until 1999, when the Luxor Hotel was placed in judicial receivership on account of Wena’s failure to pay rent.

Before the ICSID tribunal, Wena maintained that its investment in Egypt was the lease agreements for the two hotels. Egypt raised an objection to this submission on the basis that the leases had been terminated in accordance with their applicable law by tribunals established pursuant to the contractual dispute resolution mechanism. The proper law of this issue is undoubtedly Egyptian law. International law has nothing to say about whether Wena’s breaches of the agreements entailed the termination of the leases. It was open to Wena in the ICSID proceedings to assert a claim based upon a denial of justice with respect to the annulment of the second arbitral award by the Cairo Court of Appeal, but it failed to do so.

The ICSID tribunal completely sidestepped the issue of the validity of the leases, merely recording that:

[i]t is sufficient for this proceeding simply to acknowledge, as both parties agree, that there were serious disagreements between Wena and EHC about their respective obligations under the leases.85

That was the end of the analysis. The ad hoc Committee hearing the subsequent annulment application filed by Egypt endorsed this approach with the following reasoning:

The leases deal with questions that are by definition of a commercial nature. The [BIT] deals with questions that are essentially of a government nature, namely the standards of treatment accorded by the State to foreign investors.86

This rigid dichotomy between the subject matter of the leases and the investment treaty is disingenuous because the sine qua non of investment treaty protection is the investor’s attainment of private law rights which comprise an investment pursuant to the definition contained in the treaty. If there is no investment, there is no investment treaty protection. One can be left in no doubt about the ad hoc Committee’s erroneous endorsement of the tribunal’s approach to the relevance of the leases:

[T]he Tribunal declared irrelevant to consider the rights and obligations of the parties to the leases for the purpose of reaching a decision on the dispute submitted to it. The Award confirms that Wena has been expropriated and lost its investment, and this irrespective of the particular contractual relationship between Wena and EHC. The explanation thus given for not determining the respective obligations of Wena and EHC under the leases is sufficient to understand the premises on which the Tribunal’s decision is based in this respect.87

Quite simply, there can be no expropriation without something to expropriate. The tribunal was obliged to first apply the proper law of the lease contracts (Egyptian law) to determine whether they remained valid and binding. If they had been lawfully terminated in accordance with their proper law, and the procedure that led to such termination was (p. 837) unimpeachable from the perspective of international law, then that should have been the end of the matter. Subject to an affirmation of their continued validity on this basis, the tribunal was then required to determine whether Egypt’s conduct violated the minimum standards of treatment in the investment treaty in accordance with the proper law of this issue (international law).88

Before moving on to an analysis of the procedural law of ICSID arbitrations it is necessary to anticipate two criticisms that might be made of the choice of law methodology advocated here. It might first be argued that a dispute implicating an issue governed by international law is not within the ratione materiae jurisdiction of the ICSID tribunal when the legal instrument containing the consent of the parties to ICSID arbitration is an investment agreement. Such an argument would necessarily rely upon the words used in the arbitration clause, which in the standard form reads ‘any dispute arising out of or relating to this agreement for settlement by arbitration’. The use of the qualifiers ‘any’ and ‘relating to’ appears to cast the jurisdictional net wide enough to cover disputes that give rise to issues which are governed by laws different to the proper law of the investment agreement. There would be little disagreement that such wording would extend jurisdiction to an issue in tort, and there is no compelling reason to deny that issues of international law would be covered as well.

If semantic considerations were to have the draconian effect of preventing an ICSID tribunal from applying the proper law of certain issues arising in a dispute, the words chosen would have to be unequivocal indeed. Far from explicitly dictating such a result, the standard ICSID arbitration clause is formulated to cover any dispute relating to the investment agreement. If, for example, the foreign investor’s shares in a company established on the basis of the investment agreement are expropriated by the host State, then the investor’s cause of action, and the resulting issues governed by international law, is within the ratione materiae jurisdiction of the ICSID tribunal.

It might next be argued this approach to the choice of law for investment disputes submitted to ICSID is not consistent with the text of the applicable law provision in article 42(1):

The Tribunal shall decide a dispute in accordance with such rules of law as may be agreed by the parties. In the absence of such agreement, the Tribunal shall apply the law of the Contracting State party to the dispute (including its rules on the conflict of laws) and such rules of international law as may be applicable.

Let us first consider the situation where the parties have expressly chosen rules of law to govern their relationship in an investment agreement. Can the tribunal in that instance apply international law if it determines from the parties’ pleadings that, in order to dispose of a claim or counterclaim, it must rule upon an issue governed by international law? The answer must be that it can because a choice of law by the parties does not extend to matters beyond their contractual relationship. Just as municipal conflict of laws does not generally permit parties to select the law governing their conduct arising outside the contractual context (such as upon the commission of a tort),89 the autonomy of parties to an investment agreement with an ICSID arbitration clause is similarly constrained. (p. 838) Returning to the previous example, the choice of the UNIDROIT Principles to govern an investment contract does not have the effect of removing the investment activities contemplated by the contract from the regulatory system in place at the host State. If the host State justifies withholding sums due to the investor under the investment contract on the basis of the tax legislation in force, then, assuming this issue is not specifically dealt with by the contract, the tribunal cannot rule upon this issue by reference to the UNIDROIT Principles. The issue cannot be characterized as a contractual issue and is thus outside the scope of the parties’ choice of rules of law under the first sentence of article 42(1).

The law chosen to govern a contract must be distinguished from the laws applicable to the dispute in this context. The law chosen to govern a contract will apply to issues concerning the interpretation and performance of the contract, the consequences of its breach and the assessment of damages. But it is generally accepted that it does not necessarily govern issues relating to the capacity of the parties, formal validity or the mode of performance.90 Similarly, in most legal systems, whether a contractual stipulation about tortious liability will be an effective defence to a tort claim is governed by the lex loci delicti and not the law chosen by the parties to the contract.

Turning next to the default rule on the applicable law in article 42(1), it should be obvious that this provision does not provide any guidance as to when national law or international law should be applied by the tribunal. So much ink has been spilt on the import of the conjunction ‘and’ that appears between the references to the law of the host State and to the rules of international law. But the search for definitive guidance from the use of a single conjunction is surely in vain. The default rule does not purport to set out the connecting factors that would enable the tribunal to decide the proper law of a particular issue. Article 42(1) is not, therefore, a choice of law rule in the true sense of the term. It simply recognizes the competence of the tribunal to apply both national and international law. It is for ICSID tribunals to develop a coherent set of principles to guide the choice of either of these laws with respect to particular types of issues.

These limitations of article 42(1) are implicit in the Report of the Executive Directors on the ICSID Convention, which simply notes that failing a choice of law by the parties:

the Tribunal must apply the law of the State party to the dispute (unless that law calls for the application of some other law), as well as such rules of international law as may be applicable.91

The Executive Directors thus make no attempt to define the parameters of the tribunal’s competence to apply these sources of law.

It is true that the original wording of article 42(1) was even more unequivocal as a statement of the competence of the tribunal to apply diverse sources of law rather than a choice of law rule. The preliminary draft of article 42(1) read:

In the absence of any agreement between the parties concerning the law to be applied … the Arbitral Tribunal shall decide the dispute submitted to it in accordance with such rules of law, whether national or international, as it shall determine to be applicable.92

(p. 839) It is certainly true that the capital-importing States voiced concern about the possibility that ICSID tribunals might resort to ignoring domestic rules and regulations wholesale if such a broad discretion with respect to the choice of law were to be conferred by article 42(1). The revised and enacted text of article 42(1) was designed to allay this concern, but it does not transform the article into a true choice of law rule.

The early ICSID cases interpreting the default rule in article 42(1) emphasized a ‘complementary’ and ‘corrective’ function of international law vis-à-vis the national law of the host State.93 The ‘complementary’ function was said to allow an ICSID tribunal to resort to international law in the case of lacunae in the applicable national law. This role for international law must be rejected outright. Only adherence to an extreme form of positivism would permit the possibility of a finding of non liquet within a functional legal system. National judges are frequently confronted with situations where there are no specific rules from the corpus of positive law that address the particular contentious issue. In such cases, judges must arrive at a solution that best fits the existing body of decisions (legal enactments and case law) and is consistent with the fundamental principles of the legal system. The position is no different with respect to international law. The possibility of a finding of non liquet in relation to a concrete dispute arising under international law has been discredited by international scholars and tribunals for many decades.94

The purported ‘corrective’ function of international law under article 42(1) has been taken up by several ICSID tribunals.95 This explanation of the role performed by international law in accordance with the default position in article 42(1) must be treated with a measure of caution.

If the host State obtains a judicial annulment of the investment agreement in its own courts through improper means, then the investor might seek to claim damages for breach of contract and at the same time request a declaration to the effect that the court judgment is a nullity because it constitutes a denial of justice in customary international law. There might, in investment treaty cases, be reasons for the investor to prefer to sue on the investment agreement rather than the investment treaty obligations (assuming the tribunal has jurisdiction over contractual claims), such as where the contractual sum of damages would be higher than a damages award based on a breach of the international obligation due to a contractual stipulation on damages which exceeds the compensatory principles under international law. In such a case, a declaration of nullity would be an appropriate remedy and perhaps international law might be said to perform a ‘corrective’ function. But this would not be a precise description of the role performed by international law because, if the proper law of the issue approach is accepted, international law would not be ‘corrective’ with respect to an issue that should objectively be determined by a municipal law. There is no overlap in the field of application of the different sources of law because international law does not purport to regulate the particular issue in question.

(p. 840) (e)  The law applicable to the tribunal’s procedure

Just as there is no single law that governs all the substantive issues arising out of disputes submitted to ICSID arbitration within the investor/State regime, there is no single law universally applicable to the procedural issues arising out of these disputes either. The search for a tidy pigeon hole in this respect is futile and one is reminded of the vast literature on the law governing arbitrations under the aegis of the Iran-US Claims Tribunal, which even in the twilight years of that institution has yet to produce a sustained consensus.96 Rather than seeking to identify a single applicable law to ICSID procedure, it is appropriate to distinguish between various elements of the procedure which attract different applicable laws.

The ICSID Convention, in conjunction with the ICSID Arbitration Rules, seeks as far as possible to provide a comprehensive set of rules to govern the procedure of ICSID arbitrations. Among the procedural issues addressed by the Convention and Arbitration Rules are included matters relating to the constitution of the tribunal (articles 37–40, 56–58; Rules 1–12); matters relating to the conduct of the written and oral phases of the procedure (Rules 13–27, 29–32, 38) and the place of the proceedings (articles 62–3; Rule 13); the power to decide jurisdictional questions (article 41; Rule 41); evidentiary matters (article 43; Rules 33–37); the failure of a party to appear or present its case (article 45; Rule 42); the power to decide incidental claims or counterclaims (article 46; Rule 40); provisional measures (article 47; Rule 39); the procedure for rendering an award and for its interpretation, revision and annulment (articles 48–52; Rules 46–53); and costs (articles 59–61; Rule 28). Article 44 also confers upon the Tribunal the important power to decide procedural matters with respect to which the Convention and the Arbitration Rules are silent. Contrary, however, to a widespread conception of the ICSID regime, it is neither completely ‘selfcontained’, nor ‘autonomous’. The following examples demonstrate this point.

First, the parties to an ICSID arbitration can apply to municipal courts and other authorities for provisional measures for the preservation of their rights and interests either before the institution of ICSID proceedings or thereafter. It is a matter of debate as to whether the parties must consent to such in the arbitration agreement, given the uncertainty as to whether the amendment to Rule 39 of the ICSID Arbitration Rules by the ICSID Administrative Council (by the insertion of a new paragraph 5 making resort to municipal courts for this purpose conditional upon the consent of the parties) was a ‘clarification’ of article 26 (providing for the exclusivity of ICSID arbitration vis-à-vis other remedies) or an attempt to modify its application, which would be ultra vires the Administrative Council. If consent is required, then it is likely to be found to be implicit in many of the investment treaty arbitrations submitted to ICSID insofar as investment treaties often contain a provision to the effect that the submission of an investment dispute is without prejudice to the parties’ rights to apply for injunctive relief before municipal courts. For instance, article 26(3) of the 2004 Model BIT for the United States of America provides that the investor:

may initiate or continue an action that seeks interim injunctive relief and does not involve the payment of monetary damages before a judicial or administrative tribunal of the respondent, provided that the action is brought for the sole purpose of preserving the claimant’s or the enterprise’s rights and interests during the pendency of the arbitration.

(p. 841) Any such application for injunctive relief will naturally be governed by the lex fori.

There may be formidable reasons for a party to ICSID proceedings to petition municipal courts for injunctive relief in support of those proceedings and it cannot be assumed that adherence to an expansive interpretation of article 26 so as to rule out this possibility is likely to promote the effectiveness of the ICSID system. Interim measures to prevent assets from being dissipated or evidence from being destroyed, or to compel the production of documents or the attendance of witnesses, or to restrain a party from pursuing parallel proceedings in a municipal court, might play a decisive role in achieving justice in the reference to ICSID arbitration. Moreover, ICSID tribunals themselves are virtually impotent in this respect, having merely the power to ‘recommend’ (rather than to ‘prescribe’) provisional measures.97 Such non-binding ‘recommendations’ have a poor track record of compliance98 and are not generally enforceable in municipal courts.

Second, the municipal rules for the enforcement and execution of final judgments apply to the enforcement and execution of ICSID awards in the territories of Contracting States.99 For example, in AIG Capital Partners v Republic of Kazakhstan,100 AIG and the joint venture company formed for its investment in Kazakhstan petitioned the English High Court to enforce an ICSID award rendered in their favour against assets in London held by third party custodians on behalf of the National Bank of Kazakhstan. The Claimants had registered the award as a judgment under section 1 of the Arbitration (International Investment Disputes) Act 1966 and sought a Third Party Debt and Charging Order under Part 72.2 of the English Civil Procedure Rules and the Charging Orders Act 1979 to enable the Claimants to recover their award debt directly from the custodians of the assets. The orders sought by the Claimants were denied because, inter alia, the assets of the National Bank of Kazakhstan were protected by sovereign immunity from execution pursuant to section 14(4) of the State Immunity Act 1978.

Third, the law on sovereign immunity from execution (whether found in international custom, treaty or municipal law) applies to the execution of ICSID awards in the territories of both Contracting States (article 55) and non-Contracting States. Again, in AIG Capital Partners v Republic of Kazakhstan,101 the execution of an ICSID award was refused by an English court due to a blanket immunity attaching to the ‘property of a State’s central bank’ pursuant to section 14(4) of the State Immunity Act 1978.

Fourth, in the territories of non-Contracting States, ICSID awards are likely to be enforced in accordance with the rules for the enforcement of foreign arbitral awards (such as, where applicable, those contained in the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards or in municipal enactments giving effect to this Convention).

Fifth, where a party has instituted parallel proceedings in a municipal court in breach of article 26, municipal rules for the granting of a stay of court proceedings apply. In Attorney-General v Mobil Oil NZ Ltd, the New Zealand High Court stayed proceedings brought by the New Zealand Government because there was a ‘relevant relationship or nexus’ between the issues raised in these court proceedings and the pending ICSID (p. 842) arbitration that had been commenced by Mobil.102 The Court exercised its power to stay in accordance with its discretion under a domestic statute (s 8 of the Arbitration (International Investment Disputes) Act 1979).103 In MINE v Guinea, the US Court of Appeals left upon the possibility that US courts could compel an ICSID arbitration upon a petition by one of the parties under the Federal Arbitration Act. The Court ruled that MINE was estopped from raising this argument because in earlier court proceedings it had represented that the particular arbitration clause referring to ICSID arbitration was incapable of specific performance and thus American Arbitration Association (AAA) arbitration should instead be compelled.104

Sixth, some Contracting States have, by their implementing legislation passed in accordance with article 69, reserved the possibility of subjecting an ICSID arbitration to certain procedural rules contained in their municipal laws.105 To the extent that such municipal procedural rules supplement rather than modify the ICSID Arbitration Rules, it is doubtful that the Contracting State could be in violation of the ICSID Convention.

Further reading

  • P Muchlinski, F Ortino, & C Schreuer (eds), The Oxford Handbook of International Investment Law (Oxford, OUP, 2008)
  • RD Bishop, J Crawford, & WM Reisman, Foreign Investment Disputes (The Hague, Kluwer Law International, 2005)
  • J Crawford, ‘Treaty and Contract in Investment Arbitration’ (2008) 24 Arbitration International 351
  • R Dolzer & C Schreuer, Principles of International Investment Law (Oxford, OUP, 2008)
  • Z Douglas, The International Law of Investment Claims (Cambridge, CUP, 2009)
  • Z Douglas, ‘The Hybrid Foundations of Investment Treaty Arbitration’ (2003) 74 BYIL 151
  • C McLachlan, L Shore, & M Weiniger, International Investment Arbitration: Substantive Principles (Oxford, OUP, 2007)
  • CH Schreuer, L Malintoppi, A Reinisch, & A Sinclair, The ICSID Convention (2nd edn, Cambridge, CUP, 2009)


∗  The author would like to thank Dr Monique Sasson for her helpful assistance in the preparation of this Chapter.

575 UNTS 160. See, in general, CH Schreuer, L Malintoppi, A Reinisch, & A Sinclair, The ICSID Convention (2nd edn, Cambridge, CUP, 2009).

Z Douglas, The International Law of Investment Claims (Cambridge, CUP, 2009), 6.

CH Schreuer, L Malintoppi, A Reinisch, & A Sinclair, The ICSID Convention (2nd edn, Cambridge, CUP, 2009), 58–70.

Ibid, 414–30.

Ibid, 1096–1114.

Ibid, 1115–1185.

Ibid, 1273–1275.

1923, PCIJ, Series A, No 1, 20.

See, among others, Maffezini v Spain (ICSID Case No ARB/97/7), Decision on Objections to Jurisdictions, 25 January 2000, 5 ICSID Rep 396; and Salini v Morocco (ICSID Case No ARB/00/4), Decision on Jurisdiction, 23 July 2001, 6 ICSID Reports 400.

10  See eg Duke Energy v Peru (ICSID Case No ARB/03/28), Decision on Jurisdiction, 1 February 2006; World Duty Free v Kenya (ICSID Case No ARB/00/7), Award, 4 October 2006; and Noble Energy and Machalapower Cia v Ecuador (ICSID Case No ARB/05/12), Decision on Jurisdiction, 5 March 2008.

11  See, for example, Tradex Hellas SA v Albania (ICSID Case No ARB/94/2), Decision on Jurisdiction, 24 December 1996, 5 ICSID Reports 43; and Inceysa Vallisoletana v Republic of El Salvador (ICSID Case No ARB/03/26), Award, 2 August 2006.

12  Z Douglas, The International Law of Investment Claims (Cambridge, CUP, 2009), 125.

13  Ibid, 96–106.

14  Noble Ventures Inc v Romania (ICSID Case No ARB/01/11), Award, 12 October 2005. In this case an umbrella clause was at issue, and the tribunal found that: ‘where the acts of a governmental agency are to be attributed to the State for the purpose of applying an umbrella clause … breaches of a contract into which the State has entered are capable of constituting a breach of international law by virtue of the breach of an umbrella clause’ (para 85). The tribunal held that the contracts were entered into by two instrumentalities on behalf of the State and that they were attributable to the State for the purpose of the umbrella clause (para 86).

15  See Z Douglas, ‘The Hybrid Foundations of Investment Treaty Arbitration’ (2003) 74 BYIL 151, 184–193.

16  See eg B Simma, ‘Self-Contained Regimes’ (1985) 16 Netherlands Ybk of Int L 111, and above, Chapter 13.

17  W Riphagen, ‘State Responsibility: New Theories of Obligation in Interstate Relations’, in R Macdonald & D Johnston (eds), The Structure and Process of International Law: Essays in Legal Philosophy, Doctrine and Theory (The Hague, Nijhoff, 1983), 593.

18  Commentary to art 55, para 3. See also Z Douglas, The International Law of Investment Claims (Cambridge, CUP, 2009), 97.

19  Commentary to art 33, para 4.

20  See also the Commentary to art 28, para 3.

21  See also Z Douglas, The International Law of Investment Claims (Cambridge, CUP, 2009), 96–98.

22  Countermeasures have been claimed by Mexico in three NAFTA disputes (before three separate NAFTA Tribunals) against alleged violations of NAFTA by the US, relating generally to access of Mexico’s surplus sugar produce to the US market. Three separate claims have been brought by US agricultural firms against Mexico, relating to the imposition of a 20% tax by Mexico on soft drink bottlers using the sweetener High Corn Fructose Syrup (HCFS). One of the Tribunals held that countermeasures could not be invoked for a claim under Chapter XI of NAFTA, on the basis that it conferred direct rights on investors and countermeasures taken in an inter-State dispute could not interfere with those rights: Corn Products International, Inc v United Mexican States (ICSID Case No ARB(AF)/04/01), Decision on Responsibility, 15 January 2008, paras 168–169. Another of the NAFTA tribunals held that the conditions for taking countermeasures were not met in the circumstances, although in principle countermeasures might apply to a Chapter XI dispute: Archer Daniels Midland Company and Tate & Lyle Ingredients Americas, Inc v United Mexican States (ICSID Case No ARB(AF)/04/05), Award of 21 November 2007, paras 161ff. The third NAFTA tribunal agreed with the Corn Products International tribunal: Cargill, Incorporated v United Mexican States (ICSID Case No ARB(AF)/05/2), decision of 18 September 2009, paras 420–430.

23  Commentary to art 44, para 1.

24  See eg Tokios Tokelės v Ukraine (ICSID Case No ARB/02/18), Decision on Jurisdiction, 29 April 2004, 11 ICSID Reports 313.

25  Islamic Republic of Iran and United States (Case A/18) (1984) 5 Iran-US CTR 251.

26  Memorial of the Islamic Republic of Iran in (Case A/18) (21 October 1983), 25–26.

27  Islamic Republic of Iran and United States (Case A/18) (1984) 5 Iran-US CTR 251, 261.

28  Accord: SEDCO v NIOC and Iran (1985) 9 Iran-US CTR 245, 256.

29  Československá Obchodni Banka, AS v Slovak Republic (ICSID Case No ARB/97/4), Decision on Objections to Jurisdiction, 24 May 1999, 5 ICSID Reports 330.

30  Ibid, 342 (para 28).

31  See eg American Security and Trust Company Claim (1958) 26 ILR 322.

32  5 ICSID Reports 330, 343 (para 32).

33  Nottebohm (Liechtenstein v Guatemala), ICJ Reports 1955, p 4.

34  Ibid, 23.

35  E Borchard, ‘The Protection of Citizens Abroad and Change of Original Nationality’ (1933–4) 43 Yale LJ 359, 377–380.

36  Soufraki v United Arab Emirates (ICSID Case No ARB/02/07), Award, 7 July 2004, 12 ICSID Reports 156.

37  Ibid, 162 (paras 26–27).

38  Ibid, 164 (paras 42–46).

39  Ibid, 167 (paras 66, 68).

40  Ibid, 166 (para 58).

41  Ibid, 169 (para 83).

42  Soufraki v United Arab Emirates (ICSID Case No ARB/02/07), Decision on Annulment, 5 June 2007. This decision was adopted by the majority of the Committee.

43  Ibid, para 18.

44  Ibid, para 76.

45  Ibid, para 70.

46  Ibid, para 71.

47  I Brownlie, Principles of Public International Law (6th edn, Oxford, OUP, 2003), 465; this passage is not in the 7th edn (2008), but the same basic position is taken: ibid, 484–486.

48  Tokios Tokelės v Ukraine (ICSID Case No ARB/02/18), Decision on Jurisdiction, 29 April 2004, 11 ICSID Reports 313.

49  Ibid, 319 (para 25), 322–323 (para 39).

50  Ibid, 320 (paras 28–9), 323 (para 40).

51  Ibid, 324 (para 46), 326 (para 52), 332 (para 77).

52  Dissenting Opinion, ibid, 345 (para 16), 346 (para 19), 349–350 (para 28).

53  Dissenting Opinion, ibid, 342 (paras 5, 8, 9), 346 (para 19), 347 (para 23).

54  Dissenting Opinion, ibid, 347–8 (paras 21, 23, 24).

55  Dissenting Opinion, ibid, 341 (para 1).

56  Dissenting Opinion, ibid, 342 (para 8), 350–351 (para 30). See further: A Broches, ‘The Convention on the Settlement of Investment Disputes Between States and Nationals of Other States’ (1972) 136 Recueil des cours 331, 355, ‘There is no reason to have these international procedures be a substitute, even on an optional basis, for domestic procedures for the settlement of disputes between a State and its own citizens’; and Z Douglas, The International Law of Investment Claims (Cambridge, CUP, 2009), 314–317.

57  Tokios Tokelės v Ukraine (ICSID Case No ARB/02/18), Decision on Jurisdiction, 29 April 2004, 11 ICSID Reports 313, 320 (paras 28–29), 332 (para 77).

58  Ibid, 332 (para 78).

59  Ibid, 327 (para 56).

60  Report of the Executive Directors of ICSID, 1 ICSID Reports 23, 25 (para 9).

61  Loewen v United States of America (ICSID Case No. ARB(AF)/98/3), Award, 26 June 2003, 7 ICSID Reports 421, 486 (para 226).

62  1937, PCIJ Reports, Series A/B, No 70, p 4, 79.

63  See C Gray, Judicial Remedies in International Law (Oxford, OUP, 1987), 16.

64  Accord: A Broches, ‘The Convention on the Settlement of Investment Disputes Between States and Nationals of Other States’ (1972) 136 Recueil des cours 331, 400.

65  See eg in the United States: s 3(a) of the Convention on the Settlement of Investment Disputes Act 1966, 22 USC § 1650a (1976), and the Statement of Intent of the US Department of State, (1966) 5 ILM 820, 824.

66  Enron v Argentina (ICSID Case No ARB/01/3), Decision on Jurisdiction, 14 January 2004, 11 ICSID Reports 273.

67  Ibid, 289 (para 77).

68  Ibid.

69  Amco v Indonesia (ICSID Case No ARB/81/1), Award, 5 June 1990, 1 ICSID Reports 569.

70  See J Paulsson, Denial of Justice in International Law (Cambridge, CUP, 2005), 218–225.

71  Factory at Chorzów, Merits, 1928, PCIJ Reports, Series A, No 17, p 4, 28 (emphasis added).

72  See Certain German Interests in Polish Upper Silesia, 1926, PCIJ Reports, Series A, No 7, p 4, 19.

73  CH Schreuer, L Malintoppi, A Reinisch, & A Sinclair, The ICSID Convention (2nd edn, Cambridge, CUP, 2009), 545–639.

74  [1996] 1 WLR 387.

75  Asian Agricultural Products Ltd v Sri Lanka (ICSID Case No ARB/87/3), Award, 27 June 1990, 4 ICSID Reports 250, 257 (para 21). See J Crawford, ‘Treaty and Contract in Investment Arbitration’ (2008) 24 Arbitration International 351.

76  CMS Gas Transmission Company v Argentina (ICSID Case No ARB/01/8), Award, 12 May 2005, 14 ICSID Reports 152, 176 (para 116).

77  See Z Douglas, The International Law of Investment Claims (Cambridge, CUP, 2009), ch 2.

78  WM Reisman, ‘The Regime for Lacunae in the ICSID Choice of Law Provision and the Question of its Threshold’ (2000) 15 ICSID Rev—FILJ 362, 371.

79  Wena v Egypt (ICSID Case No ARB/98/4), Award, 8 December 2000, 6 ICSID Reports 89.

80  Ibid, 94 (para 17).

81  Ibid, 106 (para 60).

82  Ibid, 106–107 (para 61).

83  Ibid, 107 (para 62).

84  Ibid.

85  Ibid, 94 (para 19).

86  Wena v Egypt (ICSID Case No ARB/98/4), Decision on Annulment, 5 February 2002, 6 ICSID Reports 129, 136 (para 31).

87  Ibid, 147 (para 86) (emphasis added).

88  Further examples of a similar erroneous approach to choice of law in investment treaty disputes are provided in: Z Douglas, ‘The Hybrid Foundations of Investment Treaty Arbitration’ (2003) 74 BYIL 151, 202–205, 207–211; Z Douglas, ‘Nothing if Not Critical for Investment Treaty Arbitration’ (2006) 22 Arbitration International 27.

89  See P North, ‘Choice in Choice of Law’ in Essays in Private International Law (Oxford, OUP, 1993), 171.

90  Eg Rome Convention on the Law Applicable to Contractual Obligations, 19 June 1980, 1605 UNTS 59, arts 2(a), 9, 10.2.

91  1 ICSID Reports 23, 31 (para 40) (emphasis added).

92  Working Paper in the Form of a Draft Convention (5 June 1962) in Convention on the Settlement of Investment Disputes between States and Nationals of other States: Documents Concerning the Origin and the Formulation of the Convention, Vol 2, 19, 21.

93  Klöckner v Cameroon (ICSID Case No ARB/81/2), Decision on Annulment, 3 May 1985, 2 ICSID Reports 95, 122 (para 69); Amco v Indonesia (ICSID Case No ARB/81/1), Decision on Annulment, 16 May 1986, 1 ICSID Reports 509, 515 (paras 20–22).

94  See H Lauterpacht, The Function of Law in the International Community (Oxford, Clarendon Press, 1933), 65.

95  In addition to Klöckner and Amco, see: LETCO v Liberia (ICSID Case No ARB/83/2), Award, 31 March 1986, 2 ICSID Reports 358, 372; SPP v Egypt (ICSID Case No ARB/84/3), Award, 20 May 1992, 3 ICSID Reports 189, 207, 208 (paras 80, 83); Compañía del Desarollo de Santa Elena SA v Costa Rica (ICSID Case No ARB/96/1), Award, 17 February 2000, 5 ICSID Reports 153, 170 (paras 64–65).

96  Z Douglas, ‘The Hybrid Foundations of Investment Treaty Arbitration’ (2003) 74 BYIL 151, 160–162.

97  See ICSID Convention, art 47 and ICSID Rules, r 39. There is now some doubtful authority that ‘recommend’ actually means ‘prescribe’: see CH Schreuer, L Malintoppi, A Reinisch, & A Sinclair, The ICSID Convention (2nd edn, Cambridge, CUP, 2009), 764–765.

98  See eg CSOB v Slovakia (ICSID Case No ARB/97/4), Award, 24 May 1999, 5 ICSID Reports 330.

99  Arts 54(1), 54.3.

100  [2005] EWHC 2239 (Comm); 11 ICSID Reports 118.

101  Ibid.

102  Attorney-General v Mobil Oil NZ Ltd (1987) 118 ILR 620, 630.

103  See also, in England: s 3(2) of the Arbitration (International Investment Disputes) Act 1966, by which s 9 of the Arbitration Act 1996 applies to applications to stay in favour of ICSID arbitrations.

104  MINE v Guinea, 693 F.2d 1094, 1103–4 (12 November 1982).

105  See eg in England: s 3(1) of the Arbitration (International Investment Disputes) Act 1966, by which the Lord Chancellor can direct that ss 36, 38–44 of the Arbitration Act 1996 apply to ICSID arbitrations.