ICSID Case No. ARB/16/4
between
EURUS ENERGY HOLDINGS CORPORATION
and
KINGDOM OF SPAIN
Mr. Oscar M. Garibaldi
Arbitrator
17 March 2021
[Page 1]
1. I dissent from the decisions of my esteemed colleagues (henceforth, the Majority) (i) to hold EU law to be a part of the applicable law under Article 26(6) of the ECT and (ii) to dismiss the bulk of the Claimant’s claims grounded on Article 10(1) of the ECT. I concur, on substantially different grounds, in the Majority’s decisions that (iii) the claw-back feature of the Disputed Measures breached the first two sentences of Article 10(1) of the ECT and (iv) EU law on state aid does not bar the claim based on the claw-back feature.1 The purpose of this dissenting opinion is to explain (a) to what extent and why I disagree with the Majority’s decisions and their supporting reasons on points (i) and (ii) and (b) to what extent and why I concur in the Majority’s decisions on points (iii) and (iv) but disagree on the reasoning on which they are based.
2. I regard a dissenting opinion as an opportunity to voice criticism on important points of law or fact on which no agreement has been reached. If we are to make progress in our search for better solutions to legal issues that are serious and recurring, we should be ready to criticize the decisions of others and the grounds on which those decisions are based, as well as to open our minds to countercriticism. That is why it is not enough, in my view, for a dissenting arbitrator merely to state an alternative position, as it is often done. The dissenter should explain, respectfully, what he or she sees as flaws in the majority’s position, so that others may assess the relative strength of the opposing reasons. This is the approach to be followed in this Partial Dissent.
1 I also dissent, consequentially, from the Majority’s decisions on the quantum of compensation. If my views on the merits had been accepted, the compensation to which the Claimant is entitled should have been calculated under different principles. For the purposes of this Partial Dissent, it is not necessary to develop this point further. ↩
[Page 2]
3. I should make it clear that I hold my colleagues in the highest regard, personally and professionally. My criticism of their views is always directed ad sententias et argumenta, never ad homines.
4. As will become apparent, while I agree in some respects with the Majority’s analysis and conclusions on the matters subject to this dissent, I disagree in many (perhaps most) other respects. Because the points of agreement are intertwined with those of disagreement, a comprehensive discussion is necessary. And because I find the Majority’s analysis flawed in critical respects, I prefer to discuss most contested issues in my own words.
5. The main issue presented under the rubric of “Applicable Law” is whether the law to be applied by the Tribunal in this case should include EU law. The Majority acknowledges that Japan is not a party to the EU treaties and “is not bound by them as such.”2 Yet, the Majority takes the view that the Tribunal is allowed “to note the settled legal consequences of facts established in the course of the proceedings, for example, the fact that unnotified state aid gives rise to adverse consequences under Article 108(3) of the TFEU.”3 On this basis, the Majority concludes: “To that extent, at least, EU law is part of the applicable law under Article 26(6) of the ECT.”4 True to that conclusion, when it comes to discussing EU state aid, the Majority applies EU law as law.5
6. The Majority’s conclusion that EU law is part of the applicable law is unsupported by the ECT and contradicts the basic principle of international law that treaties are not binding on non-parties. As explained below, except to the extent that the ECT itself may contain a renvoi to EU law, the Tribunal may take cognizance of EU law only as a fact, and only to
2 Decision on Jurisdiction and Liability (henceforth, Decision), ¶ 232. Unless stated otherwise, the abbreviations used in this Partial Dissent are the same as those used in the Decision. ↩
4 Id. I note that the phrase “at least” leaves open the possibility of a broader application of EU law. ↩
5 Decision, ¶¶ 423(b), 428-32, especially 427, 432. The way in which the Majority applies EU law is discussed in detail infra, ¶¶ 121-139. ↩
[Page 3]
the extent necessary to apply the ECT and other rules and principles that are applicable as law.
7. The starting point of the analysis must be Article 26(6) of the ECT, which prescribes the law to be applied in an arbitration under Part V:
“A tribunal established under paragraph (4) shall decide the issues in dispute in accordance with this Treaty and applicable rules and principles of international law.”6
8. This provision directs the Tribunal to apply, in addition to the ECT itself, “applicable rules and principles of international law.”7 The EU treaties are (or rather contain) rules and principles of international law, but they are not “applicable” in this case, because Japan is not a party to them and, as the Majority acknowledges, is not bound thereby.8 Therefore, EU law cannot be part of the substantive law applicable in this case.
9. The Majority also refers to Article 16 of the ECT, which provides that if two or more Contracting Parties to the ECT have entered or may enter into a treaty concerning the subject matter of Part III or Part V of the ECT, neither the other treaty nor the ECT shall be construed to derogate from a provision of each other that is more favourable to the Investor or the Investment.9 In this case, Article 16 is plainly inapplicable, because Spain and Japan have not entered into any other treaty concerning the subject matter of Part III or Part V of the ECT. Nevertheless, the Majority takes the view that “Article 16 remains indirectly relevant as concerns the relation between the ECT and the TFEU, in that by clear implication the ECT prevails over the TFEU to the extent that the ECT provision ‘is more
6 CL-0001, Energy Charter Treaty (henceforth, ECT), Art. 26(6). ↩
8 RL-0010, Vienna Convention on the Law of Treaties (henceforth, VCLT), Art. 26 (“Pacta sunt servanda. Every treaty in force is binding upon the parties to it and must be performed by them in good faith.”) ↩
9 CL-0001, ECT, Art. 16 (“Where two or more Contracting Parties have entered into a prior international agreement, or enter into a subsequent international agreement, whose terms in either case concern the subject matter of Part III or V of this Treaty, (1) nothing in Part III or V of this Treaty shall be construed to derogate from any provision of such terms of the other agreement or from any right to dispute resolution with respect thereto under that agreement; and (2) nothing in such terms of the other agreement shall be construed to derogate from any provision of Part III or V of this Treaty or from any right to dispute resolution with respect thereto under this Treaty, — [applying to both (1) and (2)] where any such provision is more favourable to the Investor or Investment.”) ↩
[Page 4]
favourable to the Investor or Investment’.”10 The Majority offers no basis for this alleged implication, other than saying that “[o]ne would not expect a non-EU member state, or the nationals of such a state, to be treated less well.”11 The interpreter’s expectations, however, are not a substitute for what Article 16, properly interpreted under the rules of the VCLT, actually provides. In fact, the ECT does prevail over the TFEU in this case, but not because one would expect the rule of Article 16 to be extended to Japan or its nationals, but because the TFEU is inapplicable, and hence there is no conflict for Article 16 to resolve.
10. The Majority correctly dismisses the Respondent’s argument that, under the authority of the CJEU’s pronouncements in Achmea, the Tribunal has no competence to apply EU law.12 The Majority does so by construing Achmea narrowly, as not applying to a multilateral treaty such as the ECT, to which the EU itself is a party.13 Such a narrow construction is a possible one, although there are more fundamental reasons to conclude that the Achmea decision does not control the Tribunal’s competence, a competence which does not derive from the EU treaties but from the ECT.14 In the present case, however, the question of the Tribunal’s competence to apply EU law as law is academic, because, for the reasons stated, EU law is not part of the applicable law.
11. All that being said, the Majority is right to point out that “the EU treaties have established legal regimes for regulating matters such as state aid, which are furthermore directly applicable as part of the law of the member states,” with the consequence that “[t]o the extent that Japanese or other third-state corporations establish activities in the EU that are regulated by those regimes, they may be affected by them.”15 Those legal regimes and the consequences derived therefrom are among the factual circumstances that the Tribunal must take into consideration in applying the provisions of the ECT. To give those facts proper consideration in the proper context, it is neither necessary nor justifiable to turn
14 See, e.g., Vattenfall AB and others v. Federal Republic of Germany, ICSID Case No. ARB/12/12, Decision on the Achmea Issue, 31 August 2018. ↩
[Page 5]
them into applicable law. In this respect, EU law is in a position akin to that of Spanish law, in which it happens to have been incorporated: Spanish laws and judicial and administrative decisions are to be considered as facts, except to the extent that the ECT might require, expressly or by clear implication, that they be applied as law.16
12. The Claimant’s claims are based on (i) Article 13(1) of the ECT, which concerns the expropriation of investments, and (ii) three clauses of Article 10(1) of the ECT, which concerns the promotion, protection, and treatment of investments. The Tribunal unanimously dismisses the claim based on Article 13(1) on grounds with which I am generally in agreement. As for the claims based on Article 10(1) of the ECT, the Tribunal (by majority) dismisses them, except for the claim related to the claw-back feature of the Disputed Measures, which the Tribunal upholds under the principle of stability derived from the first and second sentences of Article 10(1) of the ECT.
13. I concur in the portion of the Majority’s decision which upholds the claim related to the claw-back feature of the Disputed Measures, although the reasons for my concurrence are different from those of the Majority. I dissent from the Majority’s dismissal of the remainder of those claims and from the reasoning supporting the dismissal. I explain in this section the extent of my partial concurrence, the extent of my dissent, and the reasons therefor.
14. The starting point of the discussion must be Article 10(1) of the ECT. The entire text of that provision, with added numbers for each sentence and italics to set out the clauses on which the Claimant relies, is as follows:
16 Certain German Interests in Polish Upper Silesia (Germany v. Poland), 1926 P.C.I.J. (ser. A) No. 7, ¶ 52 (“municipal laws are merely facts which express the will and constitute the activities of States, in the same manner as do legal decisions or administrative measures.”) For example, the ECT implicitly requires the application of Spanish law as law to determine whether the Claimant had rights subject to expropriation under the rules of Article 13 of the ECT. ↩
[Page 6]
“[1] Each Contracting Party shall, in accordance with the provisions of this Treaty, encourage and create stable, equitable, favourable and transparent conditions for Investors of other Contracting Parties to make Investments in its Area. [2] Such conditions shall include a commitment to accord at all times to Investments of Investors of other Contracting Parties fair and equitable treatment. [3] Such Investments shall also enjoy the most constant protection and security and no Contracting Party shall in any way impair by unreasonable or discriminatory measures their management, maintenance, use, enjoyment or disposal. [4] In no case shall such investments be accorded treatment less favourable than that required by international law, including treaty obligations. [5] Each Contracting Party shall observe any obligations it has entered into with an Investor or an Investment of an Investor of any other Contracting Party.”17
15. The Claimant’s first claim under Article 10(1) of the ECT is purportedly grounded on the first and second sentences taken together. The Majority likewise discusses those two sentences together, on the ground that the first sentence cannot be interpreted in isolation from the second. To simplify the presentation, I shall also address the two sentences in the same subsection, although I shall analyse them individually (Subsection B).
16. The Claimant’s second claim under Article 10(1) of the ECT, based on the italicised portion of the third sentence, will be discussed in Subsection C.
17. The Majority ends its analysis by addressing the Respondent’s defences based on EU rules on state aid. One defence is dismissed altogether and the other is dismissed in respect of the sole claim which the Majority upholds, which is based on the claw-back feature of the Disputed Measures. I concur in the dismissal of those defences, but to a different extent and on different grounds (Subsection D).
17 CL-0001, ECT, Art. 10(1) (emphasis added). ↩
[Page 7]
18. The first sentence of Article 10(1) states: “Each Contracting Party shall, in accordance with the provisions of this Treaty, encourage and create stable, equitable, favourable and transparent conditions for Investors of other Contracting Parties to make Investments in its Area.”18 This provision, as well as all other provisions of the ECT at issue in this case, must be interpreted under the rules of interpretation of the VCLT. First and foremost, the provision must be interpreted in good faith in accordance with the ordinary meaning to be given to its terms in their context and in the light of the object and purpose of the treaty.19
19. The Majority states, citing some earlier decisions in support, that the first sentence of Article 10(1) cannot be interpreted in isolation from the second sentence.20 It would be more accurate to say that the first and the second sentences form the proximate context of each other, and context is one of the primary elements of interpretation under the rules of the VCLT. Aside from the role of proximate context in the interpretation of both sentences, there is no reason why the two sentences cannot be applied separately, each in accordance with its own prescriptive content.21
20. As for the content of those obligations, the Majority states that the first sentence “does not give a general mandate to ECT tribunals to decide whether government decisions affecting investments are ‘equitable’ or ‘favourable.’”22 That statement is correct, but as a comment on the content of the first sentence it misses the mark. The first sentence does not purport
19 RL-0010, VCLT, Arts. 31 and 32, especially Art. 31.1. ↩
21 By “prescriptive content” of a treaty provision I mean the conduct that the provision qualifies as permitted, not permitted, or obligatory, including the set of conditions under which such qualification takes effect. I do not understand the Majority’s position to be that (i) the first sentence is merely hortatory or otherwise lacks prescriptive force, or (ii) the prescriptive content of the first sentence is subsumed within that of the second sentence, so that the first sentence does not really impose any separate obligation. These theories, which several tribunals have espoused, have no basis in the ECT. The first sentence is couched in the same prescriptive language (“shall”) as the second. A tribunal faced with multiple treaty provisions framed in equally mandatory language does not have discretion to select which ones it will treat as prescriptive and which ones it will not. ↩
[Page 8]
to give ECT tribunals any such “general mandate” nor does the scope of the sentence refer to (any and all) “government decisions affecting investments.” In fact, the scope of the obligations arising from the first sentence and the effects of that sentence on the Claimant’s claims are more limited than the concerns that prompted the Majority’s statement.
21. The first sentence requires the state to encourage and create “stable, equitable, favourable and transparent conditions” for the making of investments. Leaving aside for the moment what counts as stable, equitable, favourable, and transparent (a matter to be discussed later), those conditions concern only “the making of Investments.” Article 1(8) of the ECT defines “Make Investments or Making of Investments” as “establishing new Investments, acquiring all or part of existing Investments or moving into different fields of Investment activity.”23 Therefore, the first sentence of Article 10(1) concerns the conditions that the Respondent was required to create for the Claimant to carry out new investment activities. In the circumstances of this case, this amounts primarily to a reference to the “original” conditions, that is, the conditions that the Respondent had created and existed at the time the Claimant made each of its investments in eolic energy in Spain.
22. The first sentence of Article 10(1) requires that those original conditions be stable, equitable, favourable, and transparent. In this case, however, the Claimant does not allege that the original conditions for each of its investments (which were not necessarily the same) were inequitable, unfavourable, or non-transparent. With the possible exception of the requirement that those conditions be stable, the Claimant’s claims are not based on an alleged inconsistency between the original conditions and the requirements of the first sentence; they concern a subsequent change, to the Claimant’s detriment, of (i) the original conditions or (ii) conditions imposed subsequently which the Claimant had accepted or to which it had acquiesced. Therefore, again with the possible exception of the requirement of stability, the first sentence cannot provide a legal basis for the Claimant’s claims.
23. The requirement that the original conditions be “stable” is different in an important respect from the other requirements set forth in the first sentence. In the case of the other
[Page 9]
requirements, it is possible to assess at a single point (e.g., at the time the investment was made) whether the original conditions were equitable, favourable, and transparent – to such degree as may be required. But in the case of the requirement that the original conditions be “stable,” it is not possible to assess compliance with the requisite degree of stability until some time should have elapsed. Stability, whatever the required degree, is a quality that the original conditions must possess over time, i.e. a certain degree of continuity or lack of change over a certain period. Accordingly, if the stability of the original conditions is affected by the subsequent treatment of the investment, the requirement of the first sentence that the original conditions be “stable” may overlap with the requirements of the second sentence concerning the treatment of the investment “at all times.” The existence and extent of the overlap will depend on a comparative analysis of the requirement of stability in the first sentence with the requirement of fair and equitable treatment in the second sentence.
24. I conclude that, aside from the requirement that the original conditions be “stable,” the first sentence of Article 10(1) cannot serve as a basis for the Claimant’s claims. In view of this conclusion, it is unnecessary to discuss, in the context of the first sentence, the requirement that the original conditions be equitable, favourable, and transparent. The requirement of stability under the first sentence will be examined, for convenience, together with the requirement of fair and equitable treatment under the second sentence.
25. The second sentence of Article 10(1) requires a commitment to accord the investment, at all times, fair and equitable treatment. The Majority notes that Article 10(1) has been extensively discussed in earlier decisions and finds little point in going over the same ground. The Majority makes two general observations and immediately goes on to adopt the “test” formulated in Blusun v. Italy.24
[Page 10]
26. The Majority’s first general observation is that the fair-and-equitable-treatment standard does not give tribunals sitting under investment treaties general discretion to impose their own views as to “fairness” and “equity.”25 This observation is of course correct. Arbitrators do not have the power, under the guise of interpretation, to treat the terms of a treaty as empty vessels in which to pour their own policy preferences or to improve on the textual choices made by the drafters of the treaty. But, as we shall see, neither do they have the power to discard the terms of the treaty and apply instead rules, standards, or tests of their own making.
27. The Majority’s second observation is that the first and second sentences of Article 10(1) embody a legal standard, “which takes into account the prerogatives and responsibilities of governments as well as the rights and interests of investors, including their interest in stability.”26 This statement is so general that it is difficult to take issue with it, except to note that it says nothing about what that legal standard exactly is or on what it is based.
28. Then the Majority adopts as its own the following dictum in Blusun v. Italy:
“In the absence of a specific commitment, the state has no obligation to grant subsidies such as feed-in tariffs, or to maintain them unchanged once granted. But if they are lawfully granted, and if it becomes necessary to modify them, this should be done in a manner which is not disproportionate to the aim of the legislative amendment, and should have due regard to the reasonable reliance interests of recipients who may have committed substantial resources on the basis of the earlier regime.”27
29. The Blusun dictum then becomes the roadmap for the Majority’s analysis of the Claimant’s claims under the first and second sentences of Article 10(1) and for the ultimate disposition of those claims. I disagree with the so-called Blusun “test” and the way the Majority applies it, on two levels.
27 Decision, ¶ 315, quoting from CL-0083, Blusun v. Italy, ¶ 319(5). ↩
[Page 11]
30. On a more general and more fundamental level, I perceive no legal or theoretical justification for applying the Blusun “test” (or any comparable arbitrator-made “test” or standard) in place of the fair-and-equitable-treatment standard set forth in the second sentence of Article 10(1). The second sentence requires “fair and equitable treatment,” not something else. The Blusun “test” is in fact a complex arbitrator-made rule, designed to serve as the governing rule for decision in cases concerning subsidies first granted and later modified. That rule is framed in terms of tiered sub-rules, concepts, and criteria which are different from the “fair and equitable treatment” standard, and the Majority makes no effort to demonstrate that those sub-rules, concepts, and criteria derive from or are otherwise connected (in a manner consistent with the rules of the VCLT) with the “fair and equitable treatment” standard.
31. In fact, the Majority adopts and uses the Blusun rule as a substitute for the “fair and equitable treatment” standard. This is incorrect. Just as a tribunal is not authorized to impose its own views concerning what is “fair” or “equitable,” as the Majority correctly points out, a fortiori a tribunal is not authorized to toss aside the “fair and equitable treatment” standard imposed by the treaty and apply instead a substitute rule, “test,” or standard of its own choosing.
32. On a more specific level, I find the Blusun rule a poor substitute for the “fair and equitable treatment” standard. If the “fair and equitable treatment” standard is sometimes criticized on the ground that it is not precise enough, the Blusun rule is hardly an improvement. I shall analyse each component of the rule later.28 For the moment, suffice it to note that the concepts of “necessary to modify” and “due regard” are neither precise nor unambiguous. Nor is it self-evident (i) whether or how the twin restrictions set forth in the Blusun rule (proportionality to the aims of the challenged measure and “due regard” for the investor’s reliance interests) are related to the concepts of fairness and equity or (ii) if so related, whether, how, or why they are jointly sufficient to satisfy the requirement that the treatment be “fair and equitable.”
[Page 12]
33. It cannot be seriously claimed that the term “fair and equitable treatment” is exempt from the need for interpretation. But any interpretation must be done by following the rules of the VCLT, not by adopting a substitute rule or standard unconnected with the text of the treaty. It bears repeating that, under the primary rule of interpretation of the VCLT, the second sentence of Article 10(1) must be interpreted in good faith in accordance with the ordinary meaning of the terms used in their context and in the light of the object and purpose of the ECT. I shall discuss each of these elements in turn.
34. The terms of the second sentence are: “Such conditions shall include a commitment to accord at all times to Investments of Investors of other Contracting Parties fair and equitable treatment.”29 “Such” indicates that the term it modifies (“conditions”) is one that has been referred to previously – in this case in the preceding sentence. Hence, “[s]uch conditions” can only refer to the conditions mentioned in the first sentence, i.e. “stable, equitable, favourable and transparent conditions.”30
35. The ordinary meaning of “fair and equitable treatment” is not a particularly difficult issue. Several tribunals have started their analyses by quoting the dictionary meaning of “fair” and “equitable.”31 I regard that step as not quite necessary, because it is reasonable to assume that most people, and certainly all lawyers, know the ordinary meaning of these terms. The difficulty (if there is one) lies not in a lack of understanding of what “fair” or “equitable” means in ordinary language, but in the process of applying these concepts to a particular instance of treatment in the circumstances of a particular case. But that difficulty
29 CL-0001, ECT, Art. 10(1) (emphasis added). ↩
30 “Such conditions” cannot refer to the conditions “to make Investments,” that is, to the original conditions for a new investment activity, because the second sentence makes clear that the required treatment must be accorded “at all times.” ↩
31 See, e.g., CL-0114, Antin Infrastructure Services Luxembourg S.à.r.l. and Antin Energía Termosolar B.V. v. Kingdom of Spain (ICSID Case No. ARB/13/31), Award, 15 June 2018, ¶ 518 (“‘fair’ means ‘just, unbiased, equitable, impartial, legitimate”; and “‘equitable’ is defined as ‘characterised by equity or fairness’, where ‘equity’ means ‘fairness; impartiality; even-handed dealing’.”); CL-0028, MTD Equity Sdn. Bhd. and MTD Chile S.A. v. Republic of Chile, ICSID Case No. ARB/01/7, Award, 25 May 2004, ¶ 113 (“just”, “even-handed”, “unbiased”, “legitimate”). ↩
[Page 13]
(if it exists) is neither extraordinary nor unsolvable. Applying general concepts like “fair” and “equitable” is not different in kind from what adjudicators do and have always done: applying general rules and concepts to the circumstances of particular cases.
36. It is sometimes said that the term “fair and equitable treatment” expresses a legal concept. That is true, in the sense that this term and cognate terms are frequently used in legal texts and legal discourse, as well as in ordinary language. In fact, by the time the ECT was drafted and opened for signature (17 December 1994), the term “fair and equitable treatment” had been used for decades in numerous investment-protection treaties and had been interpreted and applied in multiple arbitral awards. Those awards had frequently linked the term “fair and equitable treatment” to various concepts, including consistency, transparency, stability, and respect for objectively reasonable expectations derived from antecedent state action. These observations do not mean, however, that “fair and equitable treatment” should be given a special meaning, in the sense of Article 31.4 of the VCLT, because it has not been established in this case that the parties to the ECT intended to adopt any particular exegesis of “fair and equitable treatment” made by any pre-December 1994 decision or set of decisions. Much less do these antecedents justify that “fair and equitable treatment” be regarded as an open-ended legal concept, the meaning of which should depend on “evolving” interpretations adopted by tribunals in pre- or post-ECT decisions, which in any event are not binding on this Tribunal. On the contrary, “fair and equitable treatment” must be regarded as a term which had an objective meaning at the time the ECT was framed – a meaning to be established in good faith under the rules of the VCLT.
37. As already noted, the second sentence of Article 10(1) immediately follows the first sentence and explicitly refers to a term contained therein. Accordingly, the first sentence forms the proximate context of the second, and this context must especially be taken into account in the interpretation of the latter.32 The drafters of the ECT could simply have
32 For a similar analysis, see, e.g., RL-0098, Dissenting Opinion of Kaj Hobér in Stadtwerke München GmbH, RWE Innogy GmbH, and others v. Kingdom of Spain (ICSID Case No. ARB/15/1), 20 November 2019 (henceforth, Hobér Dissent), ¶ 6. ↩
[Page 14]
provided, as other treaties do, that each Contracting Party shall accord at all times fair and equitable treatment to the Investments of Investors of other Contracting Parties. Instead of doing so, they chose a more elaborate formulation, which links the commitment to accord fair and equitable treatment to the “stable, equitable, favourable, and transparent conditions” referred to in the first sentence.
38. This observation suggests that the drafters understood “fair and equitable treatment” to belong to a larger class of “stable, equitable, favourable, and transparent conditions.” This does not imply, however, that “fair and equitable treatment” is the same as “stable, equitable, favourable, and transparent conditions” (“equitable” being redundant), because the second sentence does not say so. But this close contextual relationship does indicate that stability, favourability, and transparency (to certain degrees, as will be discussed) are among the relevant factors or circumstances that should be taken into account in determining whether a given instance of treatment is “fair and equitable.”
39. The traditional way of ascertaining the object and purpose of a treaty is to look at any statement of purpose that it might contain. Article 2 of the ECT, titled “Purpose of the Treaty,” states that the ECT “establishes a legal framework in order to promote long-term cooperation in the energy field, based on complementarities and mutual benefits, in accordance with the objectives and principles of the [European Energy] Charter.”33
40. The European Energy Charter is a declaration of goals and principles for international cooperation in the energy field, adopted in 1999.34 It consists of a very general statement of goals (Title I: Objectives), a set of more specific (but still general) ideas about pursuing those goals (Title II: Implementation), and a set of undertakings to negotiate certain specific agreements (Title III: Specific Agreements).35
34 CL-0001, European Energy Charter. ↩
35 CL-0001, European Energy Charter, passim. ↩
[Page 15]
41. As regards the protection of foreign investment, the most relevant and specific statement of “objectives and principles” can be found in Title II. Paragraph 4 of Title II states that “it is important for the signatory States to negotiate and ratify legally binding agreements on promotion and protection of investments which ensure a high level of legal security [...].”36 The ECT is the product of the negotiations called upon by this statement. Accordingly, the reference to the objectives and principles of the European Energy Charter suggests that one purpose sought by the drafters of the ECT was to provide a “high level of legal security” to investments.37
42. Another way of ascertaining the object and purpose of the ECT is to look at the overall structure and content of the treaty, that is, what the treaty actually does. What Part III of the ECT fundamentally does is to lay down certain international standards which Contracting States are required to observe in their treatment of investments of nationals of other Contracting States. This observation is a sufficient reason to reject, as inconsistent with the object and purpose of the treaty, any interpretation of “fair and equitable treatment” that seeks directly or indirectly to reduce the international standards imposed by the ECT to mere observance of the host state’s law.
43. The term “fair and equitable treatment” must therefore be interpreted in good faith in accordance with its ordinary meaning in the context and in the light of the object and purpose just discussed. Yet, stability, transparency, and favourability are matters of degree, and the goal of providing a “high level” of legal security for investments does not indicate a precise point in the metaphorical height-scale of protection. Because these are matters of degree, the preceding observations on context and object and purpose do not imply by
36 CL-0001, European Energy Charter, II.4. The immediately preceding paragraph states that “the signatories will at national level provide for a stable, transparent legal framework for foreign investments, in conformity with the relevant international laws and rules on investment and trade.” Id. (emphasis added). This passage plainly refers to actions to be taken internally, at the national level, while the immediately following paragraph (the pertinent part of which is quoted in the text) refers to the negotiation and adoption of the binding international agreements. It follows that the passage concerning actions to be taken at the national level is not relevant to determine the purposes to be served by the binding international agreements. ↩
[Page 16]
any means that for an instance of treatment to be “fair and equitable” it must satisfy an absolute degree of stability, transparency, or favourability, or to reach an absolute height of protection – if such things are capable of being so measured. In fact, those elements of context and of object and purpose indicate to the interpreter a direction, more than a destination. The direction is very important, because it indicates the intent of the framers of the ECT. But the destination, the end point, the precise degree of stability, transparency, and favourability required in each particular case is determined by what is “fair and equitable” in that case, that is, by the ordinary meaning of “fair and equitable treatment.”
44. “Fair and equitable” are the defining attributes of the treatment that the second sentence of Article 10(1) of the ECT requires. But these are not inherent attributes of a given instance of treatment, in the sense that the date of a regulation or the colour of the paper on which it was written are. What is “fair and equitable,” as these terms are ordinarily used, depends on the surrounding circumstances. Barring extreme examples, a given instance of treatment (i.e. the same state measure) may be fair and equitable in one set of circumstances and unfair and inequitable in another set of circumstances. That is why prior arbitral decisions regarding “fair and equitable treatment” must be approached with particular caution, even if they relate to the same measure under the same treaty, because not all the relevant circumstances in the two cases are necessarily the same.38 Therefore, the “fair and equitable treatment” standard should be applied in light of all the relevant circumstances of the case, not in light of a subset of those circumstances selected by arbitrators and turned into substitute standards.
45. The preceding conclusion does not require an arbitrator to ignore that, over a span of several decades, tribunals have identified various patterns of treatment as raising questions of fairness and equity. For example, the degree of stability, transparency, or consistency of a given set of treatment conditions may raise such questions and, depending on the
38 The Tribunal takes the view that “concordant decisions on the interpretation and application of the ECT merit serious consideration, especially if they rise to the level of a jurisprudence constante.” Decision, ¶ 239. I subscribe to this view in the understanding that prior decisions should be given serious consideration in a critical, not deferential, manner. Each tribunal is entitled to interpret and apply the ECT on the basis of the VCLT and in the light of the arguments presented to it. This is a moot point, however, because arbitral jurisprudence on the matters at issue in this case can hardly be considered concordant or constante. ↩
[Page 17]
remaining circumstances, compel an answer one way or another. It is no accident that the framers of the ECT chose to refer specifically to stable and transparent conditions in the first sentence of Article 10(1). But in the context of the second sentence of Article 10(1), stability, transparency, and consistency are factors to be taken into account in a fair-and-equitable-treatment analysis; they are not requirements to be used as a substitute for the “fair and equitable treatment” standard.
46. The same can be said of another widely recognized pattern of treatment which may raise questions of fairness and equity: the frustration of so-called “legitimate” expectations.39 This pattern of treatment consists of a complex set of interactions between a state and a foreign investor in which the state (i) takes actions that give rise to objectively reasonable expectations on the part of the investor and then (ii) takes some further action that frustrates those expectations. The Majority relies on a statement in Blusun which notes, among other things, that the term “legitimate expectations” does not appear in the ECT or other treaties.40 The Majority takes the view that such expectations are essentially consideranda: “they are relevant factors to be taken into account in the interpretation and application of treaty standards such as Article 10(1) of the ECT, first and second sentences.”41 I agree with this view, insofar as it concerns the application of the fair-and-equitable-treatment standard under the second sentence. The pattern of treatment that the Majority calls “legitimate” expectations is relevant to the fair-and-equitable-treatment analysis but does not exhaust all the relevant considerations. The Majority fails to appreciate, however, that if the term “legitimate expectations” does not appear in the ECT or other treaties, neither does the Blusun rule or any of its components. If the former are properly reduced to mere consideranda, a position with which I agree, by the same token the same must be true of the latter.
47. The statement in Blusun on which the Majority relies ends with the sentence: “International law does not make binding that which was not binding in the first place, nor render
39 On this terminology, see infra, ¶ 79. ↩
40 Decision, ¶ 316, quoting Blusun, ¶ 371. ↩
[Page 18]
perpetual what was temporary only.”42 This statement is overbroad. It is important to show where it goes astray, lest it be used (improperly) as an a priori constraint on the interpretation of the fair-and-equitable-treatment standard. It is true that customary international law does not obligate states to refrain from frustrating objectively reasonable expectations, as the ICJ ruled in Obligation To Negotiate Access to the Pacific Ocean Case,43 but customary international law does make binding certain unilateral statements made on behalf of a state that may not be binding otherwise, as the ICJ ruled in the Nuclear Tests Case.44 More to the point, absent a violation of ius cogens there is no reason of principle why treaty law may not make binding what was not binding in the first place (under customary international law or municipal law); it all depends on the terms of the treaty. Therefore, for all its aphoristic qualities, the quoted sentence in Blusun cannot be relied upon to foreclose the possibility that the fair-and-equitable-treatment standard of the ECT might have, in certain circumstances, the effect of holding a state to its commitments. The fair-and-equitable-treatment standard of the ECT must be interpreted under the rules of the VCLT, not under a priori conceptions about what a treaty may or may not prescribe.
48. As noted, the first sentence of Article 10(1) requires that the conditions for the making of investments be, inter alia, “stable.” This stability requirement, which can only be assessed over time, may overlap with the obligation to afford fair and equitable treatment at all times, as stability is one of the relevant circumstances to be taken into account in the application of the fair-and-equitable-treatment standard. A stability overlap can only be partial, because the two sentences have different scopes of application (the first applies to the making of investments; the second at all times). The overlap would also depend on the extent to which the meaning of “stability” and the required degree of stability are the same under both sentences. In this respect, the use of the same terms and the close contextual relation suggest that both sentences refer to stability in the same sense and to the same
42 Decision, ¶ 316, quoting Blusun, ¶ 371. ↩
43 International Court of Justice, Obligation to Negotiate Access to the Pacific Ocean (Bolivia v. Chile), Judgment of 1 October 2018, ¶ 162. ↩
44 International Court of Justice, Nuclear Tests Case (Australia v. France), Judgment of 20 December 1974, ¶¶ 43-46. ↩
[Page 19]
degree. Consequently, the discussion on stability in the next subsection is meant to apply to both sentences of Article 10(1).
49. As noted, to determine whether the Disputed Measures meet the fair-and-equitable-treatment standard of the ECT, it is necessary to consider all the relevant circumstances of the case. For the purposes of this Partial Dissent, however, I shall place particular attention to the circumstances which the Majority and I assess differently. To facilitate the analysis, the relevant circumstances can be grouped as follows: (i) circumstances related to the Respondent’s antecedent conduct, i.e. its conduct preceding the Disputed Measures; (ii) circumstances related to the Claimant’s conduct, including its expectations and reactions based on the Respondent’s antecedent conduct; and (iii) circumstances related to the Disputed Measures. I shall discuss these three sets of circumstances separately and then draw general conclusions on the application of the fair-and-equitable-treatment standard.
50. The Respondent’s relevant antecedent conduct consisted primarily of adopting laws and regulations on the generation of electricity through conversion from eolic energy.45
51. In the relevant period, that is, the period in which the Claimant made investments in Spain in the eolic sector, the Respondent adopted two statutes and four material royal decrees governing that sector. Those statutes and royal decrees established four successive legal regimes, each consisting of (i) a statute establishing a general legal framework for the regulation of this subject and (ii) a royal decree “developing” the statute, i.e. regulating the
45 The Claimant invokes also certain statements made by a representative of the Xunta in Japan concerning investment in wind plants in Galicia. I agree with the Majority, however, that those statements cannot be invoked as the source of commitments by the Respondent. See Decision, ¶ 322. ↩
[Page 20]
same matters in greater detail, within the general parameters of the statute. The last three legal regimes included, and were based on, the same statute.
52. Those four successive legal regimes were the following:
53. The First, Second, and Third Regimes are primarily relevant to an analysis of the Claimant’s conduct, a topic which will be addressed separately. For the purposes of the present discussion, the Fourth is the most relevant legal regime. The Fourth Regime (with a relatively minor change introduced later by RD 1614/2010)46 was in effect at the time the Disputed Measures were adopted and was replaced by those measures. If the Disputed Measures breached the standard of fair and equitable treatment under the ECT, it must be because (inter alia) they abrogated the Fourth Regime, not the preceding three regimes, which had generally ceased to be in effect long before the Respondent adopted the Disputed Measures.
54. Two fundamental questions must be asked in respect of the Fourth Regime:
a. Commitments: whether the regime set forth one or more commitments, undertakings, promises, assurances, guarantees, or other forms of representation (to be collectively referred to as commitments) that the investors or investments in the eolic energy sector would be treated in a certain way, including any commitment concerning the duration of any such treatment; and
b. Specificity: how specific were any such commitments.
46 C-0010, RD 1614/2010. The changes introduced by this Royal Decree are discussed infra, ¶ 85. The Claimant asserts that these changes had no significant effect on its investment. Claimant’s Memorial, ¶ 126. ↩
[Page 21]
55. The question whether the Fourth Regime contained commitments requires an objective answer, based on what the Respondent said and did in adopting that regime. The Claimant’s understanding of any such commitments and any expectations it derived therefrom are separate matters, to be discussed later as circumstances related to the Claimant’s conduct.47
56. The Fourth Regime consisted of Law 54/1997 and RD 661/2007. Law 54/1997 established a special regime for the generation of electricity from wind and other renewable sources, a pattern which had been adopted by the predecessor statute. The special-regime producers were entitled to sell their energy at a price consisting of (i) a component determined by a market-price mechanism and (ii) a premium (prima) to be determined by the Government on the basis of certain criteria.48 Article 30.4 of Law 54/1997 established the legal framework for the premiums:
“Adicionalmente, la producción de energía eléctrica mediante energías renovables no hidráulicas [...] percibirá[ ] una prima que se fijará por el Gobierno de forma que el precio de la electricidad vendida por estas instalaciones se encuentre dentro de una banda porcentual comprendida entre el 80 y el 90 por 100 de un precio medio de la electricidad, que se calculará dividiendo los ingresos derivados de la facturación por suministro de electricidad entre la energía suministrada. [...]
Para la determinación de las primas se tendrá en cuenta el nivel de tensión de entrega de la energía a la red, la contribución efectiva a la mejora del medio ambiente, el ahorro de energía primaria y a la eficiencia energética, y los costes de inversión en que se haya incurrido, al efecto de conseguir unas tasas de rentabilidad razonables con referencia al coste del dinero en el mercado de capitales.”49
47 Accordingly, I cannot agree with the Majority’s view that the Claimant’s current position (it expected stability, not immutability) has a bearing on the existence and extent of commitments made in RD 661/2007. See Decision, ¶ 324 and n. 430. ↩
48 C-0004 or R-0003, Arts. 16.1 and 30. ↩
49 Id., Art. 30.4. (My literal translation of this passage is as follows: “In addition, the production of electric energy by means of non-hydraulic renewable energies [...] shall receive a premium which shall be fixed by the Government in such a way that the price of electricity sold by these installations fall within a percentage band between 80 and 90 percent of an average electricity price, which shall be calculated dividing the revenues derived from invoicing for the supply of electricity by the electricity supplied. [...] To determine the premiums, there shall be taken into account the voltage level at which the energy is delivered to the network, the effective contribution to the improvement of the ↩
[Page 22]
57. The Majority takes the view that the second transcribed paragraph, which it quotes in a non-literal translation, states a “coherent general principle,” which “imposes some limits on what can be done.”50 The Majority acknowledges that Law 54/1997 empowered the administration to establish the premium by regulation, but argues that “[t]he stream cannot rise higher than its source or commit the state to more than the legislative framework allows.”51 On this basis, the Majority concludes that Article 30.4 of Law 54/1997 is inconsistent with the thesis that particular Royal Decrees, notably RD 661/2007, stabilized the regime.52
58. The metaphor that “the stream cannot rise higher than its source” and the reference to what “the legislative framework allows” suggest that the Majority considers a regulation such as RD 661/2007 incapable of setting forth a commitment unless such a commitment has been expressly authorized by the statute. The Majority’s legal theory appears to be, then, that for the purposes of expressing a commitment, any provision of RD 661/2007 (or any other Royal Decree issued under Law 54/1997) that purported to “stabilize” the regime (i.e. to give any aspect of it a specified duration) was ultra vires and therefore either invalid under Spanish law or otherwise to be disregarded. This theory, to be addressed in the following paragraphs, appears to be different from a related argument, to be discussed later in connection with the Claimant’ expectations, that those provisions of RD 661/2007 could not have given rise to “legitimate” expectations, because they were subject to being amended or repealed by later Royal Decrees or other norms of a higher rank, as in fact they were.53
59. In my view, the Majority’s theory is based on a misreading of the relevant legal texts and a flawed reasoning based on that misreading. To show why, we must first examine Article
environment, to the savings of primary energy, and to energy efficiency, and the investment costs incurred, with the purpose of achieving reasonable rates of return with reference to the cost of money in the capital market.”)
[Page 23]
30.4 of Law 54/1997 closely and then compare it with the provisions of RD 661/2007. A close examination of Article 30.4 leads to the following observations.
60. The first observation is an obvious point, but its consequences are often overlooked. Article 30.4 established norms of two different kinds. First, it provided that the producers were to receive a premium over the price for the energy sold, which means that it gave the producers a right to such a premium. This was a primary norm (or norm of conduct), that is, the kind of norm that creates rights and imposes obligations on persons subject thereto. The right created by that primary norm was, however, the right to a premium, not to a particular, specified premium. For the creation of the latter, Article 30.4 delegated on the Government the setting of the premium, while specifying certain criteria that the Government had to take into account in fulfilling that task. This was a secondary norm (or norm of competence), that is, a norm that empowered another organ of the system to set the premium, subject to those criteria.
61. The criteria which the Government had to apply in setting the premium included various factors to be considered (voltage level on delivery to the network, contribution to the improvement of the environment, contribution to savings of primary energy, contribution to energy efficiency, and cost incurred in the investment) and an overall objective: “to achieve reasonable rates of return with reference to the cost of money in the capital market.” These criteria were not addressed to the producers; they were directed to the Government’s conduct in setting the premium. Further, applying these criteria was not a mathematical exercise; it involved an exercise of judgment and a weighing of competing factors within a complex process of technical evaluation, quantification, and assessment of how eolic energy furthered certain economic and policy goals. Accordingly, setting the premium was an act of governmental rulemaking; it was not a mere deductive exercise from the content of Article 30.4.
62. For these reasons alone, it cannot be right to consider, as the Majority does, that any right (or “legitimate” expectation) that the producers had in respect of the premium had to be based solely on Article 30.4. Put differently, it is a mistake to regard Article 30.4 as the sum total of a producer’s rights under the Fourth Regime (or the sole source of an investor’s
[Page 24]
objectively reasonable expectations) and to treat the regulations thereunder as something that can safely be disregarded. In fact, the setting of the premium, as a separate governmental act, was an integral component of the legal regime established by the Law.
63. A second observation is that the criteria set forth in Article 30.4 concern only the quantum of the premium to be established by the Government. Law 54/1997 said nothing about the timing for setting the premium or the duration of the premium once set. In particular, the Law did not require the Government to establish that premium annually (as Law 40/1994 had done) or to establish or revise it at any time or at any particular intervals. Nor did Law 54/1997 debar the Government from designating a particular duration for the premium or for any element of the calculation. Because by necessity the setting of the premium had to happen at some time and the premium had to have some duration, it must be concluded that timing and duration are matters which the legislature deliberately left open, to be determined by the Government in the exercise of its regulatory function.
64. A third observation is that Article 30.4 set forth the criteria that the Government had to take into account in setting the premium. It did not require that the premium, once set, had to conform to those criteria at all times. Any such requirement would have been unworkable, of course, because some components of the criteria vary continually over time. It must be understood, then, that the criteria of Article 30.4 were to be applied in setting the premium for the first time and in any later revisions. But since Article 30.4 did not require that the premium be revised at any particular intervals, it did not exclude the possibility that the premium might be established (on the basis of the designated criteria) as a variable premium to be adjusted in accordance with an index.
65. A fourth, and most important, observation is that, under the criteria set forth in Article 30.4, “achiev[ing] reasonable rates of return with reference to the cost of money in the capital market” was the overall objective of the process for setting the amount of the premium.54
54 C-0004 or R-0003, Law 54/1997, Art. 30.4 (“Para la determinación de las primas se tendrá en cuenta el nivel de tensión de entrega de la energía a la red, la contribución efectiva a la mejora del medio ambiente, el ahorro de energía primaria y a la eficiencia energética, y los costes de inversión en que se haya incurrido, al efecto de conseguir unas tasas de rentabilidad razonables con referencia al coste del dinero en el mercado de capitales.”) (emphasis added). ↩
[Page 25]
It was not an overall target imposed by the statute on the actual financial results of any producer in the special regime. Therefore, there is no basis in Article 30.4, or elsewhere in the statute, for the contention that the overall objective of reasonable rates of return referred to in the statute worked as a ceiling (or a floor, for that matter) on the actual rate of return that a producer was entitled or permitted to have (or could reasonably expect) in any particular period or over the entire duration of the investment. In this respect, the tribunal in Cube correctly read Article 30.4, and the Majority’s disagreement with that decision cannot be based on the actual words of the statute.55
66. The other principal component of the Fourth Regime is Royal Decree 661/2007. This Royal Decree set forth a comprehensive, highly detailed regulation for the generation of electrical energy from a variety of sources, including wind. The producers of electricity from eolic energy (group b.2) had the right to choose between receiving (i) a regulated tariff or (ii) a market or negotiated price plus a premium.56 The alternative selected had to be in effect for at least one year.57 As for the first alternative, RD 661/2007 established that eolic plants installed on land (group b.2.1) would be entitled to a regulated tariff of 7.3228 c€/Kwh for the first 20 years and of 6.1200 c€/Kwh thereafter (“a partir de entonces”).58 As for the second alternative, the Royal Decree established a variable premium, which depended, in part, on the “reference market price,” a concept defined in the decree.59 The actual premium to be applied would be a “reference premium,” but this premium, added to the reference market price, could not exceed an upper limit or fall below
55 RL-0090, Cube Infrastructure Fund SICAV and others v. Kingdom of Spain (ICSID Case No. ARB/15/20), Decision on Jurisdiction, Liability and Partial Decision on Quantum, 19 February 2019 (henceforth, Cube), ¶ 293 (“We do not consider that the references in RD 661/2007 to a “reasonable return” were intended to have any application outside the context of reviews of the tariffs and of the upper and lower limits under Article 44.3 RD 661/2007. In particular, we do not consider that the references to a “reasonable return” signified a limit on the profit that a producer could earn from any power facility or group of facilities without suffering a reduction or lower-than-normal increase in tariffs, or that the references provided any basis for changes to the 2007 Regime outside the mechanisms set out in RD 661/2007.”); Decision, ¶ 332 (disagreeing with the quoted statement in Cube, also quoted in footnote, without referring to the text of Article 30.4 of Law 54/1997). ↩
56 C-0008 or R-0101, RD 661/2007, Art. 24. ↩
[Page 26]
a lower limit set forth in the decree.60 For eolic plants installed on land (group b.2.b), RD 661/2007 established, for the first 20 years, a reference premium of 2.9291 c€/Kwh, an upper limit of 8.4944 c€/Kwh, and a lower limit of 7.1275 c€/Kwh.61 After the first 20 years, the reference premium would be zero.62 These amounts would be adjusted annually according to any increase in the national consumer-price index (IPC) minus (i) 25 basis points until 31 December 2012 or (ii) 50 basis points thereafter.63
67. Article 44.3 of RD 661/2007 regulated the revision of the regime described above. It provided that the tariffs, premiums, upper and lower limits, and other add-ons would be revised in 2010 and thereafter every four years.64 Nevertheless, it provided that a revision of the regulated tariff or the upper and lower limits (but not including the reference premium) would not affect existing plants:
“Las revisiones a las que se refiere este apartado de la tarifa regulada y de los límites superior e inferior no afectarán a las instalaciones cuya puesta en servicio se hubiera otorgado antes del 1 de enero del segundo año posterior al año en que se haya efectuado la revisión.”65
68. Royal Decree 661/2007 provided a transition regime for existing eolic plants operating under the Third Regime. The producer had to make an irrevocable choice between the two options granted by the preceding regime: (i) selling electricity to the distribution company at a regulated tariff or (ii) selling it to the market at the market price plus a premium.66 If the producer chose the first option, the regulated tariffs of RD 661/2007 would not apply.67 If the producer chose the second option, the premium established under the Third Regime
63 Id., Art. 44.1 and First Additional Provision. ↩
64 Id., Art. 44.3. (My literal translation is as follows: “The revisions which are referred to in this subsection of the regulated tariff and of the upper and lower limits shall not affect the installations the commissioning of which shall have been granted before the 1st of January of the second year following the year in which the revision shall have been made.”) ↩
66 Id., First Transitory Provision (1). ↩
[Page 27]
would continue to apply for a transition period ending on 31 December 2012.68 In either case, the producer could opt into the regime of RD 661/2007, in which case that regime would apply in toto.69
69. The preceding review leads to the following conclusions, based on the plain texts of Law 54/1997 and RD 661/2007. Law 54/1997 (i) set forth a commitment that the relevant producers would receive a premium over the market price of the energy sold, and (ii) left to the Government the setting of the premium, subject to certain specified criteria on quantum. Royal Decree 661/2007 “developed” the regime of the law by setting forth a commitment to a remuneration regime comprising two alternatives, at the producer’s choice: (i) a regulated tariff or rate (which would have taken into account the premium, to make it a realistic choice) or (ii) a variable premium calculated on the basis of quantified factors. Royal Decree 661/2007 also set forth a commitment on the duration of each alternative and a commitment that future revisions of the regulated tariff and the upper and lower limits (but not including the regulated premium) would not affect existing plants.70
70. We can now assess the Majority’s view that the duration and stability commitments set forth by RD 661/2007 should be disregarded to the extent not expressly contemplated in Law 54/1997. As the Majority’s position could be interpreted in two different ways, I shall consider both alternatives.
71. First, the Majority’s position could be that those commitments in RD 661/2007 were invalid ab initio as a matter of Spanish law. Any such contention would be unsustainable, for various reasons. The Respondent does not claim that RD 661/2007 was, in whole or in part, invalid ab initio as inconsistent with Law 54/1997; it argues instead that RD 661/2007 was validly superseded by later norms of the same or higher rank.71 Nor is there any
69 Id., First Transitory Provision (3). ↩
70 To the same effect, see RL-0093, OperaFund Eco-Invest SICAV PLC and Schwab Holding AG v. Kingdom of Spain (ICSID Case No. ARB/15/36), Award, 6 September 2019, ¶ 485 (henceforth, OperaFund) (“[I]t is hard to imagine a more explicit stabilization assurance than the one mentioned in Article 44(3)”). ↩
71 Decision, ¶¶ 299-301 (summarizing Respondent’s arguments, centred on the possibility and validity of regulatory change). The Respondent argues separately that RD 661/2007 was invalid under EU law because it was part of a subsidies regime that was not notified to the EC. That contention is analysed infra, ¶ 125. ↩
[Page 28]
evidence on the record that the Spanish courts declared RD 661/2007 null and void ab initio as contrary to Law 54/1997. It is also highly doubtful that the Tribunal might have jurisdiction to decide, as a matter of Spanish law, that the presumption of legality of RD 661/2007 must be overridden and certain portions of the decree be declared ultra vires and invalid ab initio.
72. Second, the Majority’s position could be that, as a matter of interpretation and application of the fair-and-equitable-treatment standard of the ECT, a commitment made in a Royal Decree should be disregarded if not expressly authorized by the respective statute. The Majority offers no basis for any such contention, other than the metaphor that “the stream cannot rise higher than its source.”72 The metaphor is inapposite, however, because the source of the Government’s power to regulate is Article 97 of the Spanish Constitution, not (or not only) the relevant statute.73 In the Spanish system, in which a statute like Law 54/1997 provides a general framework to be supplemented or “developed” by detailed regulations, it is unjustified, as well as unrealistic, to expect that every aspect of the regulations will be foreseen and provided for in the statute. To be sure, in the Spanish system the regulations may not contradict the provisions of the statute, but the preceding analysis has shown that there is no inconsistency between the commitments made in RD 661/2007 and the provisions of Law 54/1997.
73. Therefore, I perceive no valid reason for an arbitral tribunal to disregard the commitments set forth in RD 661/2007. It remains to be determined whether, and to what extent, the commitments made in the Fourth Regime are sufficiently specific.
74. The specificity of a commitment is quintessentially a matter of degree. The degree of specificity of a commitment is one of the circumstances that should be taken into account
73 Constitución Española, Art. 97 (“El Gobierno [...] [e]jerce la función ejecutiva y la potestad reglamentaria de acuerdo con la Constitución y las leyes.”), http://www.senado.es/web/conocersenado/normas/constitucion/index.html. My literal translation: “The Government [...] exercises the executive function and the regulatory power in accordance with the Constitution and the laws.” ↩
[Page 29]
in assessing whether a particular instance of treatment meets the fair-and-equitable-treatment standard. Other things being equal, the less specific the commitment, the more difficult it would be to find that any expectations based thereon were objectively reasonable and, ultimately, that a subsequent measure breaching that commitment amounted to unfair and inequitable treatment.
75. By contrast, the Blusun rule and other similar rules treat the specificity of a commitment as an all-or-nothing factor, working as a diverting valve which, depending on the way it is set, leads the analysis through one conduit or another. In fact, the Blusun rule is a default rule, which purports to apply “[i]n the absence of a specific commitment.”74 The Majority does not explain what kind or degree of specificity would be sufficient to make a commitment “specific.” Nor does the Majority explain what legal consequences would follow from a “specific commitment,” although it can be surmised that the breach of such a commitment would amount to unfair and inequitable treatment, unless perhaps other (unstated) factors might change that result. In any event, the Majority seems to take the view that the Blusun rule comes into play because the commitments of RD 661/2007 should be disregarded or treated as non-existent, not because they were not sufficiently specific.75 Since I have concluded that those commitments did exist and cannot be disregarded, the question of specificity becomes critical even for the purposes of the Blusun rule and other similar approaches that turn on the existence of specific commitments.
76. In this regard, it cannot be denied that the commitments set forth in RD 661/2007 and hence those contained in the Fourth Regime as a whole are highly specific. They are highly specific in respect of the content of the commitments: the quantum of the remuneration to which the producers were entitled, the duration of the specified quantum, and the exemption of existing plants from future revisions to certain key components of the
74 Decision, ¶ 315, quoting from Blusun, ¶ 319(5). ↩
75 Decision, ¶ 321. This leads the Majority to the odd result that the highly specific commitments of RD 661/2007 are disregarded, while the much less specific criteria of Article 30.4 of Law 54/1997 are treated as the state’s sole relevant antecedent conduct. ↩
[Page 30]
specified remuneration.76 Those commitments are also highly specific in respect of the beneficiaries of the commitments, i.e. the class of producers validly registered in the RAIPRE as beneficiaries of the special regime.77
77. The preceding review of the Respondent’s relevant antecedent conduct leads to the following conclusion: by adopting the Fourth Regime comprising Law 54/1997 and RD 661/2007, the Respondent made highly specific commitments to the producers of electricity from eolic energy in respect of the remuneration to which the producers would be entitled, including the quantum of the remuneration, the duration thereof, and the exemption of existing plants from any future changes affecting certain components of the remuneration.
78. The relevant circumstances related to the Claimant’s conduct comprise the Claimant’s reactions (including its objectively reasonable expectations) in respect of the legal regime existing at the time of each investment and any subsequent changes to that regime. The Claimant’s objectively reasonable expectations in respect of those legal regimes and any commitments they contained are important consideranda in a proper analysis under the standard of fair and equitable treatment.
79. I should start with an explanation of the term “objectively reasonable expectations.” I am not aware of any substantive difference between what the Majority means by “legitimate expectations” and what I mean by “objectively reasonable expectations.” But I prefer to use the latter term, which is more descriptive, and avoid the ubiquitous, overused, ambiguous adjective “legitimate.”78 The concept, which I believe to be generally accepted
76 To the same effect, see CL-0112, Novenergia II – Energy & Environment (SCA) (Grand Duchy of Luxembourg), SICAR v. Kingdom of Spain (SCC Arbitration 2015/063), Award, 15 February 2018 (henceforth, Novenergia II), ¶¶ 665-667 (“The commitment from the Kingdom of Spain could not have been clearer”); Haigh Dissent, ¶ 10. ↩
77 For similar conclusions, see, e.g., CL-0117, 9REN Holding S.à.r.l. v. Kingdom of Spain (ICSID Case No. ARB/15/15), Award, 31 May 2019, ¶¶ 257, 295; CL-0113, Masdar Solar & Wind Cooperatief U.A. v. Kingdom of Spain (ICSID Case No. ARB/14/1), Award, 16 May 2018, ¶ 512; Haigh Dissent, ¶¶ 11-12, 50-52. ↩
78 When we use the term “legitimate,” we do so by reference to an underlying criterion of legitimacy, which may be made explicit or not. In the absence of an explicit reference to a criterion of legitimacy, it is not clear whether the underlying criterion is a moral or religious code, or a legal system, or a political theory, or a particular custom, or a ↩
[Page 31]
regardless of the terminology used, is that an investor’s expectations are relevant to the fair-and-equitable-treatment analysis only to the extent that they are reasonable from an objective standpoint, that is, the standpoint of a reasonable investor in the circumstances.79
80. The critical issues to be considered in this subsection are: (i) whether the relevant expectations of the Claimant are limited to those which arose at the time of the initial investment or may also include those which arose in connection with each change of legal regime; and (ii) whether expectations based on commitments set forth in Royal Decrees were objectively reasonable.
81. The Majority takes the view that “as a matter of general principle, a legitimate expectation is a form of reliance interest which must relate to facts or circumstances in existence at the time the investment is made.”80 This view is correct when applied to the usual case, in which the legal regime existing at the time of the investment and the expectations derived therefrom are abrogated or destroyed by the subsequent measures on which the claim is based. But in a case of a long-term investment which has been subjected to successive intermediate legal regimes, it is unrealistic and artificial to limit consideration of the investor’s objectively reasonable expectations to those held at the beginning of the process, based on a legal framework that may have long been repealed by the time the questioned measures were imposed. As already noted, whether a particular instance of treatment is fair and equitable depends on all the relevant circumstances, not on a subset of them selected by a rigid arbitrator-made rule.
82. The Majority rejects the possibility that “legitimate” expectations might change in the course of the investment on two grounds: (i) those expectations need not always change in the direction more favourable to the investor, and (ii) if the expectations can change in both
dynastic principle, etc. Leaving the reference open carries the risk of a later redefinition of the term through switching the criterion of reference.
79 See CL-0023, Ioan Micula and others v. Romania (ICSID Case No. ARB/05/20), Final Award, 11 December 2013 (henceforth, Micula), ¶ 669 (“Whether a state has created a legitimate expectation in an investor is thus a factual assessment which must be undertaken in consideration of all the surrounding circumstances.”) ↩
[Page 32]
directions, they would simply “track the state of the law from time to time, and the idea of a reliance interest would lose all autonomy.”81 The first observation is of course correct but the second is not. The investor’s expectations need not track the state of the law, because they depend on the investor’s attitude and decisions concerning each change in the law, a point to which I shall presently return. Further, the supposed “autonomy” of the notion of reliance cannot mean that reliance can be predicated only in reference to the state of affairs at the time of the initial investment. An investor can rely on something (such as a state commitment) for its initial decision to invest, but that reliance does not prevent the investor from relying on something else (such as another state commitment) for a subsequent decision concerning the investment.82 As will be seen, long-term investments like those in the present case present a more complex reality, which a proper analysis must take into account.
83. It is undisputed that the individual wind farms on which the Claimant invested were approved and commissioned at different times, under various legal regimes.83 Depending on the date of approval or start of operations, each individual wind farm in the Claimant’s inventory was subject to one, two, three, or four successive legal regimes in the period preceding the Disputed Measures.84 To simplify the analysis, let us take the three oldest plants (Pax I and II A, Barbanza, and Vicedo), which were approved under the First Regime, although they came into operation under the Second.85 Progressively simpler
82 See CL-0046, C. Schreuer and U. Kriebaum, “At What Time Must Legitimate Expectations Exist?” in J. Werner and A. Hyder Ali (eds.), A Liber Amicorum: Thomas Wälde (2009), pp. 273-274 (“A foreign investor may be presumed to know the general regulatory framework prevalent in a country at the time it first embarks upon the investment. But it is not only the framework existing at that early stage that can create legitimate expectations. If there are favourable changes to the legal framework during the establishment or during the lifetime of the investment, this may also create legitimate expectations which will be protected if the foreign investor relies on them in subsequent business decision [sic].”) The same reasoning should apply to unfavourable changes to the legal framework in which the investor acquiesces. One possible exception, depending on the circumstances, may be “creeping” detrimental changes which may not individually justify a challenge but in their combined effect may amount to unfair and inequitable treatment. This hypothetical situation, which is analogous with that of “creeping” expropriation, does not correspond to the facts of the present case. ↩
83 Claimant’s Memorial, Appendix II. ↩
84 Compare id. with the dates on which each successive regime took effect. ↩
85 Claimant’s Memorial, Appendix II. ↩
[Page 33]
versions of the same analysis will apply to the newer plants, which were subject to fewer changes in the applicable legal regime.
84. The Claimant’s expectations at the time of its initial investment in the three oldest plants derived from the First Regime (Law 40/1994 and RD 2366/1994), which was in force at the time. That legal regime was later superseded by the Second Regime (Law 54/1997 and RD 2818/1998), which was superseded by the Third Regime (Law 54/1997 and RD 436/2004), which was superseded by the Fourth Regime (Law 54/1997 and RD 661/2007), which was amended, in a relatively minor way, by RD 1614/2010. Each time the legal regime was changed, the investor faced a business decision, which was informed, in part, by (objectively reasonable) expectations concerning the treatment to be afforded by the new regime and, in particular, by the revenues to be expected from the level of subsidies to be paid thereunder.
85. Such a business decision involved, in a very real sense, the acceptance or non-acceptance of the new regime, a decision based, at least in part, on the (objectively reasonable) expectations derived from that regime. If the investor was given the choice of remaining under the old regime or to opt into the new one (as the Second and Third Regimes and to some extent the Fourth Regime did),86 the investor had to make that choice, and if it opted to come under the new regime (whether more or less favourable to its interests), it cannot seriously be denied that its expectations going forward would be those derived from the new regime. On the contrary, if the Government imposed the new regime on existing investors (as it did generally in the case of RD 661/2007 and fully in the case of RD
86 C-0004 or R-0003, Law 54/1997, Eighth Transitory Provision (producers under the preceding regime would continue to be subject thereto, for as long as a certain compensatory scheme for certain other producers remained in effect, but could opt into the remuneration regime of the new law); C-0005, RD 2818/1998, First Transitory Provision (producers under the preceding regime would continue to be subject thereto, but could opt into the remuneration regime of the decree); C-0007, RD 436/2004, Second Transitory Provision (producers under the regime of RD 2366/1994 [First Regime] would be generally exempt from the regime of the new decree, except that they would have the right to sell energy to the market and could opt into the entire regime of the new decree; producers under the regime of RD 2818/1998 [Second Regime] would continue to be subject to the previous remuneration regime for a transition period of about 6 years, but could opt into the regime of the new decree.); C-0008 or R-0101, RD 661/2007, First Transitory Provision (producers under the regime of RD 436/2004 would continue to generally subject to that regime for a transition period but could opt into the new regime; in all other respects, the new regime generally replaced the previous one, which was repealed.) ↩
[Page 34]
1614/2010)87, the investor had to decide whether to acquiesce in the new regime, or to contest it under the local law or under the ECT, or to sell the investment. In making that business decision, the investor would have to take into account a variety of factors, including the expectations derived from the new regime as compared with the expectations created by the old one. If for any reason the investor decided to keep the investment and to acquiesce in the new regime, it is unrealistic to ignore the expectations created by the new regime and unreasonable to insist that the sole expectations that count are those generated by the previous, no longer applicable regime.
86. The Claimant’s description of the facts, its insistence in the unity of its investments, and its reliance on the Fourth Regime, portrayed as the apex of the evolution of a unitary system of incentives provided to the eolic-energy sector in Spain – all indicate that the Claimant accepted every legal regime to which its investments were successively subjected, until the Disputed Measures.88 Whether the Claimant regarded each new regime as more or less favourable than its predecessor is immaterial, because it appears that the Claimant did not contest the new regime or sell any of its investments. Notably, the Claimant appears to have embraced the Fourth Regime and the amendment of RD 1614/2010 unreservedly, even though they were generally imposed on its existing investments nolens volens.89 Therefore, it must be concluded that at the time the Claimant ceased investing in Spain, it had acquiesced in the application of the Fourth Regime to all of its investments and had adjusted its expectations to the terms of that regime.
87 C-0008 or R-0101, RD 661/2007, First Transitory Provision (see preceding note); C-0010, RD 1614/2010 (Article 2 imposed reference limits on the number of hours of production for which the producer would be entitled to the premium or premium equivalent. Article 5 generally reduced, for a designated period, the reference premium applicable to certain installations under RD 661/2007.) ↩
88 See, e.g., Claimant’s Memorial, passim (treating the “Incentives Regime” as a unit); ¶¶ 46, 216 (RD 661/2007 “final update to the Incentives Regime”), 197 (investment should be considered as an “indivisible whole”), 204 (expectation that the Claimant’s projects would continue to benefit from a regime at least as favourable as that of RD 661/2007). ↩
89 The Fourth Regime was imposed on existing projects after a transition period. CW-5, Supplemental Witness Statement of Tetsuya Suwabe, ¶ 7 (“Although we opted to continue to receive remuneration under the scheme set out in Royal Decree 436/2004 for all the projects (except for the Grallas project) that were entitled to do so, the further confirmation of stability provided by Royal Decree 661/2007 directly contributed to Eurus’ ongoing decision to stay in Spain. We were particularly reassured by the fact that Eurus’ wind power facilities that chose to remain under the 2004 regime would later be entitled to the incentives regime defined by Royal Decree 661/2007, once the latter’s transitional period was over. As a result, Eurus did not pursue or consider a sale of its interests in any of the SPCs.”). ↩
[Page 35]
87. It is clear that, in the end, the Claimant’s expectations were based on the Fourth Regime, as amended by RD 1614/2010.90 Yet it is also clear, by the Claimant’s own admission, that its expectations did not exactly coincide with the terms of that regime. The Claimant appears to have expected that in the future its investments would be subject to conditions not significantly worse than those set forth in the Fourth Regime, as amended by RD 1614/2010.91 Further, the Claimant has stated that it did not expect that regime to be “frozen” or “petrified” but that any subsequent change would be either more favourable to its interests or at least would not destroy its business model without compensation.92 Therefore, the Claimant appears to have expected that the amended Fourth Regime would be stable, as distinguished from immutable.
88. The next question is whether the Claimant’s expectations based on the Fourth Regime can be considered to be objectively reasonable. The Majority considers that no “legitimate” expectations could have derived from RD 661/2007, for two basic reasons: (i) that such expectations must be based on the legal regime that existed at the time of the investment, which for all but one of the Claimant’s plants was an earlier regime;93 and (ii) that the regime of RD 661/2007 was subject to change within the framework of Law 54/1997, and there could have been no “legitimate” expectation that existing plants would be grandfathered.94 The Majority’s first reason does not apply in the complex circumstances of this case, as explained in the preceding subsection.95 The Majority’s second reason,
90 See, e.g., Claimant’s Memorial, ¶¶ 204, 213-214; Claimant’s Reply, ¶¶ 239-240. ↩
92 See, e.g., Claimant’s Reply, ¶ 240; Decision, n. 422. ↩
94 Decision, ¶¶ 332-333 (citing a 2005 decision of the Spanish Supreme Court to that effect). The Majority also states that regulations “[had] change[d] and not in any uniform direction favouring the recipients.” Id., ¶ 331. The Majority also mentions in this context that “the stream cannot rise higher than its source or commit the state to more than the legislative framework allows.” Decision, ¶ 334. As already discussed, this argument is more properly analysed as one against the existence of specific commitments arising out of RD 661/2007. For the reasons already explained, this argument is invalid. Supra, ¶¶ 75-77. ↩
[Page 36]
which can be referred to as the “revocability argument,” broadly coincides with the Respondent’s position on this matter. It will be discussed in the following paragraphs.
89. There is no doubt that, as a matter of Spanish law, the Respondent had the general legal power or authority to amend or repeal, through the appropriate organs and procedures, each component of the Fourth Regime, in particular Law 54/1997 and RD 661/2007. Moreover, the Claimant was aware of the existence of that general power96 or at any rate should have been aware of it, because it concerned a fundamental feature of the host country’s legal system, not an obscure rule or unsettled point of law. What appears to have been unsettled at first was whether the exercise of such power would require compensation to injured producers,97 but in late 2005 the Supreme Court issued a decision from which it could be inferred that a change in the regulations within the framework of Law 54/1997 would not give affected producers a right to compensation under Spanish law.98
90. The question is, however, whether as a matter of application of the standards of the ECT, the mere existence of that general power under domestic law is enough to foreclose any objectively reasonable expectations based on a state commitment that such power will not be exercised in certain respects and for a certain period. The answer must be negative, for the following reasons:
a. First, to say that the Government had or retained its general power under Spanish law to amend or repeal RD 661/2007 is to say that the commitments set forth in that Royal Decree were not binding on the Government as a matter of Spanish law. But a state commitment need not be binding under domestic law to be one of the circumstances to be considered in a fair-and-equitable-treatment analysis. Fair and equitable treatment is not limited to compliance with legal obligations under domestic law. Consequently, even though the commitments made in RD 661/2007
96 See C-0049, BHV Memorandum, October 2002, p. 77 (referring to an opinion of the Landwell law firm to the effect that Spanish legislation gave the producers of renewable energy certain rights, which constituted acquired rights, so that any reversal of such rights would obligate the Government to provide compensation) (emphasis added). ↩
98 Id.; see RL-0137, 2005 Judgment (producers do not have an unmodifiable right that the specific system of remuneration will stay the same). ↩
[Page 37]
were revocable as a matter of Spanish law, that revocability in itself does not foreclose the possibility that they might have created objectively reasonable expectations.
b. Second, from the standpoint of a reasonable investor trying to ascertain the Respondent’s intent in making the commitments set forth in RD 661/2007, only three alternatives are apparent: either (i) the state made those commitments with the intent of repudiating them at a later time; or (ii) the state made those commitments with the intent of complying with them initially, but with the (unstated) reservation that it may well change its mind in the future and repudiate them; or (iii) the state made those commitments with the intent of complying with them in accordance with their terms. The first alternative must be discarded; it would be an instance of bad faith, which cannot be presumed and has not been proved in this case. The second alternative must also be discarded; the principle of transparency would have required the state to disclose that it intended to honour the commitments unless and until it decided otherwise – a disclosure which would have deprived the commitments of all value as investment incentives. The sole remaining alternative is the third one. Therefore, a reasonable investor would have had reason to conclude that the state, presumed to act in good faith, intended to honour its commitments according to their terms, and to refrain from exercising its power to revoke them.99
c. Third, if objectively reasonable expectations could not have arisen from the commitments made in RD 661/2007 because the state had the general power to amend or repeal that decree, the same reasoning would apply to the (less specific) commitments made in Law 54/1997, because that law could have been (and in fact was) later repealed. In this sense, the logic of the revocability argument extends beyond RD 661/2007, to all norms of the Spanish legal system that are subject to
99 See RL-0090, Cube, ¶ 305 (“[A]fter careful reflection, being perfectly aware of the jurisprudence of their Supreme Court, the Spanish authorities decided that the Special Regime for the remuneration of renewable energy should be established for a long duration in order to attract sufficient investment that otherwise could not have been obtained. This was a deliberate decision from which the Respondent cannot move away lightly.”). ↩
[Page 38]
amendment or repeal. It is therefore inconsistent for the Majority and the Respondent to embrace the revocability argument while accepting that the Claimant could derive “legitimate” expectations from the reference to “reasonable rates of return” in Article 30.4 of Law 54/1997.100
d. Fourth, the theory that no objectively reasonable expectations can be based on revocable commitments contradicts common sense and the way in which real-world investors make long-term investment decisions. It follows from that theory that a “reasonable” investor would have to assume that the state would exercise its power to repudiate its commitments, notwithstanding grounds to believe that the state sincerely intended to honour those commitments. In other words, according to that theory, a “reasonable” investor must assume that any state commitment that is subject to change cannot be trusted and should be ignored. In the real world, however, governments offer incentives to investors for the intended purpose of being relied upon, and reasonable investors do not make decisions by automatically assuming that outcomes will always be the worst. (Otherwise, fewer investments would be made.) In particular, if real-world investors had to assume that state commitments such as those made in RD 661/2007 should be ignored, those commitments would never be able to achieve their intended purpose of providing incentives to long-term investors. An incentive on which no investor can rely is not an incentive.
91. It follows from the preceding discussion that, as a matter of application of the standards of the ECT, the revocability of a legal regime does not prevent that regime from giving rise to objectively reasonable expectations. It remains to determine whether the Claimant’s expectations to the stability of the amended Fourth Regime, including the commitments set forth in RD 661/2007, were in fact objectively reasonable. As on all matters concerning the standard of reasonableness, the answer depends on all the relevant circumstances.
100 Decision, ¶ 355; Id., ¶ 301 (summarizing the Respondent’s position). ↩
[Page 39]
92. As already discussed, a reasonable investor had good reasons to believe that the Respondent intended to honour its commitments according with their terms.101 Further, those commitments were made against the background of a consistent, long-standing state policy aimed at increasing the production of electricity from renewable sources, a policy consistent with EU directives and other international commitments of the host state.102 That policy had been implemented through a succession of legal regimes which generally maintained or increased the incentives provided to producers of energy from wind.103 Those successive regimes generally gave existing investors the choice of staying under the old regime or opting into the new one.104 One important exception was RD 661/2007, which was generally imposed on existing producers, but in that case grandfathering the existing remuneration regime would have perpetuated a design flaw, which was not a matter of policy but the perverse operation of an ill-chosen adjustment formula.105 Further, the Claimant's expectations were not that the Fourth Regime, as amended, would be immutable, but that it would be stable. In light of these circumstances, it must be concluded that the Claimant's expectations based the Fourth Regime, including the overall stability of the regime, were objectively reasonable.
93. The relevant circumstances related to the Disputed Measures concern the character of the measures and the effects they had on the preceding legal regime and the Claimant's objectively reasonable expectations based thereon. As already discussed, the commitment under the second sentence of Article 10(1) of the ECT to accord fair and equitable treatment is textually and contextually linked to the conditions referred to in the first sentence of the
101 Supra, ¶ 90. ↩
102 See generally, Decision, ¶¶ 97-119. ↩
103 See Id.; see CE-1, Brattle First Regulatory Report, ¶¶ 99-110. ↩
104 Supra, ¶ 85. ↩
105 Under the Third Regime, tariffs were adjusted by reference to the Tarifa Media o de Referencia, an index of the cost of electricity for final consumers in Spain. It turned out that this index was unsuitable because an increase in the generation of electricity from renewable sources would have led to higher premiums, which would have led to an increase in the index, which would have led to higher premiums, etc. See Respondent's Counter-Memorial, ¶ 473. RDL 7/2006 (BRR-101) suspended operation of the index and RD 661/2007 replaced it. C-0008 or R-0101, RD 661/2007, Second Transitory Provision. ↩
[Page 40]
same provision.106 Accordingly, an analysis of the relevant circumstances related to the Disputed Measures should include: (i) the degree to which the measures were favourable and transparent; (ii) the extent to which they affected the preceding legal regime and the degree to which the resulting treatment conditions can be considered stable; and (iii) the extent to which they repudiated specific commitments or frustrated the Claimant's objectively reasonable expectations based on the preceding regime.
94. As already explained, I perceive no legal basis for applying the Blusun rule, either as purported interpretation of the fair-and-equitable-treatment standard of Article 10(1) of the ECT or as a substitute therefor.107 Nevertheless, as part of the discussion in this subsection I shall critically examine the three conditions, factors, or criteria which would come into play under the Blusun rule in the assumed absence of a special commitment: (i) whether it became “necessary” to modify the subsidies granted by the preceding regime; (ii) whether the modification was done in a manner “not disproportionate to the aim of the legislative amendment"; and (iii) whether the modification was made with “due regard to the reasonable reliance interests of recipients who may have committed substantial resources on the basis of the earlier regime.”108
95. The Disputed Measures were undoubtedly less favourable to the producers of electricity from wind than was the preceding legal regime. The Claimant argues that the Respondent failed to provide favourable conditions for the investment.109 The question is, however, whether the concept of favourability referred to in the first sentence of Article 10(1) of the ECT (and relevant also to the second sentence) applies to treatment consisting of granting and modifying subsidies. Subsidies are indeed favourable to the recipient, but they differ fundamentally from the ordinary conditions that states create for investments; they are extraordinary, ultra-favourable benefits which states grant in exceptional circumstances to
106 Supra, ¶ 24. ↩
107 Supra, ¶ 32. ↩
108 Supra, ¶ 28. ↩
109 Claimant's Memorial, ¶¶ 206-212 and the immediately following heading. ↩
[Page 41]
sustain or promote particular economic activities. No provision of the ECT requires states to grant such extraordinary benefits. In particular, no such obligation appears in Article 19(1)(d) of the ECT, which merely requires Contracting Parties to “have particular regard [...] to developing and using renewable energy sources.”110
96. In light of this context and the extraordinary nature of subsidies, the term “favourable conditions" in Article 10(1) should be interpreted as referring to ordinary conditions for an investment, not including subsidies. For these reasons, the Claimant's argument on favourability should be rejected. Article 10(1) of the ECT did not require the Respondent to grant subsidies to the eolic-energy sector and, if it chose to grant such subsidies, a requirement that they be maintained cannot be derived from the favourability factor of Article 10(1).111
97. If the Disputed Measures are taken at face value, their prescriptive content and effects on the preceding regime are reasonable clear. But the Respondent's arguments in defence of those measures give rise to serious transparency issues, as shown by the preceding discussion and the one yet to come. For example, if it were true that the producers were entitled only to the “reasonable rates of return" referred to in Article 30.4 of Law 54/1997 and nothing more (a position that the Majority accepts), the entire regime would have been non-transparent, because it would have failed to explain (i) what rates of return the state considered reasonable or (ii) on what “cost of money in the capital markets” they were based or (iii) how future tariffs would reflect changes in (i) or (ii). Further, the Disputed Measures fail to explain why rates of return once considered reasonable are no longer so considered or what changes in the cost of money in the capital markets led the state to that conclusion. These transparency issues become less important if one rejects the false
110 CL-0001, ECT, Art. 19(1)(d). ↩
111 This conclusion refers exclusively to the favourability factor in the circumstances of this case and has no bearing on the other elements of the fair-and-equitable-treatment analysis. The Majority reaches a similar result, without discussing the favourability factor. See Decision, ¶ 317. ↩
[Page 42]
premise that the producer could have no rights or expectations other than those arising out of the reference to “reasonable rates of return” in Article 30.4 of Law 54/1997.112
98. As already discussed, the first sentence of Article 10(1) of the ECT requires that the state create "stable conditions" for the making of investments. In addition, for textual and contextual reasons the stability of the existing conditions is a relevant factor to be considered in the application of the fair-and-equitable-treatment standard. As noted, stability is not the same as immutability. The stability of a legal regime is consistent with changes that do not exceed a certain magnitude – a magnitude which is usually explained in terms of the importance or centrality of the changed elements and the effects and timing of the changes.113 Accordingly, we must review the extent of the changes that the Disputed Measures introduced in the preceding regime and then assess the magnitude of those changes and the resulting impact on the stability of that regime.
99. The Disputed Measures introduced the following principal changes in the preceding legal regime:
112 Supra, ¶ 90. ↩
113 For examples of different ways of describing the magnitude of a change that deprives the changed regime of stability see, e.g., RL-0093, OperaFund, ¶ 508 (essential characteristics of the regulatory regime relied upon by investors); CL-0116, Foresight Luxembourg Solar 1 S.à r.l. and others [Greentech] v. Kingdom of Spain (SCC Arbitration V (2015/150)), Final Award, 14 November 2018, ¶ 365 (fundamental and abrupt alteration), ¶¶ 397-398 (radical and unexpected, fundamental change); CL-0112, Novenergia II, ¶ 695 (radical and unexpected); CL-0079 or RL-0071, Eiser Infrastructure Limited and Energía Solar Luxembourg S.à r.l. v. Kingdom of Spain (ICSID Case No. ARB/13/36), Award, 4 May 2017, ¶ 382 (radical alteration that deprives investors who invested in reliance of the regime of the value of their investments), ¶ 387 (drastic and abrupt revision of the regime on which the investment depended in a way that destroyed its value); CL-0039 or RL-0050, Total S.A. v. Argentine Republic (ICSID Case No. ARB/04/01), Decision on Liability, 27 December 2010, ¶ 168 (no respect for certain basic features of the amended regime); CL-0023, Micula, ¶ 687 (formal shell of regime maintained but eviscerated of all or substantially all of its content); Hobér Dissent, ¶ 9 (fundamental and radical changes). ↩
[Page 43]
10-year Treasury bonds in the secondary market over the preceding ten years, plus 300 basis points.
[Page 44]
Measures provide that the specific remuneration would apply only to the "regulatory life” of the facility, which it sets at 20 years.114
100. The changes just described affected critical aspects of the preceding legal regime. The Claimant's expert on regulation characterized them as a “mid-stream switch in the regulatory paradigm" for existing plants.115 The preceding regime followed a paradigm based on a generally stable subsidized remuneration per MWh produced, which gave producers an incentive to build and manage projects efficiently, in the hope of achieving better financial results than those made possible by the subsidized remuneration, in exchange for taking the risk of selecting the sites, designs, and equipment. In contrast, the Disputed Measures adopted a paradigm based on imputed (not actual) costs (even for plants already built under the old regime) and a designated pre-tax target rate of return based on such imputed costs, a rate of return which operated as a ceiling on the producer's actual financial results. This change of paradigm has had the effect of depriving efficient producers of the rewards that they accrued under the previous system in exchange for the risks taken. The point here is not whether one paradigm is better than the other or the switch in mid-stream was economically efficient or inefficient. The point is the magnitude of the change, and the change was indeed radical. Consequently, the stability of the preceding legal regime was severely impaired.
101. As discussed, RD 661/2007 set forth certain specific commitments regarding the remuneration regime for producers of electricity from wind and the duration of that regime.116 By repealing and replacing the entire regime, the Disputed Measures abrogated those commitments.
114 C-0018, MO IET/1045/2014 of 16 June 2014, Art. 5. ↩
115 CE-1, Brattle's First Regulatory Report, ¶¶ 154-183. ↩
116 Supra, ¶ 77. ↩
[Page 45]
102. I have also concluded that the Claimant derived objectively reasonable expectations to the stability of the Fourth Regime, as amended.117 By radically impairing the stability of the preceding regime, the Disputed Measures frustrated those expectations.
103. The Blusun rule sets forth three criteria that the state must satisfy to modify subsidies previously granted, in the absence of a specific commitment to keep them unmodified: (i) the modification must have become necessary; (ii) it must not be disproportionate to the aim pursued; and (iii) it should have due regard to the reasonable reliance interests of recipients who may have committed substantial resources on the basis of the earlier regime.118
104. The first criterion is necessity (“if it becomes necessary to modify them”). The Majority does not discuss it in detail, but it plainly concludes that the modification of the existing subsidies regime had become necessary to address the tariff deficit.119
105. Two observations must be made regarding the criterion of necessity. The first is that the "necessity” at issue here is not a plea of necessity under customary international law, as codified in Article 25 of the ILC's Articles on Responsibility of States for International Wrongful Acts, and therefore the requirements of the latter are not applicable.120 The second observation is that in applying this criterion, the Majority does not use “necessary” in a strict sense. Strictly speaking, to say that modifying the Fourth Regime was necessary to "address" the tariff deficit is to say that the modification was a necessary condition (conditio sine qua non) for achieving that result or, in other words, that no other measures would have solved or ameliorated the problem. This is certainly not the way in which the Majority applies this criterion. Not only does the Majority fail to show that the tariff deficit could not have been addressed without modifying the Fourth Regime – it expressly declines
117 Supra, ¶ 87. ↩
118 CL-0083 or RL-0074, Blusun, ¶ 319(5); Decision, ¶ 335. ↩
119 Decision, ¶¶ 337-338. ↩
120 CL-0057, Articles on Responsibility of States for International Wrongful Acts, Article 25. ↩
[Page 46]
to consider alternative measures.121 It appears, then, that in the Majority's view, the relation between the problem (the tariff deficit) and the means to address it (the modification of the subsidies) is not a matter subject to the Tribunal's review.
106. In this light, it appears that “if it becomes necessary to modify the subsidies” in fact means “if the state considers, in its discretion, desirable to modify the subsidies." This means that the "necessity" criterion (as the Majority applies it) has no limiting function, and hence it can have no meaningful role in the evaluation of a measure under the Blusun rule.
107. The second criterion in the Blusun rule is “proportionality to the aim of the legislative enactment."122 In this case, the Disputed Measures are the “legislative enactment" and the aim of the Disputed Measures is to “address” (presumably, to eliminate or reduce) the tariff deficit.123 Accordingly, to satisfy this criterion the Disputed Measures must have been proportionate to the aim of eliminating or reducing the tariff deficit. Yet, it is not entirely clear what the Blusun rule and the Majority understand by “proportionality.” It appears that, as a starting point, the measure must be suitable to achieve the aim, but it is not self-evident at what point a suitable measure should be considered disproportionate to the aim. The way the Majority applies this criterion suggests that a suitable measure is disproportionate to the extent it is not necessary (perhaps stricto sensu) to achieve the aim, even if the absence of the measure should delay achievement of the aim.124
108. The idea of proportionality often plays a role in the application of the fair-and-equitable-treatment standard, either as a separate considerandum (or element of the standard) or as
121 Decision, ¶ 338 ("[I]t is not for the Tribunal to second guess reasonable measures taken to address the deficit (including measures affecting existing plants), to propose alternative policies that could have been adopted, or to weigh up for itself the competing demands of generators and consumers.") ↩
122 CL-0083 or RL-0074, Blusun, ¶ 319(5); Decision, ¶ 335 (“[The modification of the existing subsidies] should be done in a manner which is not disproportionate to the aim of the legislative amendment"). ↩
123 Decision, ¶ 338. ↩
124 In the Majority's analysis, all aspects of the Disputed Measures meet the criterion of proportionality, except for the claw-back feature. The claw-back feature is excluded because it "has not been shown to have been necessary to resolve the tariff deficit problem, which would have been solved in any event by the Disputed Measures without much further delay and without the element of claw-back." Decision, ¶ 355. It is not clear, however, why the same reasoning should not apply to the shortening of the relevant life of a plant, which the Majority regards as satisfying the criterion of proportionality. ↩
[Page 47]
part of the reasoning through which the various elements are assessed and weighed.125 In any case, the Majority's version of proportionality cannot meaningfully serve as an element in the fair-and-equitable-treatment analysis, much less as an overall control factor. Let us examine the characteristics of this criterion more closely. First, the aim of the measure is deemed to be unreviewable. Second, the Blusun proportionality relation is primarily an instrumental one – that of means to an end. Only to the extent that the means somehow overshoot the end are they considered disproportionate. Conceivably, a system of review of state measures focused on ends and means can combine either a strong or weak review of means with either a strong or weak review of ends. The proportionality criterion of the Blusun rule combines a weak review of means with no review of ends. The result is a very weak criterion, which puts no meaningful limits on the state's power to modify subsidies. More important, it is not apparent whether this weak version of proportionality bears any relationship to the ordinary meaning of “fair and equitable treatment.” I shall presently return to the way the Majority applies this criterion.
109. The third Blusun criterion is “due regard to the reasonable reliance interests of recipients [of the subsidies] who may have committed substantial resources on the basis of the earlier regime."126 The Majority does not explain what this criterion means, nor does the Majority appear to apply it independently of the proportionality criterion. The most obvious difficulty is the uncertain meaning of “due regard." This expression, common in diplomatic language and "soft" legal texts, normally refers to something that ought to happen in the process of taking a decision, not to the result of that process.127 This criterion appears to mean, then, that in modifying the subsidies the state ought to consider (or consider carefully) the reasonable reliance interests of the recipients of the subsidies.128 If this is so, then this third criterion can serve as an exhortation to the state about how it should act in the process of devising changes to the subsidies but not so much as a basis for a tribunal to
125 See, e.g., RL-0093, OperaFund, ¶ 555 (part of a balancing assessment). ↩
126 CL-0083 or RL-0074, Blusun, ¶ 319(5); Decision, ¶ 335. ↩
127 See, e.g., Art. 19(1)(d) of the ECT (requiring Contracting Parties to “have particular regard [...] to developing and using renewable energy sources"). ↩
128 It is also unclear whose reliance interests the state ought to consider. Depending on whether the relative sentence beginning with "who may have committed" is read as a qualification or an explanation, the class of the relevant recipients may or may not be restricted to those who committed substantial resources on the basis of the earlier regime. ↩
[Page 48]
review the objective results of that process.129 By contrast, the standard of Article 10(1), second sentence, concerns the fairness and equity of the actual treatment received by an investor, not (or not only) the fairness and equity of the process through which the state devised such treatment. Therefore, the third Blusun criterion is not really useful as a basis on which a tribunal can determine whether a particular measure constitutes fair and equitable treatment.
110. The Majority applies the Blusun rule in a piecemeal manner, to various components of the Disputed Measures,130 an approach which implicitly discards assessing the overall stability of the regime. In so doing, the Majority finds fault with only two aspects of the Disputed Measures: (i) the claw-back feature and (ii) the shortened “regulatory” life assigned to the existing installations. The Majority adjudges the claw-back feature to be inconsistent with the principle of stability and to fail the Blusun criterion of proportionality.131 As for the life assigned to the existing installations, the Majority considers it to be unreasonable, but this presumptive violation is in the end erased after applying an overall control of proportionality.132
111. Nevertheless, when it comes to applying the overall control of proportionality to all aspects of the Disputed Measures other than the claw-back feature, the Majority sets aside the (conceptual) proportionality criterion of the Blusun rule and adopts instead a quantitative test, based on the approach used in another case to quantify damages.133 The Majority's quantitative-proportionality test is based on its view of the Claimant's “legitimate" expectations to an actual rate of return. The Majority's reasoning is as follows: (i) “the only legitimate expectation the Claimant could have had was that of a ‘reasonable return'
129 A state may carefully consider the reasonable reliance interests of the recipients of subsidies and yet decide that those interests are outweighed by other interests. If such a decision were unreviewable (as a matter of application of the Blusun rule or for sheer practical reasons), this would be another criterion that places no meaningful limits on the state's discretion. ↩
130 Decision, ¶ 339. ↩
131 Decision, ¶ 355. ↩
132 Decision, ¶¶ 340-344. ↩
133 Decision, ¶¶ 357-358, relying on RL-0088, RREEF Infrastructure (G.P.) Limited and RREEF Pan-European Infrastructure Two lux S.à r.l. v. Kingdom of Spain (ICSID Case No. ARB/13/30), Decision on Responsibility and on the Principles of Quantum, 30 November 2018, ¶ 523. ↩
[Page 49]
in terms of the 1997 Law;"134 (ii) the “reasonable return" is a dynamic concept;135 (iii) the Disputed Measures set a target return of 7.398%; (iv) the Claimant's overall projected rate of return after the Disputed Measures is 13.8% (average of the experts' calculations); and (v) this rate of return is “well above the 7.398% target of the Spanish regulator.”136 On this basis, the Majority concludes that the Respondent did not breach its obligation to ensure a reasonable return and did not frustrate the Claimant's “legitimate” expectation to obtain such a return.137
112. As already discussed, the premise on which this whole approach is based is faulty. Law 54/1997 did not give producers a right to an actual rate of return (reasonable or unreasonable, floor or ceiling); it just set forth a conceptual target (reasonable rates of return with reference to the cost of money in the capital markets) for the Government to aim at in setting the premiums.138 A producer's actual rates of return could be higher or lower depending inter alia on the efficiency of its choices. For the same reason, the Claimant did not have a “legitimate” expectation to realizing any particular actual rate of return; its objectively reasonable expectation was merely to receive the premiums calculated on the basis of the Fourth Regime in the manner and for the period designated, subject to an overall expectation to the stability of that regime. Apart from being based on a false premise, the Majority's quantitative-proportionality calculation is also flawed in its economic effect: it counts against the Claimant the value of the efficiencies it realized under the abolished regime.
113. I conclude that if the Blusun criteria were intended, as surrogates for the requirement of fair and equitable treatment, to place some limits on the state's power to modify subsidies, they do not do that well or even meaningfully. I agree, though, that there is room for a meaningful criterion of proportionality in a fair-and-equitable treatment analysis. The key issue is, after all, whether it is fair and equitable for the investor or a class of investors to
134 Decision, ¶ 356. ↩
135 Decision, ¶ 366. ↩
136 Decision, ¶¶ 367-368. ↩
137 Decision, ¶ 369. ↩
138 Supra, ¶¶ 65, 111. ↩
[Page 50]
bear all or most of the cost of a policy change which the state deems desirable.139 In this respect, a criterion of proportionality that compares the burden placed on the investors with the burden placed (or which could have been placed) on an appropriate comparator group, or the population at large, would be more closely related to the ordinary meaning of “fair and equitable treatment.” The record contains no such comparison, but it does contain a related, indirect exercise. The Claimant's regulatory expert compared five alternative ways of dealing with the tariff deficit and concluded that the Respondent relied disproportionately on the Disputed Measures, which means that those measures were not a proportionate solution.140
114. In the preceding subsections I have analysed the circumstances of the present case that appear relevant to assess the claim for breach of the fair-and-equitable-treatment standard. Those circumstances include specific commitments made in the legal regime that the Disputed Measures replaced, the Claimant's objectively reasonable expectations, the degree of favourability and transparency of the Disputed Measures, and the effects of those measures on the stability of the antecedent regime, the specific commitments, and the Claimant's objectively reasonable expectations. I have also critically examined the criteria set forth in the Blusun rule. Partial conclusions have been drawn at each step of the analysis. It remains to reach a general conclusion.
115. If the Disputed Measures are considered in the light of all the circumstances analysed in the preceding subsections, it must be concluded that the treatment provided by those measures did not accord with the ordinary meaning of “fair and equitable.” Consequently, by imposing the Disputed Measures the Respondent breached its obligation under the
139 See CL-0117, 9REN Holding S.à.r.l. v. Kingdom of Spain (ICSID Case No. ARB/15/15), Award of 31 May 2019, ¶ 243 ("The question before the Tribunal however is whether such changes [changes of Spanish regulatory measures made in the exercise of Spanish sovereignty] can be made by Spain without financial consequences under the ECT.") (emphasis in the original). ↩
140 CE-1, Brattle, Changes to the Regulation of Wind Installations in Spain since December 2012, ¶¶ 132-149, esp. ¶ 149. (“The reasons behind Spain's failure to implement alternative solutions to the Tariff Deficit, including the CO2 tax and fuel tax and the profiling of FITs for renewable installations recommended by the EC or the CNE, remain unclear. The failure to exhaust other alternatives demonstrates that Spain relied disproportionately on the Disputed Measures – in other words they were not a proportionate solution.") (emphasis in the original). ↩
[Page 51]
second sentence of Article 10(1) of the ECT. Separately, for the same reasons set forth in respect of the analysis of the stability factor, the Respondent breached its obligation under the first sentence of Article 10(1) to create stable conditions for the Claimant's investments made under the Fourth Regime.
116. The Claimant argues that the Disputed Measures violated a portion of the third sentence of Article 10(1) of the ECT concerning unreasonable or discriminatory measures.141 The Respondent rejects that claim.142 The Majority takes the view that “unreasonable or discriminatory measures in the general sense are examples of measures that breach the FET standard as contained in Article 10(1), first and second sentences.”143 Then it finds that "[n]o conclusive evidence was provided to necessitate a finding of unreasonableness or discrimination under those elements of the FET standard, save insofar as the retro-active aspect of the Disputed Measures, in the form of the claw-back clause is concerned."144 On this basis, the Majority declines to make a separate finding of unreasonableness or discrimination, even in respect of the claw-back feature.145 In the end, this claim is rejected in the dispositif.146
117. The third sentence of Article 10(1) of the ECT provides that
"no Contracting Party shall in any way impair by unreasonable or discriminatory measures their [Investments of Investors of other Contracting Parties] management, maintenance, use, enjoyment or disposal."147
118. I cannot agree with the view that the quoted portion of the third sentence is somehow subsumed within the standard of fair and equitable treatment under the second sentence.
141 Decision, ¶¶ 371-377. ↩
142 Decision, ¶¶ 378-383. ↩
143 Decision, ¶ 387. ↩
144 Decision, ¶ 388. ↩
145 Id. ↩
146 Decision, ¶ 467(d), (e). ↩
147 C-0001, ECT, Art. 10(1), third sentence. ↩
[Page 52]
Legal provisions that purport to state separate rules should be applied separately, even if upon analysis their prescriptive contents partially overlap. In this respect, it cannot be assumed, without the benefit of analysis, that the prescriptive content of the quoted portion of the third sentence is a subset of the prescriptive content of the second sentence, or vice versa.
119. If the Majority had upheld the claims under the first and second sentences of Article 10(1) in full, it might have been justifiable to decline to address the claim under the third sentence, on the ground that the quantum of compensation would not have changed. But since the Majority rejects almost all aspects of the claims under the first and second sentences, it would have been appropriate for the Tribunal to analyse the prescriptive content of the third sentence and to resolve the claim thereunder. I cannot perceive any justification for the Majority's blanket dismissal of this claim.
120. For the purposes of this Partial Dissent, it is unnecessary to embark on a comparative analysis of the standard of fair and equitable treatment and that of unreasonable and discriminatory measures. But even without getting into details, one conclusion can be drawn from the preceding discussion. If the Disputed Measures breached the obligation of stability under the first sentence of Article 10(1) and the obligation to provide fair and equitable treatment under the second sentence, as I have concluded earlier, a fortiori they should be considered to have breached the obligation not to impair, by unreasonable measures, the use and enjoyment of the Claimant's investments.
121. The Respondent has raised two defences grounded on the EU legal regime on state aid. The first defence is that the Claimant did not have, and could not have had, any “legitimate” expectation to receiving any Special Regime subsidies.148 The second is that any Award rendered in this case ordering the Respondent to make payments in excess of those
148 Decision, ¶ 391. For present purposes, it is unnecessary to distinguish the various versions of the regime preceding the Disputed Measures. I shall therefore adopt the Majority's terminology and refer to it as the Special Regime. ↩
[Page 53]
provided by the Disputed Measures would itself constitute impermissible state aid and would be subject to a stand-still obligation under EU law.149
122. With regard to the second defence, I agree with my colleagues that the Tribunal is bound to decide the present case in accordance with the applicable law.150 The enforceability of the Award is a separate matter, to be addressed in a separate proceeding, outside this Tribunal's jurisdiction.151 The EC's threat to block enforcement of the Award cannot deter the Tribunal from fulfilling its function.
123. On the contrary, I disagree with the Majority's analysis and disposition of the first EU-state-aid defence. The Majority states its conclusion as follows:
“In the circumstances, the Tribunal finds[,] by majority[,] that under EU law and the law of Spain, the Claimant could not legitimately have expected that the Special Regime subsidies were, for certain, lawful. Even if the subsidies were lawful, it could not expect, under EU law and the law of Spain, that the amount of state aid granted under these measures would be paid for the lifetime of the plants. The Claimant should have known that these measures had not been notified to, let alone approved by, the EC.”152
124. The reasoning supporting these conclusions appears to consist of the following key steps: (i) under EU law, the Respondent had an obligation to notify state aid to the EC;153 (ii) state aid which is not notified to the EC, or which is implemented before it is authorized by the EC, is unlawful;154 (iii) in a case of unnotified state aid, the EC has the power to recover such aid from recipients or, alternatively, to approve such aid if it finds it compatible with
149 Id. ↩
150 Decision, ¶ 427. ↩
151 Decision, ¶ 422, quoting from Vattenfall AB and others v. Federal Republic of Germany (ICSID Case No. ARB/12/12), Decision on the Achmea Issue, 31 August 2018, ¶ 230; Decision, ¶ 427. ↩
152 Decision, ¶ 428. It is not clear what the majority means by saying that even if the subsidies were lawful, the Claimant could not have “legitimately" expected, under EU and Spanish law, that the subsidies would be paid for the lifetime of the plants. Perhaps the Majority is referring to its own prior analysis based on Spanish law, in which it concluded that the regime did not give rise to any such expectation. But then, why the reference to EU law? Or perhaps the Majority is referring to the Claimant's expectation preceding a subsequent EC decision to approve the subsidies (as in the case of the Disputed Measures), a decision which would erase the illegality. Or perhaps the Majority is alluding to something else. These uncertainties need not be resolved because the Majority takes the view, as discussed further in the text, that the subsidies regime was unlawful under EU law. ↩
153 Decision, ¶ 410. ↩
154 Decision, ¶ 411. ↩
[Page 54]
the internal market;155 (iv) the Special Regime constituted state aid and hence should have been notified to the EC;156 (v) the Respondent should have been aware of its notification duty and should have acted accordingly;157 (vi) it is undisputed that the Respondent made no such notification;158 (vii) ergo, the Special Regime was unlawful under EU and Spanish law;159 (viii) the Claimant should have known that the Special Regime had not been notified to the EC;160 (ix) ergo, the Claimant could not have had the "legitimate" expectation that the subsidies granted by the Special Regime were, for certain, lawful;161 and (x) therefore, the Claimant could not have derived any “legitimate” expectations from that regime.162
125. This reasoning, if I understand it correctly, has two important implications. The first is that the Majority's EU state-aid analysis largely obviates its own analysis under the Blusun rule. As already discussed, the Majority concludes, incorrectly in my view, that the Claimant could have derived no “legitimate” expectations from the regime of RD 661/2007. That
155 Decision, ¶ 411. ↩
156 Decision, ¶ 419. ↩
157 Decision, ¶ 426. ↩
158 The Majority acknowledges that the EC (i) was well informed of the Special Regime and even praised it; (ii) took a "rather generous" approach to such schemes; and (iii) took no enforcement action while that regime was in force or afterwards. Decision, ¶¶ 415-417, 425, 427. Yet these facts play no role in the Majority's reasoning leading to the conclusion quoted in the preceding paragraph. The absence of enforcement action does play a role in the Majority's discussion of the claw-back feature in the context of EU state-aid law, as discussed later in the text. ↩
159 This conclusion is not explicit in the Majority's argument, but it follows necessarily from the preceding steps. In addition, the Majority notes that the EC took the view that the Disputed Measures “were state aid and, not having been notified, were unlawful," in spite of which the EC approved them as compatible with the internal market. Decision, ¶ 421, citing RL-0073, EC decision of 10 November 2017, ¶¶ 84-89. The Majority relies on this EC Decision to conclude "by parity of reasoning" that the Special regime constituted state aid. Parity of reasoning would also lead the Majority to conclude that the Special Regime, not having been notified, was unlawful under EU law. Decision, ¶ 419. ↩
160 Decision, ¶ 428. ↩
161 Decision, ¶ 428. ↩
162 Decision, ¶ 423(a) (“In principle, an investor cannot have a legitimate expectation of treatment which is unlawful under the law of the host state, provided that the host state law itself is not acting inconsistently with the treaty under which the tribunal exercises its jurisdiction."). This statement may be acceptable in principle, but only in principle. It cannot be used as a technical rule to foreclose consideration of other circumstances bearing on the objective reasonableness of the expectation. See CL-0016, Southern Pacific Properties (Middle East) Limited v. Arab Republic of Egypt (ICSID Case No. ARB/84/3), Award on the Merits, 20 May 1992 (henceforth, SPP v. Egypt), ¶¶ 82, 83 ("These acts, which are now alleged to have been in violation of the Egyptian municipal legal system, created expectations protected by established principles of international law. A determination that these acts are null and void under municipal law would not resolve the ultimate question of liability for damages suffered by the victim who relied on the acts."). ↩
[Page 55]
conclusion was based on the notion that the “stream” (RD 661/2007) cannot rise higher than its "source” (Law 54/1997), but the same conclusion now flows from the Majority's EU state-aid analysis. Up to that point, both analyses of the Majority lead to the same result, but the same is not true in respect of Article 30.4 of Law 54/1997, which the Majority regards as the source (the sole source) of a “legitimate” expectation to a reasonable rate of return, which in turn becomes the benchmark for the Majority's notion of proportionality. If Article 30.4 of Law 54/1997 is part of a subsidies regime which was not notified to the EC and hence was unlawful under EU and Spanish law, it is difficult to see how (under the Majority's reasoning) Article 30.4 could be the source of an expectation of a reasonable return or how such expectation could serve as a measure of proportionality.
126. A second implication flows from the Majority's discussion of the claw-back feature in the context of the unnotified subsidies regime. The Majority dismisses the Respondent's state-aid defence on the ground that “the subsidies having been paid (and subject to any lawful recovery measures by the EC, which did not occur), they remained entitled to the benefit of the stable regime which Article 10(1), first and second sentences, of the ECT promised."163 This statement underscores the fragility of the basis on which the Majority finds a breach of Article 10(1), first and second sentences, in respect of the claw-back feature. That finding is now “subject to any lawful recovery measures by the EC.” No such recovery has occurred to date, but it may still occur (if EU law so authorizes), or at any rate it might have occurred. Therefore, in the Majority's view the Respondent's breach of the ECT in respect of the claw-back feature depends on the contingent fact that the EC has not exercised its enforcement powers under EU law.
127. The Majority's line of reasoning on EU state aid and its conclusions and implications are, in my view, incorrect. First, it is not correct to regard EU law as part of the law to be applied in this case, and to apply it as such, even though the EU Treaties, as res inter alios acta, are not binding in the relations between the Respondent and Japan. The Majority indeed applies EU law as law, notably by adopting as a major premise that unnotified state aid is
163 Decision, ¶ 430 (emphasis added). ↩
[Page 56]
unlawful,164 by holding (by parity of reasoning from the EC's Decision of 10 November 2017) that the Special Regime constituted state aid,165 and by concluding that, because the Respondent failed to notify the Special Regime, that regime was unlawful under EU and Spanish law.166 This last conclusion is, in turn, the major premise for the findings that the Claimant could have had no “legitimate” expectation to the lawfulness of the Special Regime and that the stability of that regime was subject to the EC's power to recover the subsidies paid under that regime.
128. Second, it is incorrect to accept the possibility that EU law, as applicable law, might trump a legal conclusion based on the ECT. If the EC's exercise of its powers under EU law could have foreclosed a ruling from the Tribunal that the stability obligation under Article 10(1) of the ECT was breached, the source of such powers (EU law) would have trumped the ECT. This conclusion cannot be reconciled with the primacy of the ECT under its Article 26(6) or even with the Majority's own view that, through the “indirect relevance” of Article 16 of the ECT, an ECT rule that is more favourable to the investor would prevail over a conflicting rule of EU law.167 Apart from this role as a trump card, EU law on state aid works as an overriding consideration in the fair-and-equitable-treatment analysis, superseding other considerations required by such analysis. This whole approach is inconsistent, in my view, with the system of norms which the Tribunal is bound to apply under Article 26(6) of the ECT.
129. On the contrary, from the standpoint of international law, which is the Tribunal's standpoint, the laws of Spain (which happen to incorporate EU state-aid rules) are merely facts, except to the extent that the ECT effects a renvoi to them expressly or by clear implication.168 As already explained, EU law, based on treaties which are res inter alios acta, plays a similar role. Therefore, both Spanish law and EU law contribute to the factual matrix in respect of which the provisions of the ECT must be applied. But the legal status
164 Decision, ¶ 411. ↩
165 Decision, ¶ 419. ↩
166 Decision, ¶ 421. ↩
167 See discussion supra, ¶ 9. ↩
168 Decision, ¶ 423(b). ↩
[Page 57]
of the Special Regime under EU law is not the sole factual element to be considered in applying Article 10(1) of the ECT. All the circumstances related to such legal status must be considered, including in particular the Respondent's failure to notify the regime to the EC and the context in which that failure took place.
130. Let us then examine the influence of these EU state-aid considerations on the Claimant's objectively reasonable expectations. As we have seen, the Majority takes the view that “the Claimant could not legitimately have expected that the Special Regime subsidies were, for certain, lawful,"169 and such a lack of certainty would prevent any “legitimate” expectation of treatment according to that regime.170 This requirement of certainty of lawfulness appears to be based on a series of decisions of the CJEU.171 Yet, whatever may be said about such a certainty requirement as a condition for “legitimate expectations” to arise under EU law, CJEU precedents have no bearing on the conditions for “legitimate expectations" (or objectively reasonable expectations) under the ECT.172
131. As a matter of application of the ECT, it cannot be right to require that the Claimant have expected, "for certain,” that the Special Regime was lawful. The Special Regime, like any other Spanish laws and royal decrees, bore a presumption of legality, and the Claimant was entitled to rely on that presumption, unless and until other factors came to light that were sufficient to overcome that presumption.173 Even apart from the presumption of legality, it is unreasonable to demand that, to be "legitimate” (i.e. objectively reasonable under the ECT), the Claimant's expectation on the lawfulness of the Special Regime reach the level of certainty. In the absence of a binding decision by a competent authority, a decision which did not exist at the time the Claimant formed its expectations and still does not exist
169 Decision, ¶ 428 (emphasis added). ↩
170 Decision, ¶ 423(a). ↩
171 See Decision, ¶ 412, quoting from Case C-199/06, Centre d'exportation du livre français v. Société internationale de diffusion et d'édition, 12 February 2008 (GC), ¶ 67, citing Case C-91/01 Italy v. Commission [2004] ECR I-4355, ¶ 66 ("[S]o long as the Commission has not taken a decision approving aid, and so long as the period for bringing an action against such a decision has not expired, the recipient cannot be sure as to the lawfulness of the proposed aid which alone is capable of giving rise to a legitimate expectation on his part.") (emphasis added). ↩
172 The Majority admits as much in respect of EC's pronouncements (“[...] so does the EC not have jurisdiction to impose a binding interpretation of the meaning of legitimate expectations under the ECT.”). Decision, ¶ 432 (emphasis in the original). ↩
173 See CL-0016, SPP v. Egypt, ¶¶ 82, 83. ↩
[Page 58]
today, it is difficult to see how a reasonable investor could have been certain at that time that the Special Regime was or was not in accord with EU or Spanish law.
132. The question is, then, whether having regard to all the relevant circumstances, it was objectively reasonable for the Claimant to rely on the legality of the Special Regime or, more precisely, whether from the standpoint of a reasonable investor there were sufficient elements to overcome the presumption of legality of that regime, so that it would have been objectively unreasonable to rely on it. The key facts are the following:
174 Decision, ¶ 410, quoting from Art. 108(3) of the TFEU. ↩
175 The Majority concludes that the Special Regime was subject to the notification rule, but it reaches that conclusion, by parity of reasoning, on the basis of the Decision of the EC of 10 November 2017, issued ten years after the time the Claimant was forming its expectations based on the Fourth Regime. See Decision, ¶ 419. ↩
176 Decision, ¶ 418. ↩
177 Decision, ¶ 417, quoting from uncontradicted evidence of Claimant's expert Mr. Carlos Lapuerta. ↩
178 Decision, ¶ 424. ↩
[Page 59]
133. This set of facts raises two fundamental questions: (i) did the Respondent make any initial determination, at the times the various components of the Subsidies Regime were adopted, whether they fell under Article 108(3) of the TFEU and, if so, what was that determination? and (ii) why did the Respondent fail to give notice to the EC of the various components of the Subsidies Regime?
134. On the first question, it should be recalled that the Respondent had a generic obligation under Article 108(3) of the TFEU to notify any state-aid scheme to the EC. As the obligor, the Respondent was responsible for making an initial determination, for its own compliance purposes, that the Special Regime did or did not fall under Article 108(3) of the TFEU. Such a determination, though not conclusive, would have been relevant to a decision by the Respondent whether or not to notify the Special Regime to the EC. Also to be recalled is that no competent authority has ever determined that the Special Regime was or was not subject to notification under Article 108(3). The Majority does find that the Special Regime
179 Decision, ¶ 425. ↩
180 Decision, ¶¶ 425, 427. ↩
181 CE-1, Brattle's First Regulatory Report, ¶¶ 111-112 (“Spain added about 2,089 MW per year on average from 2004 up to and including 2009, significantly more than in previous years."), relying on BRR-38 (official source). ↩
182 Decision, ¶ 421. ↩
[Page 60]
was subject to notification, but the Tribunal is not a competent authority to rule on this matter.183 In any event, the Majority's finding is based, by “parity of reasoning,” on an EC decision dated 10 November 2017.184 No such ex post facto determination can be relevant to ascertain how the Respondent viewed the matter in 1997 (when Law 54/1997 was enacted) or in 2007 (when RD 661/2007 was issued), or to assess the Claimant's objectively reasonable expectations at the relevant times.
135. The record contains no evidence of (i) any determination, made by the Respondent at any time, that the Special Regime did or did not fall under Article 108(3) of the TFEU or (ii) the reasons for the Respondent's failure to give notice to the EC under that provision.185 Nevertheless, in the absence of evidence let us consider the various possibilities from the standpoint of a reasonable investor.
136. First, the Respondent may have studied the matter and concluded that the Special Regime did not fall under Article 108(3), either because it was originally designed not to involve payments from the state treasury or for other reasons. In such case, that conclusion may very well have been the reason for not notifying the scheme to the EC. But if the state had any doubt about its initial determination (which after all was not conclusive), it would have been bound to disclose both the determination and the doubts, to comply with the requirement of transparency under Article 10(1) of the ECT. No such disclosure having been made, it would have been objectively reasonable for an investor to conclude that the state had satisfied itself (by consulting with the EC or otherwise) that the Special Regime was not subject to Article 108(3) and that no risk of illegality existed.
183 The Majority takes the view that "this Tribunal does not have jurisdiction to interpret EU law itself.” Decision, ¶ 432. In my view, the Tribunal does not have jurisdiction to rule on the question whether the Special Regime was subject to notification under Article 108(3) of the TFEU because EU law is not part of the law which the Tribunal is charged with applying in this case. ↩
184 Decision, ¶ 419. ↩
185 The burden of proof on these matters falls on the Respondent, to the extent that they are necessary elements of its defence based on EU state-aid law. That is so, for two reasons: (i) it was the Respondent which invoked its own failure to notify the Special Regime as an affirmative defence (actori incumbit onus probandi; reus in excipiendo fit actor), and (ii) the Respondent is the only party capable of providing such evidence. ↩
[Page 61]
137. Second, the Respondent may have concluded that the Special Regime indeed fell under Article 108(3). In such case, the Respondent may have had various reasons for not giving notice to the EC despite its own conclusion that notice was required. For example, the state may have decided not to notify the Special Regime to preserve a defence of illegality in the event it wished to repudiate its commitments in the future – a case of bad faith which can be entertained as a hypothesis but cannot be taken as established without adequate proof. Alternatively, the state may have considered that, given the EC's actual knowledge of the Special Regime, its favourable opinion of it, its generous approach toward the application of state-aid rules to energy projects based on renewable sources, and its willingness to approve such schemes even if they were not notified at the proper time, the required notification was as a simple formality which could have been accomplished later if necessary. As a third alternative, the state may have had other reasons for the absence of notice, including a simple oversight. In all those scenarios, the Respondent's decision not to give notice would have involved a risk to the recipients of the subsidies – a risk of which the Respondent would have been aware, having concluded (by hypothesis) that notice was indeed required. In such circumstances, transparency would have required the Respondent to disclose to the investors its conclusion that the Special Regime was subject to notice and the risks resulting from its decision not to notify. Absent such a disclosure, a reasonable investor would have had no grounds to believe that the Respondent had reached that conclusion or that it was consciously subjecting investors to the risks of illegality and disgorgement of the subsidies.
138. Third, and least likely, the Respondent may simply have overlooked making an initial determination whether the Special Regime fell under Article 108(3) and such oversight may have been the reason for its failure to notify the EC. Under that hypothesis, the absence of notice and the consequent illegality of the Special Regime under EU law would have been the result of the state's negligence. In such case, it would be unreasonable and inconsistent with the ECT standard of fair and equitable treatment to place the harmful consequences of the state's negligence on the investor's shoulders. A reasonable investor
[Page 62]
cannot be held to a higher standard of diligence than the state itself, on a matter concerning the state's compliance with its own obligations.186
139. In light of the relevant facts and the preceding analysis of alternative answers to questions not covered by the evidence, it must be concluded that, under the fair-and-equitable-treatment standard of the ECT, no elements were available to a reasonable investor to overcome the presumption of legality of the Special Regime – even if the investor knew or could have discovered that the Respondent had failed to notify the Special Regime to the EC. The key issue is who should bear the injurious consequences of the Respondent's own failure to notify the Special Regime to the EC, if any such notification was indeed required. Whether that omission was in good or bad faith, malicious or negligent, the result is the same: it was the state's own conduct which caused the illegality, or risk of illegality, of the Special Regime under EU law, while the state benefited from the investments induced by that regime over a span of more than sixteen years. It is hardly consistent with the principle of good faith for the state now to seize on its own breach of duty to shift the entire burden of that breach on the investor, nor does it seem consistent with a sense of juridical propriety for a tribunal to condone such result.187
186 See RL-0090, Cube, ¶ 306 ("The obligations regarding State aid were incumbent upon the Respondent, and investors were entitled to assume that they had been taken into account by the Respondent when drafting its legislation. It was not for the Claimants to second-guess the Respondent's legislature. Moreover, at the time that the investments were made it was not at all clear that the tariff regime should be regarded as State aid, let alone as impermissible State aid."); RL-0096, Dissenting Opinion of Horacio Grigera Naón, ¶¶ 29-42 (analysis coinciding in several respect with that in the text), in RL-0095, BayWa r.e. Renewable Energy GmbH and BayWa r.e. Asset Holding GmbH v. Kingdom of Spain (ICSID Case No. ARB/15/16), Decision on Jurisdiction, Liability and Directions on Quantum, 2 December 2019. In Cavalum, the tribunal noted that neither Spain nor the EC ever had any concerns that the RD 661/2007 regime was contrary to state aid rules. The tribunal stated that, given its conclusions, the respondent's state-aid argument had not arisen, “but if it had arisen, the Tribunal would have dismissed it on the basis that there is no necessary connection between an investor's legitimate expectation of a reasonable rate of return and a failure by the State to notify state aid, and that in any event it was not now open to Spain in the light of its prior conduct to raise it.” RL-0102, Cavalum SGPS, S.A. v. Kingdom of Spain (ICSID Case No. ARB/15/34), Decision on Jurisdiction, Liability and Directions on Quantum, 31 August 2020, ¶ 611 (emphasis added). ↩
187 See CL-0105, Cheng, Bin, GENERAL PRINCIPLES OF LAW AS APPLIED BY INTERNATIONAL COURTS AND TRIBUNALS, Cambridge 1953, 2006, pp. 149 et seq. ("A State may not invoke its own illegal act to diminish its own liability"), citing, e.g., PCIJ, Advisory Opinion No. 15 (1928) (“Poland could not avail herself of an objection which [...] would amount to relying upon the non-fulfilment of an obligation imposed upon her by an international agreement."). ↩
[Page 63]
140. For the reasons and to the extent set forth in the preceding sections, I dissent from the Majority's decisions concerning (i) the application of EU law, (ii) the claims based on Article 10(1) of the ECT, and (iii) the effects of EU law on state aid, and from the Majority's reasoning supporting those decisions.
141. I have explained to what extent and why the Tribunal should have upheld the Claimant's claims based on the first three sentences of Article 10(1) of the ECT. If this conclusion had been accepted, the quantum of compensation would have had to be calculated under principles different from those set forth in the Decision. Accordingly, my dissent should be understood consequentially to include the directions on quantum and other ancillary results.
[Signed]
Oscar M. Garibaldi
Arbitrator