ON THE DENUNCIATION OF THE ICSID CONVENTION, CONSENT TO ICSID JURISDICTION, AND THE LIMITS OF THE CONTRACT ANALOGY
Abstract and Keywords
This chapter argues that Professor Schreuer's interpretation of Article 72 is incorrect. The contract analogy, which is unobjectionable as an aid to explaining the basic workings of the consent regime, ceases to be useful when it dictates conclusions that are inconsistent with the text of Article 72 and the context of the ICSID Convention. At that point, applying the contract analogy to Article 72 becomes an instance, in the context of legal discourse, of what J. S. Mill called the fallacy of false analogies. Recent studies have spent more time on the development of alternative theories than on pointing out the limits of the contract analogy. Studies must not only offer an alternative analysis based on those rules of treaty interpretation, but show why the opposing theories are inadequate. The chapter presents an analysis of this issue supplemented with brief observations on the limits of the contract analogy and on the inadequacy of other theories that, though by no means attributable to Professor Schreuer, may tempt those searching for an extra-textual justification for interpreting ‘consent’ in Article 72 as an ‘agreement to arbitrate’.
There is another [mode of error in the employment of arguments of analogy], more properly deserving the name of fallacy; namely, when resemblance in one point is inferred from resemblance in another point, though there is not only no evidence to connect the two circumstances by way of causation, but the evidence tends positively to disconnect them. This is properly the Fallacy of False Analogies.
(John Stuart Mill)1
[…] et idem indignor quandoque bonus dormitat Homerus.
Even good Homer sometimes nods off, complained Horace. Even great scholars sometimes err, we say, not to complain but to take comfort in our shared human condition.
For me and many of my colleagues, Professor Schreuer's great work on the ICSID Convention is inseparable from our practice of investment arbitration. We turn to it first, and often last, for teaching and insights into all matters related to the Convention. It is a familiar companion, an indispensable source of wisdom, an old mentor with whom we always expect to agree.
In the spring and summer of 2007, events led me to study the interface between denunciation of the Convention and consent to ICSID jurisdiction. To my surprise, I found myself in disagreement with Professor Schreuer's interpretation of Article 72 of the Convention. Shortly afterwards, I sought out Professor Schreuer at a conference and told him that I feared I could not agree with his reading of Article 72. ‘Well, that's all right!’ he beamed. Over lunch I explained the substance of my interpretation. He listened attentively and encouraged me to send him a more detailed email message. He responded (p. 252 ) kindly and generously to that message and gave me reason to believe that he was taking my interpretation seriously. Yet, I do not presume to know Professor Schreuer's current views on Article 72 or the extent to which my comments (and perhaps those of others) may have influenced those views. For this and many other weightier reasons, I eagerly await publication of the second edition of his Commentary.
A true scholar, Professor Schreuer was not only willing to listen to my criticism, but he also encouraged me to put it in writing and, later, he repeatedly urged me to publish it. Now that I have completed the article, it is fitting that I offer it as a small contribution to this Liber Amicorum, in tribute to Professor Schreuer, his exemplary scholarship, and his vast contribution to our field.
On 1 May 2007, the Republic of Bolivia gave written notice of denunciation of the ICSID Convention, the first contracting State to do so.3 Article 71 of the Convention permits denunciation by written notice to the depositary and provides that any such denunciation shall take effect six months after receipt of the notice.4 Bolivia's denunciation of the Convention took effect on 2 November 2007, six months after the World Bank, the depositary of the Convention, received the notice.
At the time Bolivia gave notice of denunciation, it was a party to several bilateral investment treaties providing for ICSID jurisdiction of defined classes of investment disputes between each State party and nationals of the other State party.5 Within the six-month period following the notice of denunciation, at least one foreign investor gave its own (p. 253 ) consent to ICSID jurisdiction and submitted to the Centre an investment dispute with Bolivia under the Netherlands-Bolivia bilateral investment treaty.6
Bolivia's denunciation of the Convention has brought to the forefront of practical issues one that had been confined to academic discussion: What are the effects of a notice of denunciation of the Convention on consent to ICSID jurisdiction given by the denouncing State before the notice of denunciation was received by the World Bank? This issue can also be stated in operational terms: Does the notice of denunciation wipe out any pre-notice consent? May an investor who has a dispute with the denouncing State validly give its own consent to ICSID jurisdiction within the six-month period following the notice of denunciation? May the investor achieve the same result after the end of that six-month period?
The traditional way of looking at this issue stems from the writings of Professor Christoph Schreuer, the foremost scholar on the Convention. As we shall see in detail, Article 25(1) of the Convention provides that one of the requirements for ICSID jurisdiction is the consent of both parties to the dispute and further prescribes that, once both parties have given their consent, no party may withdraw its consent unilaterally.7 There is an obvious analogy between this regime and the rules of most legal systems on the formation of contracts. Just as contracts are formed by an offer and a matching acceptance, which precludes any subsequent withdrawal of the offer, one may look at ‘consent’ as being formed by an ‘offer’ of consent by one party to the dispute, which may be ‘accepted’ by the other party, thus ‘perfecting’ the consent. Professor Schreuer uses this contract analogy (as I shall call it) and the terminology based thereon as the major premise of his analysis of the interface between consent and denunciation.8
The effects of denunciation on consent are governed, in principle, by Article 72 of the Convention. This article provides that the notice of denunciation ‘shall not affect the rights or obligations under this Convention of [the denouncing State] […] arising out of consent to the jurisdiction of the Centre given by [the denouncing State] before such notice was received by the depositary’.9 Applying the contract analogy, Professor Schreuer has proposed to read ‘consent’ in Article 72 as ‘perfected consent’.10 As ‘[c]onsent is only perfected after it has been accepted by both parties,’ he has written, ‘a unilateral offer of consent by the host State through legislation or a treaty before a notice [of denunciation] would not suffice. The effect of the continued validity of consent under (p. 254 ) Art. 72 would only arise if the offer was accepted in writing by the investor before the notice of denunciation […].’11
It is one thesis of this chapter that Professor Schreuer's interpretation of Article 72 is incorrect. The contract analogy, which is unobjectionable as an aid to explaining the basic workings of the consent regime, ceases to be useful when it dictates conclusions that are inconsistent, as we shall see, with the text of Article 72 and the context of the Convention. At that point, applying the contract analogy to Article 72 becomes an instance, in the context of legal discourse, of what J.S. Mill called the fallacy of false analogies. Professor Schreuer's authority on the Convention is unmatched, and generally recognized to be so, but (to use an analogy of my own drawn from my other avocation) even a master sculptor occasionally drives his chisel too far, and it often falls to his admiring colleagues to call attention to the slip, out of respect for the master's art and a shared commitment to the pursuit of unattainable perfection.
Several articles and notes on the effects of a notice of denunciation on consent have been published in the wake of Bolivia's notice. Most of those papers reject Professor Schreuer's analysis and conclude, on various grounds, that consent expressed in a treaty is binding on the denouncing State for the life of the treaty and may be matched by the investor at any time during that period or (in the case of one commentary) until the denunciation takes effect.12 (p. 255 ) A minority of the articles reach the opposite conclusion, based on an uncritical application of the contract analogy.13
I throw my hat into this increasingly crowded ring for two main reasons. The first is that my own views about the effects of denunciation on consent do not exactly coincide with any of the analyses published so far. While I am in agreement with many of the conclusions, and some of the reasoning, of most of the recent articles, there are important aspects of those studies that I cannot endorse. One troubling aspect of some recent discussions is insufficient attention to the text of Convention; another is a tendency to resort to meta-theories, such as a ‘dynamic-evolutionary’ method of treaty interpretation, in an attempt to justify a departure from the text of the Convention which, upon analysis, is only apparent.14 As we shall see, the perplexing problem of the effects of denunciation on consent can be solved by a straightforward application of the principles of treaty interpretation codified in the Vienna Convention.15 In particular, the difficulties of this issue largely disappear if one pays close and systematic attention to the text of Articles 25(1) and 72 of the Convention and the context formed by the other provisions of the treaty, read in the light of the object and purpose of the Convention. The bulk of this chapteer is an attempt to prove this point. My aim is not to restate the historical and political background of the issue, which is well covered in the literature, but to centre on the Convention itself.16
The second main reason for this chapter concerns the menu of competing theories that are now available to interested readers and decision-makers. After reviewing Professor Schreuer's account and the more recent literature, a reader may be left with the impression that the choice among competing theories depends on the paradigm that the reader decides to adopt. If one embraces the contract analogy—the reader may think—then it follows that Article 72 refers only to ‘perfected’ consent, with the consequence that an investor may not ‘perfect’ consent after the notice of denunciation is received by the World Bank. If one rejects the contract analogy—the reader may continue to think—and adopts instead, for example, the view that a State is internationally bound by its unilateral expression of consent, then it follows that the investor may validly give its own consent, and submit the dispute to the Centre, for as long as that expression of consent is binding, or at (p. 256 ) least until the notice of denunciation takes effect. The hypothetical reader could hardly be blamed for forming the impression that the matter rests on a choice of paradigm—the interpreter's choice—because the recent studies have spent more time on the development of alternative theories than on pointing out the limits of the contract analogy.
In my view, treating this issue as a mere choice of paradigm is an untenable position. Treaty interpretation is not a matter of choice, or ‘approach’, or philosophical preference; it is an endeavour governed by rules based on the uniform practice of States and the common conviction of their binding force, and codified in the Vienna Convention on the Law of Treaties. That is the common paradigm on which any proper attempt to interpret the Convention must be based. It is therefore important not merely to offer an alternative analysis based on those rules of treaty interpretation, but to show why the opposing theories are inadequate. Accordingly, I shall supplement my own analysis of the issue with brief observations on the limits of the contract analogy and on the inadequacy of other theories that, though by no means attributable to Professor Schreuer, may tempt those searching for an extra-textual justification for interpreting ‘consent’ in Article 72 as an ‘agreement to arbitrate’.
There is a manifest political tension between a State's denunciation of the ICSID Convention and its pre-existing consent to ICSID jurisdiction. As the case of Bolivia illustrates, the consent and the denunciation will likely be acts of different administrations of opposite political stripes and the denunciation may well be an aspect of a campaign by the later administration to undo the political legacy of its predecessor. But to acknowledge these political facts is not to admit that they have, or ought to have, legal consequences that are inconsistent with the Convention and general international law. The effects of denunciation on consent are a matter of law. While the denouncing State would ordinarily have an interest in no longer being subjected to ICSID arbitration and the investor would ordinarily have an opposite interest, the interpreter's task is not to find a ‘balance’ between those two interests but to determine, in good faith, where the line has been drawn by the Convention and the applicable rules of customary international law.17
(p. 257 ) B. Consent
The system of dispute resolution created by the ICSID Convention is based on the consent of both parties to the dispute.18 More precisely, it is a necessary condition for ICSID jurisdiction under the Convention that both parties consent to submit the dispute to the Centre. This principle is established by the first sentence of Article 25(1) of the Convention:
This provision, considered for the moment in isolation, imposes four individually necessary and jointly sufficient conditions for ICSID jurisdiction: (i) that the dispute be a legal one; (ii) that it arise directly out of an investment; (iii) that it be a dispute between a contracting State (or an appropriately designated constituent subdivision or agency) and a national of another contracting State, and (iv) that both parties consent in writing to submit the dispute to the Centre.20 As we shall see, Article 72 of the Convention qualifies, in an important respect, the requirement that the parties to the dispute be a contracting State (or a designated constituent subdivision or agency) and a national of another contracting State.
The jurisdiction of the Centre shall extend to any legal dispute arising directly out of an investment, between a Contracting State (or any constituent subdivision or agency of a Contracting State designated to the Centre by that State) and a national of another Contracting State, which the parties to the dispute consent in writing to submit to the Centre.19
The fourth requirement of Article 25(1) of the Convention is that both parties to the dispute consent to submitting it to the Centre. In other words, the consent that is required for ICSID jurisdiction is common consent.21 Article 25(1) further provides that a party that gave its consent may withdraw it as long as the other party has not given its own consent. This rule follows from the second sentence of Article 25(1), which states: ‘When the parties have given their consent, no party may withdraw its consent unilaterally’.22
(p. 258 ) In these provisions, and throughout the Convention, the term ‘consent’ is used to refer to the consent of a party to the dispute, that is to say, to each party's individual consent. For example, Article 25(1) refers to the parties giving ‘their’ consent and no party being authorized to withdraw ‘its’ consent unilaterally; Article 25(3) refers to ‘consent by a constituent subdivision or agency of a Contracting State’; Article 25(4) disclaims that the unilateral notification it contemplates might constitute the consent required by Article 25(1); and, as we shall discuss in detail, Article 72 refers to the consent given by the denouncing State or by its national before the notice of denunciation is received by the depositary. These references would be inexplicable if ‘consent’ had been used in the Convention only in the sense of common consent.
When the Convention refers to common consent, as in Article 25(1) (‘both parties consent’), it does so by referring to the consent of both parties to the dispute. If the word ‘consent’ had been used in the Convention to refer only to common consent (as distinguished from individual consent), it would have been unnecessary to qualify ‘consent’ as the consent of both parties. In short, there is no textual or contextual basis for the assumption that ‘consent’ in the Convention cannot mean anything other that common (or matched) consent, that is, that the Convention speaks of ‘consent’ only when both parties have consented. On the contrary, there is overwhelming evidence in the Convention that the term ‘consent’ was used in the sense of individual consent.23
These textual and contextual observations, we shall see, are critical to a proper analysis of the effects of denunciation of the Convention on previously expressed consent. An analysis that starts from the premise that ‘consent’ in the Convention always means common (or ‘perfected’) consent is bound to reach the wrong conclusions. As already noted, that premise is based on what I have called the contract analogy, that is, an analogy between the regime of consent to ICSID jurisdiction under the Convention and the familiar rules of domestic law on the formation of contracts. The contract analogy is helpful when it is used to explain the regime of consent in the Convention, to the extent that a reference to a more familiar institution may be needed, but it ceases to be helpful—and becomes positively misleading—when it is relied upon to derive conclusions that contradict the text and context of the Convention. As discussed in more detail in Section D, this unwarranted application of the contract analogy may derive (perhaps unconsciously) from the ambivalence of the word ‘consent’, which can be used to refer, depending on the context, both to individual consent and to common consent. By contrast, the word ‘contract’ is never used as a synonym of ‘offer’ or ‘acceptance’; it is used unambiguously to refer to the legal product of these two distinct unilateral acts.
(p. 259 ) C. Denunciation and Consent
The ICSID Convention permits a contracting State to denounce the treaty by a unilateral act of a prescribed form and regulates the time at which denunciation takes effect. Article 71 provides as follows: ‘Any Contracting State may denounce the Convention by written notice to the depositary of this Convention. The denunciation shall take effect six months after receipt of that notice.’24
Under the rules of customary international law, codified in the Vienna Convention, any denunciation of the ICSID Convention would have to take place in conformity with this provision.25 The prescribed form of denunciation is written notice to the depositary of the ICSID Convention which, according to Article 73, is the World Bank.26
Article 71 prescribes that the denunciation takes effect six months after the World Bank receives the notice of denunciation. This is a general, unqualified provision. Consequently, unless there is a specific exception in the Convention for matters in respect of which the denunciation takes effect earlier, the notice of denunciation has no effect in the six-month period following receipt. It is sometimes assumed, incorrectly, that Article 72 of the Convention introduces an exception to the general rule of Article 71. To examine this assumption, we must first consider the function of Article 72 in the system of the Convention.
The function of Article 72
A superficial reading of Article 72 of the Convention might leave the impression that it modifies the rule that denunciation has no effect in the six-month waiting period.27 In fact, such a reading has no basis in Article 72 and is in fact contradicted by the text of Article 72 and the context of the Convention. Article 72 provides as follows:
As a first approximation, we should note that Article 72 is a negative prescription. Article 72 specifies (or clarifies) that the notice of denunciation (let us ignore a notice under Article 70 for present purposes) shall not have certain effects. It provides that notice of denunciation shall not affect the rights and obligations under the Convention arising out of what may be called ‘pre-notice’ consent, that is,the consent given by the denouncing State or its national before the notice was received by the depositary. Article 72 does not (p. 260 ) add any effects to the notice of denunciation, nor does it introduce any change in the general rule of Article 71 on the timing of such effects. In other words, Article 72 does not deal with the effects of a notice of denunciation ratione temporis. The article deals with the effects of such notice ratione materiae, but only to exclude certain matters from the scope of those effects.
Notice by a Contracting State pursuant to Articles 7028 or 71 shall not affect the rights or obligations under this Convention of that State or of any of its constituent subdivisions or agencies or of any national of that State arising out of consent to the jurisdiction of the Centre given by one of them before such notice was received by the depositary.29
By restricting the effects of a notice of denunciation, Article 72 introduces an important exception to the requirement of Article 25(1) that the parties to the dispute be a contracting State (or a designated constituent subdivision or agency) and a national of another contracting State. To anticipate part of the discussion on the content of Article 72, there is general agreement that, at least if both parties to the dispute have given their consent to ICSID jurisdiction before the notice of denunciation is received by the World Bank, the dispute may be submitted to ICSID at any time thereafter, even after the notice of denunciation has taken effect under Article 71.30 That point was made abundantly clear in the travaux préparatoires of the Convention.31 After the notice of denunciation takes effect, however, the denouncing State is no longer a contracting State and its nationals are no longer nationals of a contracting State. If the denouncing State and its nationals may be parties to a dispute even though they are no longer a contracting State and nationals of a contracting State, respectively, it must be because the in personam requirement of Article 25(1) is modified by Article 72.32 The extent of the modification depends on the content of Article 72, to which we now turn.
The content of Article 72
As already noted, Article 72 refers to the individual consent given by the denouncing State (or any of its constituent subdivisions or agencies) or by a national of the denouncing State.33 There is no reference in either case to any consent by the other party to the (p. 261 ) dispute34 and no basis to suppose that ‘consent’ here refers to consent matched by such other party. So, what does Article 72 mean? It simply means what it says: that a notice of denunciation does not affect the rights or obligations under the Convention of the denouncing State or its national arising out of any individual consent to ICSID jurisdiction that either of them may have given before the notice was received by the World Bank.
What, then, are the rights and obligations that are unaffected by the notice of denunciation? As a first general observation, they are rights and obligations under the Convention.35 Second, they are rights and obligations that arise out of a pre-notice individual consent to ICSID jurisdiction. At a minimum, a right or obligation under the Convention can be said to arise out of such consent when such consent is one of the antecedents (or conditioning facts) for such right or obligation. Because a party's individual consent is a necessary condition for ICSID jurisdiction, as already discussed, the class of rights and obligations arising out of such consent must therefore include the broad array of rights and obligations established by the Convention regarding the settlement of disputes and the conduct of proceedings.36
For a more detailed analysis, let us consider the different states of affairs that may exist in the ‘pre-notice’ period (ie the period preceding the time when the depositary received the notice of denunciation) regarding consent: (i) both parties to the dispute have consented to ICSID jurisdiction; (ii) only the denouncing State (or its constituent subdivision or agency) has consented; and (iii) only the national of the denouncing State has given its consent. For the sake of completeness, we shall also consider the case in which neither the denouncing State nor its national has given consent in the pre-notice period.
(p. 262 ) Consent given by both parties to the dispute
Let us first consider the simplest and least controversial case: both parties to the dispute have given their respective individual consents in the pre-notice period. This means that, both in the case of a dispute between the denouncing State and a national of another contracting State and in the case of a dispute between a national of the denouncing State and another contracting State, each party to the dispute has given its individual consent.
In each such case, Article 72 prescribes that the notice of denunciation shall not affect the rights or obligations under the Convention arising out of the consent given by the denouncing State or its national. For example, the denunciation will not affect the right of the denouncing State or its national to submit the dispute to the Centre, as long as the other requirements for ICSID jurisdiction are met, nor will it affect the procedural and other rights guaranteed to the parties by the Convention if the dispute is submitted to the Centre.37 The rights so preserved do not include, however, the right of the denouncing State or its national to withdraw the individual consent given in the pre-notice period, because that right was extinguished, under Article 25(1), at the time that individual consent was matched by the consent of the other party to the dispute.38 Likewise, the denunciation will not affect the obligations of the denouncing State and its national derived from their respective individual consents, including the obligation to suffer any submission of the dispute to the Centre by the other party to the dispute and, once the dispute has been submitted by either party, the obligation to abide by the procedures established by the Convention and to comply with the terms of the award as required by the Convention.39
Consent given only by the denouncing State
Most of the debate about Article 72 has centred, for good and obvious reasons, on the case in which the denouncing State (or any of its constituent subdivisions or agencies) gave its consent to ICSID jurisdiction in the pre-notice period but the investor that is the other party to the dispute did not give its matching consent by the time the depositary received the notice of denunciation. May the investor give its consent within the six-month period contemplated by Article 71? May it do so afterwards? May the denouncing State withdraw its consent at any time? Does the notice of denunciation operate as a withdrawal of the State's pre-notice consent? Let us separately consider the rights and the obligations arising out of the denouncing State's consent.
(p. 263 ) Rights arising out of the State's consent
Under Article 72, the notice of denunciation does not affect the rights or obligations under the Convention arising out of the State's pre-notice consent. Leaving aside for the moment the question whether the notice of denunciation necessarily wipes out the State's consent, the rights in question include the State's right to submit the dispute to the Centre, as long as the other party consents, the investor's right to give its own matching consent (as long as the State's consent has not been withdrawn), and the due-process rights guaranteed by the Convention to each party to the dispute.
One critical right arising out of the denouncing State's pre-notice consent is the right, based on Article 25(1) of the Convention, to withdraw that consent at any time before the other disputant has given its (matching) consent. Under Article 72, that right is unaffected by the notice of denunciation. As far as the Convention is concerned, then, the denouncing State retains the right to withdraw its consent until that right is extinguished according to Article 25(1). The question is, how may the denouncing State validly (or effectually) withdraw that consent? The Convention provides no answer. Just as the Convention contains no rules about how consent must be given (except that it must be in writing), it does not prescribe how consent may validly be withdrawn (except that it may not be withdrawn after the other party to the dispute has given its consent).40 The answer, accordingly, must be sought elsewhere.
The question of how a Contracting State (or a national of a Contracting State) may validly withdraw its consent does not exclusively arise in a denunciation context. Each party to the dispute has the right, under Article 25(1), to withdraw its individual consent before the other party has matched that consent, but the Convention does not specify how or in what circumstances such consent may be validly withdrawn. It is generally admitted that unmatched consent may not be withdrawn if it is made irrevocable by its own terms.41 It is also admitted, at least outside the denunciation context, that consent expressed in a treaty may not be withdrawn unilaterally except as permitted by the law of treaties.42 If that much is admitted outside the denunciation context, it is difficult to see why the case of denunciation should require a different result, as Article 72 merely preserves the rights and obligations under the Convention derived from pre-notice consent. For this reason, any theory on the effectual withdrawal of consent must apply to all cases of withdrawal, including withdrawal of consent in the context of denunciation of the Convention.
When the application of a treaty requires the filling of a lacunain its provisions, such as the absence of specific rules about withdrawal of consent, it is necessary to resort to the rules and principles of general international law.43 Chief among those principles are those (p. 264 ) of good faith and pacta sunt servanda.44 To determine how, in the light of those principles, a State may validly withdraw its consent, we must consider the various (written) forms in which that consent may have been given in the first place.
A State may give its consent to ICSID jurisdiction in essentially four different forms: (i) a clause in a contract, normally made with the investor that is the other party to the dispute; (ii) a unilateral instrument, such as a letter to the ICSID Secretariat or a letter to the investor; (iii) a piece of national (iedomestic) legislation, such as a law for the promotion of investments; or (iv) a bilateral or multilateral treaty, normally dealing with the promotion and protection of investments. Let us discuss each case separately.
Consent by contract
In the case of consent given by contract, usually (but not necessarily) the other party to the contract (and the dispute) would have given its consent as well, with the consequence that each party's right to withdraw its consent would have been extinguished at the time the contract was made and the case would have to be analysed as one of common consent. It would be unusual, but conceivable, for the State to consent to ICSID jurisdiction (eg for the benefit of an affiliate of the co-contractant) without a matching consent by the other party to the eventual dispute. In such cases, the question whether the State's consent may be validly withdrawn would be governed by the terms of the contract and the applicable law, as controlled by the international principle of good faith.45
Consent by unilateral instrument
The State may have given its consent to ICSID jurisdiction by means of a written unilateral communication to one or more investors or by a unilateral notice of consent delivered to the Centre or to the depositary. Such unilateral instruments of consent are not precluded by the Convention, and seem to have been contemplated by the drafters, but in fact they have been seldom, if ever, used.46 A unilateral statement of consent by notice to ICSID would be akin, in form and function, to a declaration of acceptance of the compulsory jurisdiction of the International Court of Justice under Article 36 of the Statute of the Court.47
In principle, consent given by unilateral instrument may be withdrawn in accordance with the terms of the same instrument, normally by a subsequent instrument of the (p. 265 ) same kind issued by a competent State authority. Exceptionally, such consent may not be withdrawn if, in the circumstances of the case, the instrument must be regarded as a binding unilateral declaration.48 This possibility, which arises from the recognition of binding unilateral declarations in the Nuclear Tests Case, is so dependent on the facts of each particular case that it need not be examined further here.49
(p. 266 ) It remains to consider whether the notice of denunciation operates automatically as withdrawal of the consent given by unilateral instrument. As there is nothing in the Convention prescribing that a notice of denunciation has the effect of withdrawing a pre-notice (non-matched) consent, the issue must be decided by considerations of good faith. If the consent was given in a direct unilateral communication to the investor, for example, the principle of good faith demands that the State be required to withdraw its consent by means of another direct communication of the same kind. On the contrary, if the State gave its consent by a generic notice to ICSID, it does not seem contrary to good faith to attribute withdrawal effects to the notice of denunciation, which is delivered to what is, in effect, ICSID's parent institution. The same reasoning would apply, a fortiori, if the consent had been given by a generic notice to the depositary.50
(p. 267 ) Consent by national legislation
The State may have given its consent to ICSID jurisdiction in a piece of domestic legislation, often a statute designed to promote foreign investment. In that context, the consent may have been given, as an incentive to foreign investors, in respect of a generality of investors and a generality of potential disputes. While it is tempting to regard consent by national legislation and any withdrawal thereof as exclusively domestic matters, doing so would be an error. Consent and withdrawal of consent, even by domestic legislation, are operative acts under the Convention; they are defined juridical acts (acts-in-the-law) that carry defined legal consequences under the Convention, just as the State's act of expressing its consent to be bound by a treaty (which is also a domestic act) is defined and carries legal consequences under the Vienna Convention and general international law.51 Therefore, the issue of withdrawal of consent given by legislation must be considered in the light of both national law and international law.52
(p. 268 ) From the standpoint of national law, consent given in a domestic piece of legislation may later be withdrawn in the form and to the extent required or authorized by higher norms such as the constitution of the State. As a matter of form, consent given by legislation would normally be withdrawn by a subsequent piece of legislation, of equal or higher rank, that repeals or supersedes the original provision. For example, if the consent was contained in a statute, it would normally be withdrawn by a later statute or a constitutional amendment. This case starkly illustrates why a notice of denunciation may not validly serve as withdrawal of consent given by statute. A notice of denunciation would normally be given by an executive department (such as the Ministry of Foreign Affairs), and it would be very unusual for the constitution of the denouncing State to allow such acts to repeal legislation. The extent to which the consent may be withdrawn would also depend on the substantive constraints that the national legal order may impose on the content of legislation. For example, if the constitution protects vested rights or legitimate expectations of foreign investors, the consent may be irrevocable as a matter of domestic law, and consequently any attempt to withdraw it may be ineffectual under the Convention.
From the standpoint of international law, the rules and principles of domestic law governing the withdrawal of consent by legislation would normally be given effect, except to the extent that they may contravene the obligations of the State under treaties or general international law, including in particular the principle of good faith. Depending on the circumstances, good faith may require that a particular attempt to withdraw consent given by legislation be disregarded as having no international effects.53
Consent by treaty
Consent by treaty is at present the most common form of consent to ICSID jurisdiction given by a State. Such consent takes the form of a specific clause to that effect in a bilateral or multilateral treaty, usually a treaty for the promotion and protection of investments. Consent to ICSID jurisdiction is generally regarded as a valuable benefit conferred by each State party to the treaty on the nationals of the other States parties. The effects of such clauses depend on their terms, as interpreted in accordance with the accepted rules of treaty interpretation codified in the Vienna Convention. Consent clauses vary in scope and the extent to which the consent is subject to conditions.54
(p. 269 ) Whether and to what extent consent given by the State in a particular treaty is capable of being withdrawn depends, first and foremost, on the terms of the clause and the remaining provisions of the treaty. If the treaty provides consent to ICSID jurisdiction, then to the extent that the consenting State is bound by the treaty under international law, the consent will be binding on the State and hence irrevocable, unless the treaty itself provides otherwise. The irrevocability of such consent flows not from the ICSID Convention, which as we have seen says nothing about the conditions for valid withdrawal, but from the binding force of the treaty, which in turn results from the rule pacta sunt servanda under general international law.55
It follows that consent to ICSID jurisdiction given in a treaty may not be validly withdrawn except as provided in the treaty, as the treaty itself may be terminated in accordance with its provisions and general international law. For example, Article IX(4) of the Bolivia-United States bilateral investment treaty provides:56
Article IX(3) of the same treaty states:
Each Party hereby consents to the submission of any investment dispute for settlement by binding arbitration in accordance with the choice of the national or company under paragraph 3(a) (i), (ii) and (iii) or the mutual agreement of both parties to the dispute under paragraph 3(a) (iv). This consent and the submission of the dispute by a national or company under paragraph 3(a) shall satisfy the requirement of:
(a) Chapter II of the ICSID Convention (Jurisdiction of the Centre) and the Additional Facility Rules for written consent of the parties to the dispute…
Nothing in Article IX(4) of the Bolivia-United States treaty suggests that a State Party that gives its consent under that provision reserves the right to withdraw it unilaterally. The principle pacta sunt servanda would prevent any such unilateral withdrawal, absent the agreement of the other State Party, for as long the treaty remains in effect.
(a) Provided that the national or company concerned has not submitted the dispute for resolution under paragraph 2 (a) or (b),57 and that three months have elapsed from the date on which the dispute arose, the national or company concerned may submit the dispute for settlement by binding arbitration:
(i) to the Centre, if the Centre is available; or
(ii) to the Additional Facility of the Centre, if the Centre is not available…
(p. 270 ) Under customary international law, a treaty can be terminated in accordance with its provisions or by agreement of the parties.58 For example, Article XVI of the Bolivia-United States treaty provides that the treaty is to be in force for an initial period of ten years and thereafter it will remain in effect unless one of the States parties gives one year's written notice of non-renewal. Even after termination, the treaty will remain in effect for a further period of ten years from the date of termination in respect of any covered investments established or acquired prior to that date.59 This particular treaty was signed on 17 April 1998 and entered into force on 6 June 2001. Absent termination by agreement of the parties, the treaty will remain in effect in accordance with its terms.60
Therefore, as long as the obligations of the treaty remain in effect for the denouncing State, any attempt by it to withdraw its consent given in the treaty would be ineffectual.
Obligations arising out of the State's consent
Article 72 of the Convention does not affect the obligations of the denouncing State that arise out of its pre-notice consent. Those include the obligations arising from the pre-notice consent, together with any later consent validly given by the other party to the dispute and from any resulting submission of the dispute to the Centre, as well as the obligation to abide by the other duties imposed on the parties by the Convention, including compliance with the award.
As we have seen, Article 72 does not affect the denouncing State's right to withdraw its consent before the other party to the dispute has consented, but the validity of any such withdrawal depends, under the principles of international law, on the instrument by which that consent was given and other relevant circumstances. In those cases in which the consent may not be effectually withdrawn, and for as long as such consent remains in effect, the other party to the dispute will have the right to give its matching consent at any time and, if such consent is given, the denouncing State will be obligated to abide by the ICSID rules of procedure and to comply with the award. The other party to the dispute may give its consent at any time because (absent an effectual withdrawal of consent) the Convention puts no time limit on that party's right to consent or the denouncing State's obligations derived from its pre-notice consent.
In particular, there is no basis to suppose that the denouncing State's obligations arising out of its pre-denunciation consent end when the notice of denunciation takes effect under Article 71.61 Here we should recall the earlier discussion on the function of Article 72. This article qualifies the requirement of Article 25(1) that the parties to the dispute be a contracting State and a national of another contracting State. It does so by giving effect to pre-notice consent even though the denouncing State ceases to be a Contracting State when the notice of denunciation takes effect. Consequently, the obligations derived (p. 271 ) from any pre-notice consent that the State has not effectually withdrawn survive beyond the date on which the notice of denunciation takes effect under Article 71. The potential long-term survival of the State's consent-related obligations under the Convention follows from the content and function of Article 72, was specifically contemplated by the drafters in the particular case of common consent,62 and is fully consistent with the development of long-term investment treaties as the main repositories of State consent to ICSID jurisdiction.63
Consent given only by the national of the denouncing State?
Article 72 of the Convention applies equally to a national of the denouncing State. Let us call that person a ‘disputant national’. Article 72 provides that the notice of denunciation shall not affect the disputant national's rights or obligations under the Convention that arise out of the individual consent that the disputant national may have given before the notice was received by the World Bank. As the analysis of this case is parallel to that of the denouncing State, and indeed simpler, a summary discussion should suffice.
The rights of the disputant national that arise under the Convention out of the disputant national's pre-notice consent include the right to submit the dispute to the Centre, as long as the other contracting State that is a party to the dispute has given or gives its consent and the other requirements for ICSID jurisdiction are met. If the dispute is submitted to the Centre, the rights that are unaffected by the notice of denunciation will include the due-process rights guaranteed by the Convention to the parties in dispute. The disputant national's unaffected rights also include the right to withdraw its pre-notice consent, although that right can be expected to have a lesser practical value than in the case of the denouncing State. As the disputant national's pre-notice consent would have normally been given by unilateral instrument, it could be withdrawn by an instrument of the same kind, unless there are special circumstances indicating that the withdrawal must be considered invalid under the principles of international law. As the contracting State that is the other party to the dispute would ordinarily have no interest in contesting any such withdrawal, the preceding observation is merely theoretical.
The disputant national's obligations under the Convention arising from its pre-notice consent include the obligation to abide by the consequences of any consent given by the other Contracting State that is a party to the dispute. If the dispute is submitted to ICSID by either party, the disputant national will retain its obligations related to the dispute-resolution process, including the obligation to comply with the award.
Absence of consent by the denouncing State or its national
To complete the analysis, we must briefly consider the case in which neither the denouncing State nor any of its nationals has given its consent before the notice of denunciation was received by the World Bank. In such a case, Article 72 of the Convention does not apply and the rights and obligations of the parties to the respective dispute must be determined by reference to other provisions of the Convention as well as general international law.
As already noted, Article 71 of the Convention prescribes that the notice of denunciation shall take effect six months after it is received by the depositary. Accordingly, during (p. 272 ) that six-month period, the denouncing State continues to be a contracting State and its nationals continue to be nationals of a contracting State. Therefore, during that period, the denouncing State and its nationals continue to enjoy the rights and to be subject to the obligations that respectively pertain, under the Convention, to contracting States and to nationals of a contracting State.
In particular, however unrealistic the case might be, nothing in the Convention would debar the denouncing State from giving its consent to ICSID jurisdiction, during the six-month period following the notice of denunciation, in respect of a dispute with a national of another contracting State. Nor is there anything in the Convention that would prevent the national of the other contracting State from giving its own matching consent within the same period, if it has not been given earlier. If both parties to the dispute give their consent, either party may submit the dispute to the Centre, but only within the same six-month period. The reason is that, after the denunciation takes effect, the requirement of Article 25(1) that one of the parties to the dispute be a contracting State would no longer be met. Although that requirement is qualified by Article 72 for cases of pre-notice consent, no provision of the Convention imposes a corresponding qualification for cases of post-notice consent.
Similar considerations apply, mutatis mutandis, to the much more likely scenario of a national of the denouncing State that wishes to consent to ICSID jurisdiction before the denunciation takes effect. Within the six-month period in which that person will continue to be a national of a contracting State, that person will have the right to give consent to ICSID jurisdiction in respect of a dispute with another contracting State, whether that other State has previously given its own consent or not. If both the national of the denouncing State and the other contracting State have given their consent, the dispute may be submitted to the Centre, but only before the denunciation takes effect, because after that time the investor that is a party to the dispute will no longer be a national of a contracting State.
D. The Limits of the Contract Analogy
Legal concepts, such as ‘consent’, are indispensable components of any minimally developed legal system. Rule-makers depend on concepts to present legal systems in ways that can be understood and applied. In creating new rules, rule-makers continually devise new concepts and re-use and redefine old ones.
When a new concept is created by a rule-maker (eg a legislator or two or more States making a treaty), an interpreter, or the rule-maker itself, may attempt to explain it by stressing ways in which the new concept resembles an old, familiar one. This form of explanation by analogy is useful and extremely common in legal discourse.64 That is why the concept of ‘consent of both parties’, as it appears in Article 25(1) of the Convention, can profitably be explained by analogy to a contract. The consent of one party resembles (p. 273 ) an offer, because it may be withdrawn before the other party consents, and the consent of the other party resembles an acceptance, because once given it makes the first consent irrevocable. When both parties have consented, there is common (or ‘perfected’) consent, which resembles a contract in that it produces binding legal consequences. None of this is objectionable, especially taking into account that international arbitration is traditionally viewed as based, fundamentally, on the parties’ agreement to arbitrate. In this sense, viewing common consent as an agreement to arbitrate serves the additional explanatory purpose of inscribing the ICSID system within that broad tradition.
If resort to the contract analogy ended there, there would be no reason to object. The analogy is carried too far, however, when the interpreter attempts to extract legal consequences that are inconsistent with the rules governing the concept that is being explained. As we have seen, there is no basis in the Convention for interpreting the term ‘consent’ as if it meant, exclusively, common consent. On the contrary, there is ample evidence that the term ‘consent’ is used in the Convention to refer to a party's individual consent or, in the case of consent by both parties, to common consent. Consequently, an interpretation that assumes that ‘consent’ in Article 72 must mean common consent, even though the text clearly refers to the party's individual consent, is in error. The error may be the result of confusion, because the same word (‘consent’) is used to refer both to individual consent and (once both individual consents have been given) to common consent. No such confusion is possible between ‘offer’ or ‘acceptance’, on the one hand, and ‘contract’, on the other, because these are all different terms.65
Let us now turn to two possible variants of the approach just described. While none of them can be attributed to Professor Schreuer, it would be easy for an interpreter persuaded by the contract analogy, especially one educated in the civil-law tradition, to take the further steps to be presently discussed. One word of caveat for the reader trained in the common-law tradition: the types of conceptual reasoning described below may seem exotic and difficult to comprehend, but the fact is that they are extremely common in the civil-law world, for perfectly understandable reasons. Civil-law systems are more highly systematized than, for example, American law, in the sense that they tend to rely to a much greater extent on integrated subsystems (codes) that are internally organized according to legal norms and legal concepts hierarchically arrayed according to increasing degrees of generality. A system that relies so much on hierarchies of concepts can be expected to develop pathologies in the direction of excessive conceptualism, just as American law tends to develop pathologies in the direction of excessive policy-oriented pragmatism.
A first variant of a conceptualist approach would be to inquire into the ‘legal nature’ of the concept of ‘consent’. At the most basic (and most defensible) level, the inquiry would be a matter of classification: to what legal category (or higher concept) does ‘consent’ belong? Whenever a new concept or legal institution is created, the first question that a civil-law interpreter usually asks is whether it fits into a recognized legal category. If it does, then all the legal consequences attributed to that category by the system will automatically be applicable to fill any lacunaein the legal regime that came with the new concept. Thus, an (p. 274 ) interpreter may be tempted to inquire into the legal nature of ‘consent’ and conclude, for example, that ‘consent’ is a species of ‘contract’.66
If the analysis is presented in these terms, one may quibble with some of its practical aspects, but in principle there seems to be no major theoretical difficulty, as long as it is clear that (i) the ‘consent’ that is subsumed in the category of ‘contract’ is common consent, that is, the consent given by both parties to the dispute and (ii) this classification exercise does not apply to the individual consent referred to in the Convention. One pitfall of this mode of reasoning is that the interpreter may be carried away and conclude that ‘consent’, as used in Article 72, can only mean common consent (ie‘perfected’ consent), because that has been previously established to be the legal nature of the concept of consent. That conclusion would be unacceptable because, as already shown, it is inconsistent with the text of the Convention. A further danger of this type of analysis is the temptation to classify the new concept not on the basis of the characteristics attributed to it by the rules that created it, but on the basis of the interpreter's intuition into the ‘essence’ of the concept or (what usually amounts to the same thing) the characteristics that, in the eyes of the interpreter, the concept ought to have. This unvarnished version of conceptualism is closely related to the approach to be discussed next.
A still more radical version of a conceptualist approach might maintain that, whatever Article 72 says and whatever the intent of the drafters, the concept ‘consent’ has a separate ontological reality or, in plainer terms, a life of its own. So, the argument would go, the meaning of ‘consent’ in Article 72 should not be determined by reference to the ordinary meaning of the words used in the Convention, viewed in context or otherwise; it must be determined by examining the concept as a separate, independent, evolving entity, the examination to be guided by such extra-legal considerations as the interpreter may consider relevant or, as in the previous case, by his intuition into the essence of the concept. It is easy to see that, by detaching the concept from the legal rules in which it appears, this approach may lead wherever the interpreter's preferences may beckon.
Alf Ross, the eminent Danish philosopher and international legal scholar, offered a most elegant critique of this way of thinking in his famous essay ‘Tû-Tû’.67 Ross started by referring to certain practices of the Noît-cif tribe, a primitive people living on the Noîsulli Islands in the South Pacific, as described by the Illyrian anthropologist Ydobon.
And yet, Ross observes, the word ‘tû-tû’ has a function to fulfil in the ordinary language of the tribe. Consider these two statements:
This tribe, according to Mr. Ydobon, holds the belief that in the case of infringement of certain taboos—for example, if a man encounters his mother-in-law, or if a totem animal is killed, or if someone has eaten of the food prepared for the chief—there arises what is called tû-tû. The members of the tribe also say that the person who committed the infringement has become tû-tû. It is very difficult to explain what is meant by this. Perhaps the nearest one can get to an explanation is to say that tû-tû is conceived as a kind of dangerous force or infection which attaches to the guilty person and threatens the whole community with disaster. For this reason a person who has become tû-tû must be subjected to a special ceremony of purification.(p. 275 )
It is obvious that the Noît-cif tribe dwells in a state of darkest superstition. ‘Tû-tû’ is of course nothing at all, a word devoid of any meaning whatsoever. To be sure, the above situations of infringement of taboo give rise to various natural effects, such as a feeling of dread and terror, but obviously it is not these, any more than any other demonstrable phenomena, which are designated as tû-tû. The talk about tû-tû is pure nonsense.68
(1) If a person has eaten the chief's food, he is tû-tû.
(2) If a person is tû-tû, he shall be subjected to a ceremony of purification.
(3) If a person has eaten the chief's food, he shall be subjected to a ceremony of purification.
This explains why tû-tû statements are meaningful as used prescriptively by the tribe or descriptively by an anthropologist, in each case in the context of the tribe's taboo system. Even though tû-tû does not refer to anything, or more precisely to anything external to the system, it serves as a useful technique to link a disjunctive set of conditioning facts (that a person has eaten the chief's food or has encountered his mother-in-law, etc) with a set of consequences imposed by the taboo system (that the person must be subjected to a ceremony of purification).
Ross notes that the tû-tû formulation, in spite of these technical advantages, ‘could in certain cases lead to irrational results if against all better judgment the idea that tû-tû is a reality is allowed to exert its influence’.69 In such a case, continues Ross, ‘it must be the task of criticism to demonstrate the error and to cleanse one's thinking of the dross of such imaginary ideas. But even so there would be no grounds for giving up the tû-tû terminology.’70
At this point Ross reveals the hoax and explains that the tû-tû allegory concerns ourselves and our use of legal concepts. Like tû-tû, says Ross, legal concepts are devoid of semantic reference; they are empty words inserted between a set of conditioning facts and a set of legal consequences, serving solely as convenient tools for organizing, systematizing, and simplifying the presentation of the rules of the system.71 I would prefer to say that typical legal concepts have no independent meaning, and certainly no independent existence, outside the legal system to which they refer.72 A legal concept such as ‘consent’ is like tû-tû; statements including the word ‘consent’ in reference to the Convention have no semantic reference outside the system of the Convention. The extreme conceptualism (p. 276 ) that attributes to such concepts a life of their own has no better foundation than the superstitious beliefs of the fictional Noît-cif.
The preceding critique is directed against conceptualist approaches because those are the paths that the followers of the contract analogy might more easily be tempted to tread. But it is worth noting that, just as conceptualism may lead to ignoring the Convention, similar results could be reached through a diametrically opposite way of conceiving the task of interpretation. If treaty interpretation is seen, not as a rule-bound exercise to ascertain the meaning of rules, but fundamentally as a matter of competing policies, or a power struggle to instil meaning in a deconstructed text, or an exercise in balancing opposing interests, or an attempt to shape the treaty to fit new circumstances or needs, all as perceived by the interpreter—in short, if the treaty itself does not matter and can be safely tossed aside and replaced by anything-goes policy-oriented instrumentalism (or nihilism)—then ‘consent’ in Article 72, and indeed the Convention as a whole, may well acquire ‘a life of its own’ and come to mean whatever the interpreter wishes it to mean. A criticism of these views at the level of meta-theory would greatly exceed the scope of this chapter. Suffice it to say, at the level of theory, that this approach to interpretation has little or nothing to do with the interpretation that has traditionally been conducted, and we are directed to conduct, under the Vienna Convention.73
Article 72 of the Convention must be interpreted in good faith, in accordance with the ordinary meaning of its terms in context and in the light of the object and purpose of the Convention. Applying this canon, we have concluded that Article 72 means exactly what it says: the notice of denunciation does not affect the rights and obligations under the Convention arising out of the individual consent to ICSID jurisdiction given by the denouncing State or its national before the notice was received by the depositary. When those rights and obligations are closely analysed, the operational questions posed at the beginning of this chapter can be readily answered.
If the investor has matched the State's pre-notice consent before the depositary received the State's notice of denunciation, either party may submit the dispute to the Centre and both parties have the continuing obligation to abide by the ICSID process and to comply with the award. In principle, the denouncing State retains the right to withdraw any pre-notice consent that the investor has not yet matched, but the means of effectually withdrawing such consent depend on the means through which the consent was given in the first place and the general principles of international law. In particular, consent given by treaty may be withdrawn only to the extent permitted by the treaty or through termination of the treaty in accordance with its provisions and general international law. The investor may give its own consent at any time until the State's consent is effectually withdrawn and, once the consent is matched, either party may submit the dispute to the Centre. Finally, there is no justification for the thesis that the notice of denunciation may serve as withdrawal of the State's pre-notice consent, except perhaps (p. 277 ) in the exceptional case in which that consent was expressed in a communication to ICSID or the World Bank.
In closing, let us return to the theme of the manifest political tension between denunciation and consent. From a policy standpoint, a State that has given its individual consent, by contract, unilateral declaration, legislation or treaty, has accepted the possibility of having a dispute or a category of disputes submitted to ICSID arbitration. Also as a matter of policy, a State that has denounced the Convention has done so because it no longer wishes to be subject to ICSID arbitration. As already noted, the point of interpretation is not to find a ‘balance’, whatever that may mean, between the old pro-ICSID arbitration policy and the new anti-ICSID arbitration policy, or between the interests of the State, however they may be ascertained, and the interests of the investor that is the other party to the dispute. A policy calculus is interesting, and important de lege ferenda, but it is no substitute for a legal analysis such as that proposed in this chapter.
But even within the terms of a policy calculus, it does not follow that a current policy against ICSID arbitration should necessarily trump an earlier policy of consent to ICSID arbitration. If the current wishes or intent of a party could trump a prior inconsistent intent to which the system has attached legal consequences, mankind would not have progressed beyond instantaneous mutual exchanges. No system of law can work if binding promises are binding only for so long as the promissor has not changed its mind. That is why the binding force of a treaty does not depend on a State Party's present intent to be bound, but rather on the intent to be bound that the State properly expressed at the time it became a party to the treaty.74 Therefore, the theory that denunciation necessarily trumps prior consent is not only wrong as a matter of law, as discussed in the body of this work, but it is also inconsistent with the most basic policies underlying the international legal system.
(*) The views expressed in this chapter are solely those of the author and do not necessarily reflect the positions of Covington & Burling LLP or any past or present client thereof. The author gratefully acknowledges the able assistance of his colleague Luisa F. Torres in the research for this chapter. For information on the author, the reader is invited to review the author's biographical note at <http://www.cov.com>.
(1) J.S. Mill, A System of Logic (8th edn, 1900) 554.
(2) Horace, Ars Poetica, lines 358–9.
(3) See 46 ILM (2007) 969. The ICSID Convention is the Convention on the Settlement of Investment Disputes between States and Nationals of Other States, done at Washington (1965 (henceforth ICSID Convention or the Convention), in ICSID Convention, Regulation and Rules (2006) 11. Bolivia's denunciation of the Convention is a further instance of what I have called the rediscovery of the Calvo Doctrine. See O.M. Garibaldi, ‘Carlos Calvo Redivivus: The Rediscovery of the Calvo Doctrine in the Era of Investment Treaties’, in Proceedings of the 57th Annual Conference on Oil and Gas Law (2006); reprinted in 3(5) Transnational Dispute Management (TDM) (2006).
(4) Article 71 ICSID Convention, above n. 3.
(5) Bolivia is a party to at least seventeen bilateral investment treaties containing ICSID arbitration clauses. This statement is based on information from various sources, which are not always consistent with one another. See United Nations Conference for Trade and Development, Collection of Investment Instrument Online (henceforth UNCTAD BIT Collection), available at <http://www.unctad.org/sections/dite_pcbb/docs/bolivia.pdf> (last accessed 22 May 2008), reporting that as of 1 June 2007 Bolivia had signed twenty-two bilateral investment treaties, nineteen of which were in force. Fifteen of the treaties that UNCTAD reports as being in force include ICSID arbitration clauses. See also ICSID Database of Bilateral Investment Treaties (henceforth ICSID BIT Collection), available at <http://icsid.worldbank.org/ICSID/FrontServlet> (last accessed 23 May 2008), reporting that Bolivia has signed twenty-three bilateral investment treaties, nineteen of which are in force. Seventeen of the treaties that ICSID reports as being in force include ICSID arbitration clauses. (The text of the treaty with Denmark, reportedly in force, was unavailable for review.) See also R. Castro de Figueiredo, ‘Euro Telecom v. Bolivia: The Denunciation of the ICSID Convention and ICSID Arbitration under Bits’, Transnational Dispute Management (18 March 2008), Provisional Issue, no. 4, reporting that Bolivia is a party to twelve treaties with ICSID arbitration clauses, citing <http://www.sice.org/ctyindex/BOL/BOLBITs_e.asp>. For an analysis of ICSID arbitration clauses in selected treaties to which Bolivia is a party, see C. Tietje, K. Nowrot, and C. Wackernagel, ‘Once and Forever? The Legal Effects of a Denunciation of ICSID’, Transnational Dispute Management (18 March 2008), Provisional Issue, 28–30.
(6) See ETI Euro Telecom International NV v Republic of Bolivia, ICSID Case No. ARB/07/28. The request for arbitration was registered by the ICSID Secretariat on 31 October 2007. See <http://icsid.worldbank.org/ICSID/FrontServlet> (last accessed 20 May 2008).
(7) Article 25(1) ICSID Convention, above n. 3.
(8) See eg C. Schreuer, The ICSID Convention: A Commentary (2001) (henceforth Schreuer, Commentary) 218–19 (Article 25, paras 302–4). See also Ibid 1286 (Article 72, para 4); C. Schreuer, ‘Consent to Arbitration’, 2(5) Transnational Dispute Management (2005) (henceforth Schreuer, ‘Consent I’), 7; C. Schreuer, ‘Consent to Arbitration’ (27 February 2007) (henceforth Schreuer, ‘Consent II’) available at <http://www.univie.ac.at/intlaw/con_arbitr_89.pdf> (last accessed 20 May 2008), 7–8.
(9) Article 72 ICSID Convention, above n. 3.
(10) ‘Art. 72 is an expression of the rule, contained in Art. 25(1), that consent, once given, cannot be withdrawn unilaterally […] The rights and obligations arising out of an agreement to consent to ICSID jurisdiction are preserved and insulated from later legal developments.’ Schreuer, Commentary, above n. 8, 1285 (Article 72, para 2).
‘In order to benefit from the continued validity under Art. 72, consent must have been given before the denunciation of the Convention. […] Consent is only perfected after it has been accepted by both parties. Therefore, a unilateral offer of consent by the host State through legislation or a treaty before a notice under Art […] 71 would not suffice. The effect of the continued validity of consent under Art. 72 would only arise if the offer was accepted in writing by the investor before the notice of denunciation […].’, Ibid. In a more recent paper, Professor Schreuer has expressed the following view:
‘In the case of arbitration clauses contained in treaties a possible withdrawal of an offer of consent before its acceptance is less of a problem than in the case of national legislation. An offer of arbitration in a treaty remains valid notwithstanding an attempt to terminate it, unless there is a basis for termination under the law of treaties. Nevertheless, in order to avoid complications early acceptance is advisable also in the case of consent contained in BITs. Once the arbitration agreement is perfected through the acceptance of the offer contained in the treaty it remains in existence even if the States parties to the BIT agree to amend or terminate the treaty.’ Schreuer, ‘Consent II’, above n. 8, 8 (emphasis added). If consent expressed in a treaty remains valid notwithstanding an attempt to terminate it, there is no good reason for that validity to be extinguished by a notice of denunciation. At a minimum, the passage just quoted casts doubts on Professor Schreuer's current views on the effects of Article 72. The community of those who practise international investment law eagerly awaits a second edition of his Commentary.
(12) For example, Professor Emmanuel Gaillard has taken the view that, if ‘consent’ in Article 72 had been intended to refer to an agreement to consent (‘perfected’ consent, in Professor Schreuer's terminology), the drafters would have so provided. E. Gaillard, ‘The Denunciation of the ICSID Convention’, New York Law Journal (26 June 2007). Professor Gaillard concludes that unqualified consent given by treaty is protected by Article 72 and survives denunciation of the Convention. Ibid. Michael Nolan and F.G. Sourgens note that Professor Gaillard's approach amounts to re-characterizing the State's ‘offer’ of consent as a ‘firm offer’, that is, an offer that is irrevocable. See M.D. Nolan and F.G. Sourgens, ‘The Interplay between State Consent to ICSID Arbitration and Denunciation of the ICSID Convention: The (Possible) Venezuela Case Study’, Transnational Dispute Management (September 2007), Provisional Issue (henceforth Nolan and Sourgens). Nolan and Sourgens propose, as an alternative to the contract analogy, what they call the ‘international obligation approach’, Ibid. This amounts to the view that consent expressed by treaty or domestic legislation is binding under international law. Ibid. The analysis offered by Nolan and Sourgens places less emphasis on the content of Article 72 than on the extra-Convention legal bases for the binding effect of consent expressed in treaties or legislation. Professor Christian Tietje, Dr Karsten Nowrot, and Mr Clemens Wackernagel also conclude that the State's consent given by treaty is binding for the life of the treaty, but do so by reinterpreting the term ‘obligations under the Convention’ in Article 72 on the basis of ‘the principle of dynamic-evolutionary treaty interpretation’, which requires treating the binding consent given in the treaty as if it were an obligation under the Convention. See Tietje et al, above n. 5. Sébastien Manciaux has taken the view, without analysis, that in the case of pre-denunciation consent by the denouncing State, the investor may give its matching consent at any time until the notice of denunciation takes effect under Article 71. Sébastien Manciaux, ‘La Bolivie se retire du CIRDI’, 4(5) Transnational Dispute Management (2007).
(13) Castro de Figueiredo uncritically accepts the contract analogy as a starting point for his analysis which, I regret to say, is deeply flawed. See Castro de Figueiredo, above n. 5, passim. For an example of the flaws, see below n. 32. Julien Fouret also starts from the assumption that ‘consent’ in Article 72 can only mean common consent, and devotes a lot of effort to proving the obvious: that consent expressed by a State in a bilateral investment treaty is consent by the State, not the common consent of the parties to the dispute. See J. Fouret, ‘Denunciation of the Washington Convention and Non-Contractual Investment Arbitration: “Manufacturing Consent” to ICSID Arbitration?’ 25 Journal of International Arbitration (2008) 71.
(14) See Tietje et al, above n. 5, 22 et seq. For a criticism of this view, see below n. 34.
(15) Vienna Convention on the Law of Treaties (1969), 1155 UNTS 331 (henceforth Vienna Convention),
(16) For a comprehensive discussion of the historical and political background of the issue and the relevant travaux préparatoires, see especially Schreuer, Commentary, 1284–6 (Article 71, paras 1–3; Article 72, paras 1–6); Tietje et al, above n. 5, 13–28; and Nolan and Sourgens, above n. 12, 5–20.
(17) Tietje et al advocate a ‘balanced approach’ to interpretation, citing a few decisions. See Tietje et al, above n. 5, 14. ‘Balanced approach’ is one of those concepts that look unobjectionable until one realizes that they are ambiguous and one of its possible meanings embodies all manner of mischief. If by ‘balanced approach’ is meant only that the interpreter ought to apply the rules of interpretation of the Vienna Convention without a pre-conceived bias in favour of the State or the investor, the proposal is not only unobjectionable but devoutly to be endorsed. This was the approach adopted in Azurix v Argentine Republic, ICSID Case No. ARB/01/12, Award, 14 July 2006 (Rigo Sureda (President), Lalonde, Martins), <http://ita.law.uvic.ca/alphabetical_list.htm>, para 307, in which the tribunal said: ‘The Tribunal does not consider that the BIT should be interpreted in favor or against the investor. The BIT is an international treaty and should be interpreted in accordance with the interpretation norms set forth by the Vienna Convention on the Law of the Treaties (“the Vienna Convention”), which is binding on the States parties to the BIT. Article 31(1) of the Vienna Convention requires that a treaty be “interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in light of its object and purpose.”?’ See also Ceskoslovenska obchodni banka, a.s. v Slovak Republic, ICSID Case No. ARB/97/4, Award on Jurisdiction, 24 May 1999) (Buergenthal (President), Bernardini, Bucher), <http://ita.law.uvic.ca/chronological_list.htm>, para 34. But if the call for a ‘balanced approach’ means that the interpreter must strike a balance between the interests or positions of the State and the investor—a more sophisticated version of the time-honoured arbitral practice of ‘splitting the baby’—I must disagree. It is not the function of arbitral tribunals to strike a balance between the parties’ interests or positions; it is to interpret the treaties in good faith and under the accepted rules to give effect to the balance struck by the Contracting States. A statement in Saluka Investments BV (The Netherlands) v The Czech Republic, UNCITRAL Arbitration, Partial Award, 17 March 2006 (Watts (President), Fortier, Behrens), <http://ita.law.uvic.ca/chronological_list.htm>, para 300, is sometimes cited in support of this second view, but in that case the tribunal did not purport to lay down a general approach to interpretation; it merely took into consideration that the preamble to the treaty referred to other broad economic aims alongside that of encouraging foreign investment and ruled that, to take into account the object and purpose of the treaty, all those aims had to be balanced in the process of interpretation.
(18) See ‘Report of the Executive Directors on the Convention on the Settlement of Investment Disputes between States and Nationals of Other States’ (18 March 1965), in ICSID Convention, Regulation and Rules (2006) 37, para 23; Schreuer, Commentary, 191 (Article 25, para 241). For a general discussion of consent, using the contract analogy as a framework, see Schreuer, ‘Consent I’ and Schreuer, ‘Consent II’, above n. 8.
(19) Article 25(1) ICSID Convention, above n. 3 (emphasis added). For a history of the drafting of this provision, see Nolan and Sourgens, above n. 12, 5–13.
(20) Conditions (i) and (ii) are sometimes referred to as requirements for jurisdiction ratione materiae and condition (iii) as a requirement for jurisdiction ratione personae. Condition (iv) is sometimes referred to as a requirement for jurisdiction ratione voluntatis; at other times as a requirement for jurisdiction ratione materiae. The requirement that the dispute fall within the temporal sphere of application of the Convention or the instrument expressing a party's consent is sometimes referred to as a basis for jurisdiction ratione temporis.
(21) The term ‘common’ is used here in the sense of ‘held in common’.
(22) Article 25(1) ICSID Convention, above n. 3 (emphasis added).
(23) While the drafters often explained the provisions of the Convention using the contract analogy, the very use of the terms ‘offer’ and ‘acceptance’ interchangeably with ‘consent of a party to the dispute’ shows that ‘consent’ was not always used in the sense of common consent. For example, the drafters made clear that the consent of both parties need not be expressed in a single instrument and that each disputant may give its individual consent separately. As Mr Broches put it, ‘a host state might in its investment promotion legislation offer to submit disputes arising out of certain classes of investments to the jurisdiction of the Centre, and the investor might give his consent by accepting the offer in writing’. ‘Memorandum from the General Counsel and Draft Report of the Executive Directors to Accompany the Convention’ (19 January 1965), in History of the ICSID Convention (1970), volume II-2, 956 (emphasis added). If ‘consent’ had been intended always to mean common consent, the drafters would not have referred to the investor's consent (‘his consent’).
(24) Article 71 ICSID Convention, above n. 3. For a summary history of the drafting of Articles 71 and 72, see Nolan and Sourgens, above n. 12, 13–17.
(25) Article 54(a) Vienna Convention, above n. 15.
(26) Articles 71 and 73 ICSID Convention, above n. 3.
(27) Article 72 ICSID Convention, above n. 3.
(28) Article 70 provides: ‘The Convention shall apply to all the territories for whose international relations a Contracting State is responsible, except those which are excluded by such State by written notice to the depositary of this Convention either at the time of ratification, acceptance or approval or subsequently’. A notice under Article 70 is unrelated to denunciation of the Convention.
(29) Article 72 ICSID Convention, above n. 3 (emphasis added).
(30) See eg Schreuer, Commentary, above n. 8 (Article 72, paras 1, 4) 1285–6. See also Fouret, above n. 13, 76.
(31) At the session of 3 March 1965, Mr Mejía-Palacio ‘asked what would happen if a State which was a party to the Convention signed an agreement with a company and later withdrew from the Centre while no disputes were pending. If, say, ten years later a dispute arose—would that dispute still be under the jurisdiction of the Centre?’ Mr Broches replied ‘that if the agreement with the company included an arbitration clause and that agreement lasted for say 20 years, the State would still be bound to submit its disputes with that company under that agreement to the Centre’. History of the ICSID Convention (1968), volume II-2, 1009–10. Earlier at the same session, Mr Malaplate had ‘stated that […] the Article [article 73, currently Article 72] as it stood, now could not seem to protect a consent given before the denunciation or exclusions if proceedings under that consent had not yet started’. Mr Broches had ‘replied that the intention under Article 73 in the text submitted to the Directors was to make it clear that if a State had consented to arbitration, for instance by entering into an arbitration clause with an investor, the subsequent denunciation of the Convention by that State would not relieve it from its obligation to go to arbitration if a dispute arose’. Ibid 1009.
(32) Tietje et al, above n. 5, 7, reach the same conclusion. See also, Schreuer, Commentary, above n. 8, 1285 (Article 72, para 3).
(33) Castro de Figueiredo criticizes this position as follows: ‘According to those who support the idea that the consent given under BITs is not affected by the denunciation of the ICSID Convention, Article 72 only refers to the consent given by one party to the dispute. Therefore, regardless of the consent of the investor, the consent given by the host State would be preserved after the denunciation. However, the allusion to “one of them” seems to refer to “State or any of its constituent subdivisions or agencies or of any national of that State,” and not to one of the disputing parties. The rule contained in Article 72 intends to preserve not only the consent given by the host State, but also the consent given by the investor.’ Castro de Figueiredo, above n. 5, 9. Castro de Figueiredo realizes that the ‘investor’ in question is the national of the denouncing State (as his next sentence refers to ‘a Bolivian investor’), but he seems not to realize that he has not only failed to answer the position he criticizes, but in fact has confirmed that the ‘consent’ referred to in Article 72 is the individual consent of the denouncing State or its national.
(34) In the case of a dispute to which the denouncing State is a party, the other party to the dispute must be a national of another contracting State. In the case of a dispute to which a national of the denouncing State is a party, the other party to the dispute must be a contracting State.
(35) Tietje et al argue that consent expressed by a State in a treaty is binding and irrevocable and that this conclusion is ‘strongly indicated’ by the object and purpose of the Convention. Tietje et al, above n. 5, 22. Yet, these authors believe that their theory is inconsistent with the reference to ‘rights and obligations under this Convention’ in Article 72, because the rights derived from the treaty are not rights under the Convention. To solve the problem, they speak of a ‘need for a reconceptualized understanding of the scope of “obligations under the Convention” unaffected by a denunciation’. Ibid, 24. This amounts to a proposal to reinterpret ‘obligations under the Convention’, in accordance with what they call the ‘dynamic-evolutionary theory of interpretation’, to include consent given by treaty. This manoeuvre is unnecessary. As explained in the text, below, there is no contradiction between the text of Article 72 and the conclusion that consent given in a treaty is effectual for as long as the provision containing it is in force. For a criticism of the ‘dynamic-evolutionary theory of interpretation’, which may just as soon be invoked to support the authors’ theory as a diametrically opposite one, see below. Ironically, the authors hint at an alternative line of argument that is not very different from the one developed in this chapter. See Tietje et al, above n. 5, 27.
(36) Not every right or obligation created by the Convention arises out of consent to ICSID jurisdiction. For example, a contracting State has the right to designate four persons to the Panel of Conciliators and four to the Panel of Arbitrators, regardless of whether it has consented to ICSID jurisdiction. See Article 13(1). Likewise, a contracting State is obligated to recognize and enforce an award rendered under the Convention against another contracting State even if it has never given its consent to ICSID jurisdiction. See Article 54(1).
(37) Schreuer, Commentary, above n. 8 (Article 72, para 6) 1286.
(38) Of course, in the unusual case in which the terms of the common consent permit one party to withdraw its consent unilaterally, such terms will be effectual. At the meeting of the Committee of the Whole held on 25 February 1965, Mr Mejía Palacio inquired about those cases in which the ‘agreement had no definitive duration but provided that they could be terminated by denunciation’. Mr Broches ‘remarked that in the case of an arbitration clause which could be terminated by one of the parties, the jurisdiction of the Centre would come to an end on termination of the clause’. History of the ICSID Convention (1968), volume II-2, 1010.
(39) See above n. 30; Schreuer, Commentary, above n. 8 (Article 72, para 6) 1286. Additional obligations arising from common consent are those resulting from the rule that (unless otherwise stated) such consent excludes any other remedy. See Article 26 ICSID Convention, above n. 3.
(40) Article 25(1) ICSID Convention, above n. 3. Article 25(4) indicates that a statement that the State would consider submitting a class of cases to ICSID jurisdiction does not constitute the consent required by Article 25(1).
(41) See Schreuer, ‘Consent to Arbitration’, in UNCTAD, Dispute Settlement, International Centre for Settlement of Investment Disputes (2003), chapter 2.3, 37.
(42) See Schreuer, ‘Consent II’, above n. 8, 8, text quoted above n. 11. See also Articles 42(2), 54 Vienna Convention, above n. 15.
(43) Article 31(3)(c) Vienna Convention, above n. 15.
(44) Ibid, Article 26. On the application of the principle of good faith to determine ICSID jurisdiction, see Inceysa Vallisoletana SL v Republic of El Salvador, ICSID Case No. ARB/03/26, Award, 2 August 2006 (Oreamuno (P), Landy, von Wobeser), <http://ita.law.uvic.ca/chronological_list.htm>.
(45) For example, if the contract was valid and binding under the law of the State at the time it was made, the principle of good faith would prevent the State from withdrawing its consent by retroactively cancelling or annulling the contract.
(46) See below n. 48. Article 25(4) of the Convention suggests that the drafters contemplated the possibility of unilateral statements of consent by notice to the Centre. Article 25(4) provides that a contracting State may, ‘at the time of ratification, acceptance or approval of the Convention or at any time thereafter, notify the Centre of the class or classes of disputes which it would or would not consider submitting to the jurisdiction of the Centre. […] Such notification shall not constitute the consent required by paragraph (1).’ (Emphasis added.) A statement that the State would only consider submitting a class of cases to ICSID jurisdiction naturally does not amount to consent, but the caveat would have been unnecessary if the drafters had foreclosed the possibility of a contracting State consenting to ICSID jurisdiction by means of unilateral notice to the Centre.
(47) Article 36, paras 2 to 5 Statute of the International Court of Justice, available at <http://www. icj-cij.org/documents/index.php?p1=4&p2=2&p3=0>.
(48) The International Law Commission has articulated the conditions under which unilateral declarations of States are the source of legal obligations under customary international law. See Guiding Principles Applicable to Unilateral Declarations of States Capable of Creating Legal Obligations, in United Nations, Report of the International Law Commission Fifty-Eighth Session (1 May–9 June and 3 July–11 August 2006), General Assembly, Official Records Sixty-First Session, Supplement No. 10, UN Doc. A/61/10, in Yearbook of the International Law Commission (2006), volume II-2, also available at <http://daccessdds.un.org/doc/UNDOC/GEN/G06/636/20/PDF/G0663620.pdf?OpenElement>, 362 et seq. Those guiding principles ‘relate only to unilateral acts stricto sensu, i.e. those taking the form of formal declarations formulated by a State with the intent to produce obligations under international law’ and are summarized as follows:
(‘1.) Declarations publicly made and manifesting the will to be bound may have the effect of creating legal obligations. When the conditions for this are met, the binding character of such declarations is based on good faith; States concerned may then take them into consideration and rely on them; such States are entitled to require that such obligations be respected;
(2.) Any State possesses capacity to undertake legal obligations through unilateral declarations;
(3.) To determine the legal effects of such declarations, it is necessary to take account of their content, of all the factual circumstances in which they were made, and of the reactions to which they gave rise;
(4.) A unilateral declaration binds the State internationally only if it is made by an authority vested with the power to do so. By virtue of their functions, heads of State, heads of Government and ministers for foreign affairs are competent to formulate such declarations. Other persons representing the State in specified areas may be authorized to bind it, through their declarations, in areas falling within their competence;
(5.) Unilateral declarations may be formulated orally or in writing;
(6.) Unilateral declarations may be addressed to the international community as a whole, to one or several States or to other entities;
(7.) A unilateral declaration entails obligations for the formulating State only if it is stated in clear and specific terms. In the case of doubt as to the scope of the obligations resulting from such a declaration, such obligations must be interpreted in a restrictive manner. In interpreting the content of such obligations, weight shall be given first and foremost to the text of the declaration, together with the context and the circumstances in which it was formulated;
(8.) A unilateral declaration which is in conflict with a peremptory norm of general international law is void;
(9.) No obligation may result for other States from the unilateral declaration of a State. However, the other State or States concerned may incur obligations in relation to such a unilateral declaration to the extent that they clearly accepted such a declaration;
(10.) A unilateral declaration that has created legal obligations for the State making the declaration cannot be revoked arbitrarily. In assessing whether a revocation would be arbitrary, consideration should be given to:
(49) In the Nuclear Tests case, the International Court of Justice recognized, inter alia, the following principles: (i) ‘[…] declarations made by way of unilateral acts, concerning legal or factual situations may have the effect of creating legal obligations […] When it is the intention of the State making the declaration that it should become bound according to its terms, the intention confers on the declaration the character of legal undertaking […] An undertaking of this kind, if giving publicly, even though not made within the context of international negotiations, is binding […] nothing in the nature of a quid pro quo nor any subsequent acceptance of the declaration, nor even any reply or reaction from other States, is required for the declaration to take effect […]’, Nuclear Tests (Australia v France; New Zealand v France), Judgments, 20 December 1974, ICJ Reports (1974), 457, para 43; (ii) ‘[…] not all unilateral acts imply obligation; but a State may choose to take up certain position […] with the intention of being bound—the intention is to be ascertained by interpretation of the act. When States make statements by which their freedom of action is to be limited, a restrictive interpretation is called for’, Ibid, para 44; (iii) ‘Whether a statement is made orally or in writing makes no essential difference […]’, Ibid, para 45; (iv) ‘One of the basic principles governing the creation and performance of legal obligations, whatever their source, is the principle of good faith. Trust and confidence are inherent in international co-operation […] Just as the very rule of pacta sunt servanda in the law of treaties is based on good faith, so also is the binding character of an international obligation assumed by unilateral declarations […] interested States may take cognizance on unilateral declarations and place confidence in them, and are entitled to require that the obligations thus created be respected’, Ibid, para 46. See also Ibid, para 51.
(50) These cases of unilateral consent may have been what some of the drafters had in mind when they suggested that denunciation (not a notice of denunciation) might have the effect of withdrawing an unmatched general declaration of consent. At the meeting of the Committee of the Whole held on 25 February 1965, Mr Gutiérrez Cano referred to ‘the case in which there was no agreement between the State and the foreign investor but only a general declaration on the part of the State in favor of submission of claims to the Centre and a subsequent withdrawal from the Convention by that State before any claim had been in fact submitted to the Centre. Would the Convention still compel the State to accept the jurisdiction of the Centre?’, History of the ICSID Convention (1968), volume II-2, 1010 (emphasis added). Mr Broches replied that ‘a general statement of the kind mentioned by Mr. Gutiérrez Cano would not be binding on the State which had made it until it had been accepted by the investor. If the State withdraws its unilateral statement by denouncing the Convention before it has been accepted by the investor, no investor could later bring a claim before the Centre. If however, the unilateral offer of the State has been accepted before the denunciation, then disputes arising between the State and the investor after the date of denunciation will still be within the jurisdiction of the Centre.’ Ibid (emphasis added). This colloquy is sometimes read as suggesting that a notice of denunciation was intended to wipe out the denouncing State's prior consent given by legislation or treaty. See eg Tietje et al, above n. 5, 25; Nolan and Sourgens, above n. 12, 15–16; see also Schreuer, Commentary, above n. 8, 1286 (Article 72, paras 4–5), reaching the same conclusion without citing this passage. The exchange, however, does not support that interpretation. First, Mr Gutiérrez Cano referred to a ‘general declaration’, which Mr Broches understood to be a reference to a ‘unilateral statement’ or ‘unilateral offer’. There is no indication that either speaker was referring to State consent given by legislation or treaty. While in some sense all unmatched State consent is unilateral and may be general in scope, it is hard to accept that the speakers intended to disregard the critical features that distinguish legislation (which is created and repealed by complex constitutional processes) and treaties (which are created and terminated by complex bilateral or multilateral international processes) from simple unilateral statements, which are more akin to the executive acts required to denounce the Convention. Second, Mr Broches referred to withdrawal of the unilateral statement ‘by denouncing the Convention’, not by giving notice of denunciation. Under Article 71, the denunciation takes place six months after receipt of the notice. Under the rules of the Vienna Convention, an inconclusive exchange like this cannot override the ordinary meaning of the terms of Article 72. Under Article 32 Vienna Convention, above n. 15, recourse may not be had to the preparatory work to determine the meaning unless the interpretation resulting from applying the general rules of Article 31 leaves the meaning ambiguous or obscure or leads to a result that is manifestly absurd or unreasonable.
(51) Article 46(1) Vienna Convention, above n. 15, ‘A State might not invoke the fact that its consent to be bound by a treaty has been expressed in violation of a provision of its internal law regarding competence to conclude treaties as invalidating its consent unless that violation was manifest and concerned a rule of its internal law of fundamental importance’. See also, Ibid, Article 27, ‘A party may not invoke the provisions of its internal law as justification for its failure to perform a treaty. This rule is without prejudice to Article 46.’ The issue of consent, being one of jurisdiction, is not subject to Article 42(1) of the ICSID Convention, which determines the law that is applicable to the merits of the dispute. Tribunals have consistently held that Article 42(1) does not govern issues of consent. For example, in CMS v Argentine Republic the tribunal said: ‘Article 42 is mainly designed for the resolution of disputes on the merits and, as such, it is in principle independent from the decision on jurisdiction, governed solely by Article 25 of the Convention and those other provisions of the consent instrument which might be applicable, in the instant case the Treaty provisions.’ CMS Gas Transmission Company v Argentine Republic, ICSID Case No. ARB/01/8, Decision on Jurisdiction, 17 July 2003 (Orrego Vicuña (President), Lalonde, Rezek), <http://ita.law.uvic.ca/chronological_list.htm>, para 88. See also eg Azurix v Argentine Republic, ICSID Case No. ARB/01/12, Decision on Jurisdiction, 8 December 2003 (Rigo Sureda (President), Lauterpacht, Martins), <http://ita.law.uvic.ca/alphabetical_list.htm>, paras 48–50; Enron and Ponderosa Assets v Argentine Republic, ICSID Case No. ARB/01/3, Decision on Jurisdiction, 14 January 2004 (Orrego Vicuña (President), Tschanz, Gros Espiell), <http://ita.law.uvic.ca/chronological_list.htm>, para 38; Siemens v Argentina, ICSID Case No. ARB/02/8, Decision on Jurisdiction, 3 August 2004 (Rigo Sureda (President), Brower, Bello Janeiro), <http://ita.law.uvic.ca/chronological_list.htm>, paras 29–31; Camuzzi International SA v The Argentine Republic, ICSID Case No. ARB/03/2, Decision on Objections to Jurisdiction, 11 May 2005 (Orrego Vicuña (President), Lalonde, Morelli), <http://ita.law.uvic.ca/alphabetical_list.htm>, paras 15–17, 57; AES Corporation v The Argentine Republic, ICSID Case No. ARB/02/17, Decision on Jurisdiction, 26 April 2005 (Dupuy (President), Böckstiegel, Bello Janeiro) <http://ita.law.uvic.ca/alphabetical_list.htm>, paras 34–9; Jan de Nul NV and Dredging International NV v Arab Republic of Egypt, ICSID Case No. ARB/04/13, Decision on Jurisdiction, 16 June 2006 (Kaufmann-Kohler (President), Mayer, Stern), <http://ita.law.uvic.ca/chronological_list.htm., paras 65–8; Schreuer, ‘Consent II’, above n. 8, 40.
(52) In the CSOB case, the tribunal noted that ‘[t]he question of whether the parties have effectively expressed their consent to ICSID jurisdiction is not to be answered by reference to national law. It is governed by international law as set out in Article 25(1) of the ICSID Convention.’ CSOB v Slovak Republic, above n. 17, para 35. This position is correct to the extent that Article 25(1) of the Convention provides a solution to the issue. When consent has been given by statute and the issue is effectual withdrawal, there is no reason not to take national law into consideration, subject always to international law. In the Zhinvali case, the tribunal faced the question whether consent to ICSID jurisdiction had been given by a Georgian statute. The tribunal said that ‘if the national law of Georgia addresses this question of “consent”, which the Tribunal finds that it does, then the Tribunal must follow that national law guidance but always subject to ultimate governance by international law.’ Zhinvali Development Ltd v Republic of Georgia, ICSID Case No. ARB/00/1, Award, 24 January 2003 (Robinson (President), Jacovides, Rubin), 10 ICSID Reports (2006) 3, para 339 (emphasis added). Likewise, in the SPP case, the tribunal held that ‘[w]hile Egypt's interpretation of its own legislation is unquestionably entitled to considerable weight, it cannot control the Tribunal's decision as to its own competence’. Southern Pacific Properties (Middle East) Limited v Arab Republic of Egypt, ICSID Case No. ARB/84/3, Award on Jurisdiction, 14 April 1988 (Jiménez de Arechaga (President); El Mahdi, Pietrowski), 3 ICSID Reports (1995) 131, para 60.
(53) Nolan and Sourgens appear to maintain that consent given by a State through domestic legislation constitutes an ‘independent international obligation’ under the international doctrine of binding unilateral declarations and, for that reason, denunciation of the Convention ‘should not necessarily be viewed as immediately putting an end to the investor's ability to invoke ICSID jurisdiction for an arbitration against the state’. Nolan and Sourgens, above n. 12, 31. For the reasons explained in the text, one could reach conclusions similar to those of Nolan and Sourgens without assuming that every piece of legislation providing for consent to ICSID jurisdiction meets the conditions of a binding unilateral declaration under international law.
(54) Not every clause in a treaty referring to ICSID jurisdiction expresses consent. For example, some treaties contain clauses to the effect that the State may consent or shall sympathetically consider consenting to ICSID jurisdiction after a dispute arises. While every treaty must be interpreted on its own terms, in the absence of special circumstances it is doubtful that such formulations would satisfy the requirement of consent. See Schreuer, Consent, above n. 8, 7. Professor Gaillard correctly notes that some treaties do not contain unqualified consent to ICSID jurisdiction. Gaillard, above n. 12, 51.
(55) Article 26 Vienna Convention, above n. 15, declaratory of customary international law. Nolan and Sourgens oddly maintain that consent given by treaty is an ‘independent international obligation’ under the doctrine of unilateral declarations. Nolan and Sourgens, above n. 12, 31. They explain their position with the notion that the law of unilateral declarations lies at the basis of the law of treaties. Ibid, n. 74. While it is generally recognized that the principle pacta sunt servanda is an expression of the more general principle of good faith, which also underlies the doctrine of unilateral declarations, there is no need to resort to more general theories when nobody disputes that treaties are binding.
(56) Treaty between the Government of the United States of America and the Government of the Republic of Bolivia Concerning Encouragement and Reciprocal Protection of Investment (1998) (henceforth Bolivia-United States BIT).
(57) These articles refer to submission to the courts or administrative tribunals of the party that is a party to the dispute or in accordance with any applicable, previously agreed dispute settlement procedure.
(58) See eg Article 54 Vienna Convention, above n. 15. Article 56 establishes rules for denunciation or withdrawal in cases in which the treaty (i) contains no provisions regarding termination and (ii) does not provide for denunciation or withdrawal. Since the Bolivia-United States treaty does contain rules on termination and withdrawal (in the form of non-renewal), the provisions of Article 56 do not apply.
(59) Article XVI Bolivia-United States BIT, above n. 56. The ten-year survival rule does not apply to the provisions of the treaty dealing with the establishment or acquisition of covered investments. Ibid.
(60) For an analysis of other ICSID arbitration clauses in selected treaties to which Bolivia is a party, see Tietje et al, above n. 5, 28–30.
(61) Manciaux adopts, without analysis, this position. See Manciaux, above n. 12, para 9.
(62) See above n. 30.
(63) This development is well described by Tietje et al, above n. 5, 18 et seq.
(64) In addition to this explanatory use, analogy is frequently used by decision-makers and interpreters as a justification for extending the legal consequences of a rule to a state of affairs that is not fairly included within the conditioning facts referred by the rule but resemble, in certain respects deemed relevant, those conditioning facts. On this use of analogy, see A. Ross, On Law and Justice (1974) 149–50, 153.
(65) This is the extent of my friendly criticism of Professor Schreuer's interpretation of Article 72. The discussion that follows does not apply to his views.
(66) For example, Nolan and Sourgens engage in this type of exercise in classifying the State's individual consent given by treaty, as interpreted by Professor Gaillard, within the category of ‘firm offer’, that is, an offer that by its terms is irrevocable. See Nolan and Sourgens, above n. 12, 22.
(67) A. Ross, ‘Tû-Tû’, 70 Harvard Law Review (1957) 812.
(72) This is not to deny that certain concepts that started as legal concepts may have acquired semantic reference through incorporation in ordinary language and that such semantic baggage may be re-imported into the same or another legal system when the concepts are incorporated in new rules. But even in such cases of semantic feedback, the concept would be, first and foremost, a legal concept in the system that has incorporated it. Whatever semantic baggage the concept may carry must yield, in case of conflict, to the presentation function it performs in the system in which it is incorporated.
(73) I should make clear, once again, that none of this criticism applies to Professor Schreuer's work.
(74) See eg Article 18 Vienna Convention, above n. 15, ‘A State is obliged to refrain from acts which would defeat the object and purpose of a treaty when: (a) it has signed the treaty or has exchanged instruments constituting the treaty subject to ratification, acceptance or approval, until it shall have made its intention clear not to become a party to the treaty; or (b) it has expressed its consent to be bound by the treaty, pending the entry into force of the treaty and provided that such entry into force is not unduly delayed’; Article 27, ‘A party may not invoke the provisions of its internal law as justification for its failure to perform a treaty. This rule is without prejudice to Article 46’; Article 42, ‘1. The validity of a treaty or of the consent of a State to be bound by a treaty may be impeached only through the application of the present Convention. 2. The termination of a treaty, its denunciation or the withdrawal of a party, may take place only as a result of the application of the provisions of the treaty or of the present Convention. The same rule applies to suspension of the operation of a treaty’; Article 70(1), ‘1. Unless the treaty otherwise provides or the parties otherwise agree, the termination of a treaty under its provisions or in accordance with the present Convention: (a) releases the parties from any obligation further to perform the treaty; (b) does not affect any right, obligation or legal situation of the parties created through the execution of the treaty prior to its termination.’