LLM sample dissertation

LLM SAMPLE DISSERTATION

A bilateral investment treaty, as its name suggests, is a treaty concluded between two countries hoping to encourage reciprocal private investment. Such treaties or 'BITs' achieve this end through a fairly standardized list of mutual guarantees, with each nation promising to admit the other's investors into its territory, to accord investors 'fair and equitable' and 'most-favoured nation' treatment, to provide 'constant protection and security' to investments and investors, to pay adequate compensation upon any expropriation, and to resolve disputes with investors arising under the BIT in a given forum or in a given way. One author summarizes the elements, stating that 'virtually all bilateral investment treaties cover four substantive areas: admission, treatment, expropriation and the settlement of disputes'.

A BIT's promise of most-favoured-nation treatment finds expression in a most-favoured-nation or 'MFN' clause. The clause will, roughly speaking, contain a promise by each contracting State that it will treat the other's investments and their investors no less favourably than it treats investments and investors from anywhere else.

Promises regarding the resolution of investor disputes likewise reside in specific clauses, provisions that were considered downright revolutionary when they first emerged on the scene. There was a time when an aggrieved investor could only petition his government to champion his cause against the country in which he had invested (the 'host State'). Redfern and Hunter explain that 'a radical reform' of this unsatisfactory state of affairs 'was made possible by the creation of the ICSID mechanism through the conclusion of the ICSID Convention of 1965 [which] was aimed primarily at creating a new arbitral forum for the resolution of disputes between investors and states . . . .' Today's BITs almost always allow investors to compel the host State to arbitrate disputes arising under the treaty, usually under the auspices of the International Centre for the Settlement of Investment Disputes ('ICSID').

Dispute resolution clauses tend, however, to have limited reach. A clause might cover only expropriation claims, leaving an investor harmed by some violation of the BIT not involving expropriatory conduct in the unenviable position of litigating his or her claims in the allegedly offending country's national courts. A clause might allow for arbitration of the claim at issue, but only after a months-long waiting or 'cooling off' period during which the investor can do little but bear witness to continuing harm.

MFN jurisdiction, to the extent one acknowledges its existence, relies on the combined force of an MFN clause in one BIT (the 'basic treaty') and a dispute resolution clause in another (the 'third-party treaty"). An investor from one country whose basic treaty contains a generous MFN clause but stingy dispute resolution clauses could point to and claim the benefit of kinder dispute resolution clauses in third-party BIT on the ground that not having the benefit of those clauses would subject him to less favourable treatment. If, for example, the basic treaty permits arbitration only of expropriation claims, but a third-party treaty makes room for arbitration of all claims, the investor could compel arbitration of a non-expropriation claim on the theory that he or she must be treated as favourably as investors from the third-party country.

The question when, if ever, a tribunal should accept MFN jurisdiction has received and continues to receive substantial attention in the secondary sources and has even given rise to displays of emotion amongst some in the usually stolid international arbitration community, particularly amongst those who find the idea of MFN jurisdiction objectionable. One particularly well-respected and typically sober commentator went so far as to describe one tribunal's decision to accept MFN jurisdiction as 'heretical.'

The debate, which continues unabated, feeds in largest part on a three-way split in the awards, with five awards establishing a presumption in favor of MFN jurisdiction, three awards erecting a presumption against, and two tribunals eschewing presumptions to focus instead on the terms of the BIT and MFN clause at issue. The split is exemplified in the following passages, the first from Gas Natural (presumption in favor), the second from Plama (presumption against), and the third from RosInvest (no presumption at all):

[A]n MFN provision in a basic treaty does not incorporate by reference dispute settlement provisions in whole or in part set forth in another treaty, unless the MFN provision in the basic treaty leaves no doubt that the Contracting Parties intended to incorporate them.

Unless it appears clearly that the state parties to a BIT or the parties to a particular investment agreement settled on a different method for resolution of disputes that may arise, most-favored-nation provisions in BITs should be understood to be applicable to dispute settlement.

[I]t is the primary function of this tribunal to decide the case before it rather than developing further the general discussion on the applicability of MFN clauses to dispute-settlement provisions. [The provisions of the BIT at issue are ] not identical to [those] in any of such other treaties considered in . . . other [MFN-jurisdiction] decisions. Therefore, they must be interpreted by themselves.

This dissertation examines the sources of and, in the process, tests the validity of the presumptions for and against MFN jurisdiction. It concludes in the end that the presumptions find scant support in logic and law and that any assessment of MFN jurisdiction should confine itself to interpreting the BIT at issue.

Before turning to the presumptions themselves, however, the dissertation first looks at the nature and effect of the MFN standard as expressed in the MFN clause. The point warrants close preliminary consideration because the tribunals establishing presumptions in one direction or the other have consistently either disregarded the standard to cater to a presumption or misunderstood the standard, thereby allowing a presumption to take root.

THE MOST-FAVOURED-NATION STANDARD

The Great Leveller

One might reasonably imagine that a country selected to receive 'most-favoured nation' treatment has earned a place of particular privilege, has been lifted up from amongst the international hoi polloi to become an object of singular honor, goodwill and beneficence. The truth is rather different. Most-favoured-nation treatment does not hold the promise of kind treatment; certain countries have been known to treat their most-favoured nations downright shabbily. Nor does it signal uniquely superior or even separate treatment. Most-favoured-nation treatment is, in a word, nothing of the sort, and favouritism is the very thing a nation receiving such treatment should not expect.

This is so because the most-favoured-nation or 'MFN' principle exists precisely to undermine favoritism, to remove the burden that preferential treatment and discrimination place on international relations generally and on international trade, commerce and investment in particular. It has a 'harmonizing and levelling effect,' serves as 'an instrument for generalizing concessions and for furthering and insuring international commercial equality'. It replaces bias with balance, exchanges partiality for parity.

The MFN principle, as noted, operates through and finds expression in the so-called MFN clause, a provision in a treaty or other compact by which signatory States promise to treat one another and one another's citizens not better than but rather no worse than they treat any other nation or citizenry. Endre Ustor-who was appointed by the International Law Commission ("ILC") in 1968 as its Special Rapporteur on the topic of MFN clauses and whose eight reports to the Commission over the period of some ten years likely remain the most detailed analysis to date of the clause's purpose and effect-explained in his preliminary report, 'In the most simple form of the clause, the conceding State or promiser undertakes an obligation towards another State-the beneficiary-to treat it, its nationals, goods, etc., on a footing not inferior to the treatment it has been giving or will be giving to the most-favoured third State in pursuance of a separate treaty or otherwise.' In more prosaic terms, the clause refers to 'a treaty provision whereby two contracting parties A and B agree that if one of them subsequently concludes with a third State C a trade treaty granting C special trade advantages, those advantages will ipso facto be granted to the earlier contracting party'.

MFN clauses come in every size and stripe, are as varied in their terms as are relations between nations. As Lord McNair put it, 'Although it is customary to speak of the most-favoured-nation clause, there are many forms of the clause, so that any attempt to generalize upon the meaning and effect of such clauses must be made, and accepted, with caution'. Gauging the scope of MFN treatment, then, requires, gauging the scope of the clause, itself an exercise in treaty interpretation.

That interpretive exercise is informed in large part, indeed, primarily by the rule of ejusdem generis, which holds, in simplest terms, that a 'most-favoured-nation clause can only attract matters belonging to the same category of subject as that to which the clause itself relates'. This suggests, and it is commonly understood, that an application of the principle comes in two steps, which might be termed the 'scope step' and the 'sameness step'. In the scope step, one measures the reach of the MFN clause at issue, that is, identifies the 'matters . . . to which the clause itself relates'. In the sameness step, one assesses whether the matters yielded in the scope step 'belong . . . to the same category of subject' as matters in the third-party treaty.

This two-pronged nature of the ejusdem generis principle is made clear in the ILC's Draft Articles on Most-Favoured-Nation Clauses ('ILC Draft Articles'), which devote one article to the issue of scope and another to the issue of sameness. Draft Article 9 states that a beneficiary under an MFN clause 'acquires . . . only those rights which fall within the limits of the subject-matter of the clause.' Draft Article 10 provides that the beneficiary under the clause 'acquires the right to most-favoured-nation treatment only if the granting State extends to a third State treatment within the limits of the subject-matter of the clause.'

A Clause With Claws

The MFN clause's promise to obviate preference-its ability to 'generalize[ ] automatically the advantages granted by one State to any other included in the MFN arrangement' and thereby to 'prevent[ ] discrimination and establish[ ] equality of opportunity on the highest possible plane'-has rendered it a mainstay of international trade, the very 'corner-stone of all modern commercial treaties.' Schwarzenberger writes:

Whatever has been or may be the internal political, social or economic structure of sovereign States, the egalitarian function of the standard corresponds to one of their permanent interests. This explains the historical continuity in the application of the m.f.n. standard.

Lofty talk of egalitarianism and fairness should not, however, be allowed to disguise the clause's darker side. A country granting another MFN treatment makes a real concession, one entailing very real consequences. An MFN clause 'becomes a hindrance to the freedom of action of a nation,' traps a country in a downward spiral of ever greater concessions.

Consider the clause's effect on Freedonia, a smallish, rocky, completely imaginary place, whose primary point of pride (and source of income) is its singularly fine barley biscuits. Having little choice but to import the bulk of its biscuit barley, and yet fiercely protective of its few barley farmers, Freedonia has over the years concluded some forty trade treaties, each of which imposes an import tariff on barley of twenty kroner per bushel. These forty treaties also contain stock MFN clauses providing, in effect, that Freedonia will treat barley imports and barley importers from each signatory country no less favourably than it treats those from any other country.

Freedonia later decides that it would do well to negotiate an import treaty with Ruritania, which, while just as imaginary as Freedonia, is a leading barley producer. For one reason or another-perhaps because Ruritania drives a hard bargain or has market power, perhaps because Freedonia generally seeks closer relations with Ruritania or simply likes the cut of Ruritania's jib, perhaps because Ruritania has itself made concessions on some other point-the resulting Freedonio-Ruritanian treaty places an import tariff of only ten kroner per bushel of barley.

This sets the phones in the Freedonia Ministry of Trade to ringing, with every one of the other forty countries with which Freedonia has a treaty-from Agrabah to Lilliput to Zembla-calling to demand an immediate ten-kroner-per-bushel tariff reduction. Freedonia promised, after all, to treat each of the forty's barley importers no less favourably than it treats any other country's importers, and paying twenty kroner per bushel is by any measure less favourable than paying ten.

There can be no question that the forty countries would be acting within their rights. An MFN clause brings parity; 'if any favors, rights, privileges, also particularly tariff-reductions are granted by either of the contracting parties to any third party . . . such grants inure to the benefit of the other party simultaneously and without any compensation'. Having undertaken to provide most-favoured treatment, Freedonia's deal with Ruritania effectively spawned forty new deals, deals that when taken together reduce Freedonia's barley-related tariff revenue by roughly half.

Freedonia, of course, might have anticipated and to some extent even welcomed this result. It might have intended to effect a blanket tariff adjustment, perhaps believing that its barley farmers no longer needed protection or that its 20-kroner-per-bushel tariff policy was hurting its international reputation or its position in other trading areas. Its concession to Ruritania would allow it to make that adjustment in short order and with minimal disruption. Scott Vesel, specifically addressing MFN clauses in bilateral investment treaties, writes:

[T]he MFN clause, if construed broadly, reduces transaction costs for states wishing to adopt more investor-friendly policies. Rather than renegotiating a large number of BITs to incorporate the change, a host state can simply agree to a single BIT with the more favorable provision with the knowledge that by operation of the MFN clauses, the new treatment will apply to investors from any country with a BIT.

Yet the fact that Freedonia might not have wanted to effect a sweeping change in tariff policy, might have failed to foresee the demands for tariff reductions or otherwise appreciate the cascade effect of its deal with Ruritiania is immaterial; the effect of the MFN clause is immediate and automatic. Nor does it matter that the twenty-kroner tariff in Freedonia's earlier treaties might have been in each instance the product of specific and close negotiation, that the forty countries refuse to answer the reduction in the Freedonian tariff by reducing their own tariffs on Freedonian goods or otherwise compensating Freedonia, or that the Freedonio-Ruritanian Treaty might have specifically provided that the 10-kroner tariff would not apply to other countries.

An MFN clause, then, can work to startling and at times harsh effect, obviating in a single stroke specific and hard-fought concessions. It has been likened to a ratchet that works in only one direction and multiplies the force of every compromise, described as a haven for 'free riders' and a tool for 'putting the services of the shrewdest negotiator of a third country gratuitously at the disposal of one's own country'.

Still, this powerful ratcheting effect is, in the end, the clause's very purpose, its raison d'etre, and nations include MFN clauses precisely to enjoy that effect. Potency inheres in the clause. That the clause promises in a given instance to produce a potent result should come as no surprise.

By all appearances, however, the power of the MFN clause to displace treaty provisions wholesale has, in fact, surprised and even confounded the lion's share of the tribunals confronting a claim of MFN jurisdiction and has played no mean role in persuading those tribunals either to erect a presumption against such jurisdiction or to limit the circumstances in which jurisdiction will lie. That is, while the various tribunals cater to a number of predilections and advance several lines of argumentation in formulating their presumptions for and against MFN jurisdiction, misapprehension of and apprehension about the MFN standard serve as a unifying theme and frustrate what should be a straightforward exercise in treaty interpretation.

TESTING THE PRESUMPTIONS FOR AND AGAINST MFN JURISDICTION

The various assumptions and understandings informing the presumptions for and against MFN jurisdiction tend to run together, with tribunals (for ease of reference, and for lack of a better term, 'presuming tribunals') as often as not seeking support for one predilection in another. One can, however, tease out of the awards five points from which the conflicting presumptions arise.

  1. Disregard of the Rules of Treaty Interpretation - This first point differs from the remaining four in that it involves not a dispute between the two sides over some issue of law or principle but rather an approach taken by each of the presuming tribunals, purposefully or not, which leaves room for their respective presumptions to grow. The tribunals, without exception, state their intent to abide by the rules of treaty interpretation laid out in Articles 31 and 32 of the Vienna Convention on the Law of Treaties ('Vienna Convention'). However, again without exception, although to varying degrees, they apply those rules only insofar as they cater to a predetermined outcome.

  2. The Debate Over the Relevance and Effect of Decisional Authority - The first point of contention centers on the import of a 1956 arbitral award and three 1950s ICJ decisions to the question of MFN jurisdiction. The presuming tribunals' take on the meaning and effect of that authority could hardly vary more. It is for this tribunal inapposite, for that, illustrative. One tribunals finds in it ample support for accepting MFN jurisdiction. Another sees it as undermining the very notion that such jurisdiction could exist.

  3. The Debate Over the Importance of the Importance of Investor-State Arbitration - A second area of dispute revolves around the extent, if any, to which the interpretive exercise should be informed by the uniformly accepted fact that investor-State arbitration plays an important role in protecting foreign investors and fostering foreign direct investment. Certain awards accepting jurisdiction consider the point highly relevant, going on at length about 'abusive practices of the past' and describing resort to arbitration as 'essential . . . to the protection of the rights envisaged under the pertinent treaties'. Other awards flatly reject the importance of arbitration as an appropriate consideration.

  4. The Debate Over Evidence of Consent - The third point of disagreement starts from a point of agreement, specifically, that arbitral jurisdiction exists only by the consent of the parties in arbitration. Tribunals presuming against MFN jurisdiction maintain that the evidence of party consent must be 'clear and unambiguous' and urge that an MFN clause which does not explicitly and expressly state its intent to reach the basic treaty's dispute-resolution provisions does not, regardless of its breadth, cut the evidentiary mustard.

  5. The Debate Over the Extent to Which MFN Jurisdiction Impairs Upsets the Balance of the Contracting States' Bargain - The final and in many ways central point most directly illustrates the tribunals' fear and misunderstanding of the MFN standard. Tribunals presuming against MFN jurisdiction see such jurisdiction as depriving the host State of the 'benefit of the Bargain', of robbing that State of the dispute-resolution mechanisms for which it specifically negotiated and to which it agreed. Those presuming in favor of MFN jurisdiction share those same concerns but believe that jurisdiction may nonetheless be had if the disruption to the contracting States' bargain is minimal. Only those tribunals rejecting presumptions-Siemens and RosInvest-accept the MFN standard for what it is, accepting that it operates by its nature to displace provisions in a basic treaty.

    The Tribunal's Resistance to the Vienna Convention

    The Vienna Convention, the work of the ILC, is essentially the treaty of treaties. It deals with, among a great many other things, the capacity of states to conclude treaties, the authentication of a treaty's text and the means by which a State may express its consent to be bound by a treaty. It also devotes two Articles, Articles 31 and 32, to rules governing treaty interpretation. Some view these rules as having incorporated already well-settled rules of treaty interpretation, while others contend there never was a "settled rule and consequently any codification would inevitably contain a strong element of "progressive development"'. In all events, the Vienna Convention is a treaty-the law between nations signatory to it-and its rules of interpretation form a part of that law.

    Article 31 sets out a 'general rule of interpretation', stating that '[a] treaty shall 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 the light of its object and purpose.' The general rule explains that 'context' refers to the text of the treaty (including its preamble and annexes) as well as to agreements made by the parties and any instruments prepared by one party and accepted by the other in connection with the conclusion of the treaty. It then provides that the interpreter must consider, in addition to context, subsequent agreements concerning the application or interpretation of the treaty, subsequent applications of the treaty that shed light on the treaty's meaning, and relevant rules of international law. Finally, it states that 'a special meaning shall be given to a term if it is established that the parties so intended'.

    This general rule should, in the ordinary course, be the only rule. That is, the interpreter is to conclude the interpretive task by determining a meaning of the terms at issue on the basis of the factors listed in Article 31, i.e., ordinary meaning, object and purpose, etc. Only then may the interpreter turn to Article 32, which allows for resort 'to supplementary means of interpretation', for example, travaux preparatoires, 'to confirm the meaning resulting from the application of article 31' or to determine meaning if the Article-31 analysis 'leaves the meaning ambiguous or obscure or leads to a result which is manifestly absurd or unreasonable'.

    This approach, which greatly restricts the materials to be considered in the interpretive process and allows resort to additional indicia only in the exceptional case, has been described as a 'triumph of textualism over intentionalism', as maintaining a 'steadfast focus on the words of the treaty with extraneous evidence of intended meaning relegated to a secondary role.' One author explains:

    The division between Articles 31 and 32 is . . . based on the assumption that the text must be presumed to be the authentic expression of the intentions of the parties; and that, in consequence, the starting point of interpretation is the elucidation of the meaning of the text, not an investigation ab initio into the intentions of the parties.

    This textualist approach comports nicely with a point made earlier, that because MFN clauses vary so widely in their terms, each clause must be considered independently, on its own terms and without recourse to preconceived notions about its meaning. Ustor explained in his commentary to the Draft Articles:

    When attempting to define the notion of most-favoured-nation clause the warning of McNair was kept in mind: 'although it is customary to speak of the most-favoured-nation clause, there are many forms of the clause, so that any attempt to generalize upon the meaning and effect of such clauses must be made, and accepted, with caution.' Expressed in other words: 'Speaking strictly, there is not such thing as the most-favoured-nation clause: every treaty requires independent examination

    Each of the presuming tribunals departs from the Vienna Convention's textualist approach to one degree or another. Plama, it must be said, serves as a conspicuous example. The award establishes a strong presumption against the existence of MFN jurisdiction, holding that because evidence of a State's consent to arbitrate a dispute must be 'clear and unambiguous', no general MFN clause-no matter how long its reach-should be interpreted as embracing dispute resolution. It frankly acknowledges its view that interpretation is beside the point, that its presumption withstands the very meaning of the MFN clause itself. The award states at one point, for example, that 'it is not clear whether the ordinary meaning of the term "treatment" [in the MFN clause] includes or excludes dispute settlement provisions in [third-party] BIT's . . . . But as will be seen below, it is not relevant to address that question.'

    While an outlier, Plama is not alone in disregarding the Vienna Convention to cater to a presumption or predilection. Gas Natural focuses not on the term's of the MFN clause at issue but rather on the perceived importance of investor-State arbitration to foreign-direct investment, an overly teleological approach which has a BIT's 'object and purpose' swallowing its words whole. Berschader effectively abandons interpretation in favor of the unsupported proposition that '[t]here is a fundamental difference as to how an MFN clause is generally understood to operate in relation to the material benefits afforded by a BIT, on the one hand, and in relation to dispute resolution clauses, on the other hand.'

    These tribunals, which placed penchant before the plain terms of the BIT at issue, did more than merely frustrate the interpretive process. The Vienna Convention-at least where, as in all the MFN-jurisdiction cases, the State parties to the BIT at issue are also Vienna Convention signatories-is a binding treaty like any other and must be applied according to its terms. A tribunal paying only lip service to those terms effectively abets a violation of that treaty.

    The Debate Over the Meaning of Earlier Authority

    Six tribunals assessing MFN jurisdiction, four of them presuming tribunals, accepted guidance from or at least considered the relevance of three decisions of the ICJ decisions and an arbitral award related to one of those decisions. The three ICJ decisions issued in the Ambatielos Case (Merits: Obligation to Arbitrate) (Ambatielos Case) the Case Concerning Rights of Nationals of the United States of America in Morocco (Nationals in Morocco), and the Anglo-Iranian Oil Company Case (Anglo-Iranian). The arbitral award, which followed on the heels of the Ambatielos Case was made in The Ambatielos Claim (Ambatielos Award).

    One genuinely wonders at times whether the tribunals considering the importance of this authority to the question of MFN jurisdiction had, in fact, read the same cases. They are on the one hand, only 'marginally relevant' and yet on the other, address many of the issues relevant to the point. The Nationals in Morocco decision, depending on whom one asks, either 'do[es] not address the issue' of MFN jurisdiction or 'accepted that MFN clauses may extend to provisions related to jurisdictional matters'. As for the Ambatielos Award, it came nowhere near 'formulating a general principle concerning the scope of an MFN clause' and yet single-handedly establishes the general principles to be followed in gauging the scope of an MFN clause. The decisions, in short, provide lusty support for accepting MFN jurisdiction, while utterly undermining the very notion that such a thing could exist.

    The truth lies somewhere in the middle, although, in fairness, it favours the tribunals that have accepted MFN jurisdiction. The ICJ decisions and the Ambatielos Award, while not creating a presumption, do provide important support for the concept of MFN jurisdiction.

    The Ambatielos Case and the Ambatielos Award

    The Ambatielos decision and award centered on a claim by Mr. Ambatielos, a Greek citizen, that the British Government had ruined him by, among other things, leveling unfounded claims for back income tax, breaching a contract to deliver nine ships by a certain date and refusing to reimburse £500,000 paid to secure that delivery date, and working a 'denial of justice' by preventing him from introducing evidence and otherwise pressing his claims against the Government before the British Board of Trade and the British courts. Greece, acting on Ambatielos' behalf, filed a claim with the ICJ seeking to compel arbitration under an arbitration clause in the 1886 Greece-U.K. Treaty of Commerce and Navigation ('1886 Treaty') and a 1926 Declaration extending the arbitration clause to private claims based on the 1886 treaty.

    The U.K. argued that Ambatielos' claims fell outside any protections afforded by the 1886 Treaty and, by extension, the arbitration clause. Greece responded that arbitral jurisdiction existed because Ambatielos had effectively alleged a breach of a provision in the treaty promising that the 'subjects of each of the two Contracting Parties in the dominions and possessions of the other shall have free access to the Courts of Justice for the prosecution and defence of their rights'. Perhaps understanding that Ambatielos-who, after all, had pursued his claim through to the Court of Appeal in London before seeking to arbitrate-had very likely enjoyed 'free access to the Courts', Greece also argued that Ambatielos' claim alleged a breach of the treaty's MFN clause. It observed in this regard that the clause explicitly stated its intent to place 'the trade and navigation of each country . . . in all respects . . . on the footing of the most-favoured-nation' and provided MFN treatment in 'all matters relating to commerce and navigation' and in respect of 'any privilege, favour or immunity whatever' granted to a third State. It then pointed to a provision in a Denmark-U.K. treaty by which the States promised to 'cause justice and right to be speedily administered to [one another's] subjects . . . without tedious and unnecessary delays and charges.'

    After making clear that it could not consider the merits of the dispute and that it would compel arbitration if it found Greece's arguments regarding a breach of the treaty 'sufficiently plausible', the Court ordered the claim to arbitration. It stated, 'Having regard to the contentions . . . with respect to the scope and effect of the most-favoured-nation clause . . . as well as the divergence of views concerning the meaning of the expression "free access to the Courts of Justice" . . . the Court must conclude that this is a case in which [Greece] is presenting a claim "based on the provisions of the [treaty]"'.

    Four judges dissented, relying on the ejusdem generis rule for the proposition that the matters embraced by the MFN clause, 'all matters relating to commerce and navigation', were not of the same kind as provisions in third-party treaties concerning the 'administration of justice'. They noted in support of their determination , and on a point that will prove of interest later, that the basic treaty specifically addressed the administration of justice in its 'free access' clause and this clause 'contain[ed] no reference to most-favoured-nation treatment'.

    The arbitral tribunal addressing the merits of Ambatielos' claim disagreed with the ICJ dissenters. It acknowledged the vitality of the ejusdem generis rule, holding that an MFN clause 'can only attract matters [in third-party treaties] belonging to the same category of subject as that to which the clause itself relates'. However, unlike the dissenters, it found the rule met:

    It is true that the 'administration of justice,' when viewed in isolation, is a subject-matter other than 'commerce and navigation,' but this is not necessarily so when it is viewed in connection with the protection of the rights of traders. Protection of the rights of traders naturally finds a place among the matters dealt with by Treaties of commerce and navigation.

    Therefore it cannot be said that the administration of justice, in so far as it is concerned with the protection of these rights, must necessarily be excluded from the field of application of the most-favoured-nation clause, when the latter includes 'all matters relating to commerce and navigation.'

    The tribunal provided further support for its determination by noting that the basic treaty promised to place the 'trade and navigation . . . in all respects . . . on the footing of the most-favoured-nation.'

    Salini and Plama reject the Ambatielos Award as significant to the question of MFN jurisdiction, noting that Greece relied on the MFN clause in the Greece-U.K. treaty not to invoke 'procedural' dispute-resolution provisions in third-party treaties but rather to invoke those treaties 'substantive' promises of 'treatment in accordance with justice, right, equity and the principles of international law'. Putting aside for a moment any putative distinction between substance and procedure in this context, Salini and Plama correctly point out that Ambatielos was not an MFN-jurisdiction case.

    The Ambatielos Award nonetheless provides significant conceptual support for MFN jurisdiction. Firstly, and as observed by a number of tribunals and commentators, the award establishes that a broad and non-specific MFN clause may enjoy a broad and non-specific sweep. Matters in third-party treaties that at first blush seem quite remote from matters embraced by such a clause might, on closer inspection, be sufficiently similar to satisfy the ejusdem generis rule.

    Second, and perhaps more important to the debate over MFN jurisdiction, the Ambatielos Award stands for the proposition that a general MFN clause is more likely to embrace a given subject matter if the basic treaty addresses that subject matter. Like the ICJ dissenters, the Ambatielos arbitration tribunal took note of the fact that the 1886 Treaty particularly addressed the issue of administration of justice in its 'free access' clause. Unlike the dissenters, however, it found this fact as supporting rather than undermining the extension of the 1886 Treaty's MFN clause to matters concerning the administration of justice. It reasoned that if commerce-and-navigation treaties generally contain provisions relating to the administration of justice, and more particularly if the basic treaty itself contains such an administration-of-justice provision, it is all the more appropriate to interpret an MFN clause including 'all matters relating to commerce and navigation' as embracing matters concerning the 'administration of justice', notwithstanding that '"administration of justice", when viewed in isolation, is a subject matter other than "commerce and navigation"'.

    This boasts an almost undeniable logic. If a treaty of type y specifically addresses subject matter x, and if type y treaties almost always specifically address subject matter x, one would fairly assume that a reasonably broad and indiscriminate MFN clause in a type y treaty-which has as its goal to eliminate discrimination in the treatment of type-y-treaty signatory States-should be understood to embrace subject matter x, even if the MFN clause does not specifically identify x as falling within its purview. The logic gains force when one recalls that MFN clauses often take a non-specific and broad form precisely because one party cannot possibly divine the beneficence of the other to third-party States in futuro.

    With the exception of Siemens, none of the tribunals assessing MFN jurisdiction has appreciated that this point is the point to be taken from the Ambatielos Award. Tribunals declining MFN jurisdiction have, like the ICJ dissenters, understood that MFN jurisdiction should not lie where the basic treaty specifically addresses the same subject matter as the third-party treaty on which the MFN beneficiary seeks to rely, that accepting jurisdiction in such circumstances would upset the 'benefit of the bargain'. Telenor states:

    When concluding a [BIT] with specific dispute resolution provisions, states cannot be expected to leave those provisions to future (partial) replacement by different dispute resolution provisions through the operation of an MFN provision, unless the States have explicitly agreed thereto. . . .

    This, of course, cannot be right, something made clear by the earlier examination of the MFN standard. Extending Plama's reasoning to the Fredonia hypothetical has Fredonia properly refusing to reduce its tariffs precisely because its earlier treaties set tariffs. The Ambatielos Award takes the point one step further, explaining, again with resort to the hypothetical, that the MFN clauses on which Fredonia's trading partners rely are more likely to be understood as embracing tariff provisions because they reside in basic treaties containing tariff provisions.

    Case Concerning Rights of Nationals of the United States of America in Morocco

    Nationals in Morocco, a case brought by France against the United States, centered on the Morocco-U.S. Treaty of 1836, a capitulatory treaty giving the U.S. Consul jurisdiction over disputes between U.S. citizens living in Morocco. The treaty contained two general MFN clauses, one promising to place 'commerce with the United States . . . on the same footing as . . . that with the most favored nation', the other entitling the U.S. to 'whatever indulgence, in trade or otherwise [that might] be granted to any of the Christian Powers'.

    The U.S. argued, and the court and France accepted, that these clauses allowed the U.S. to claim the benefit of a provision in the Morocco-U.K. Treaty of 1856 which allowed the U.K. to exercise consular jurisdiction not only over disputes between U.K. citizens but also over any dispute in which a U.K. citizen was a defendant. 'Accordingly', held the court, 'the United States acquired by virtue of the most-favoured-nation clauses, civil and criminal consular jurisdiction in all cases in which United States nationals were defendants.'

    Nationals in Morocco goes beyond the Ambatielos Award in establishing theoretical underpinnings for MFN jurisdiction. Like Ambatielos, it involved a basic treaty with a broad MFN clause together with provisions addressing the same subject matter as addressed by provisions in the third-party treaty. This, as the Ambatielos Award establishes, at least suggests that the MFN clause in the basic treaty meant to bring consular jurisdiction within its scope.

    Nationals in Morocco goes further in that it involved treaty provisions a great deal more specific than those at issue in Ambatielos. The administration-of-justice provisions discussed in Ambatielos bordered on the hortatory, with the basic treaty guaranteeing 'free access to the Courts of justice' and the third-party treaties promising 'treatment in accordance with justice, right, equity and the principles of international law'. The provisions at issue in Nationals in Morocco, by contrast, were quite exact, precisely laying out the scope of and exceptions to the U.S.' consular jurisdiction. This proves important to the discussion because it calls into question the determination by several tribunals addressing MFN jurisdiction that a broad MFN clause should not be allowed to upset 'specifically negotiated' dispute-resolution provisions in the basic treaty.

    Second, while not an MFN jurisdiction case, Nationals in Morocco does establish that a generally-worded MFN clause may serve to import jurisdictional or 'procedural' provisions from a third-party treaty. This raises serious doubts about statements in Telenor, and echoed in other awards, that a general MFN clause promising MFN treatment to investments should be understood to mean that an 'investor's substantive rights in respect of investments are to be treated no less favourably . . . and there is no warrant for construing the above phrase as importing procedural rights as well.'

    Nationals in Morocco distinguishes itself from the MFN-jurisdiction cases primarily in that it involved not an attempt to invoke jurisdiction available under a third-party treaty but rather an attempt by a State to broaden its own jurisdiction under a third-party treaty. Yet this distinction is more apparent than real. As the full name of Morocco suggests-Case Concerning Rights of Nationals of the United States in Morocco-the U.S. litigated the case on behalf of U.S. citizens living in Morocco, who were seeking to import kinder jurisdictional provisions from another treaty because they quite naturally preferred to defend claims brought by Moroccan nationals before the U.S. Consul rather than in a Moroccan court. Read in this light, Nationals in Morocco provides almost direct precedential support for MFN jurisdiction.

    Anglo-Iranian Oil Company Case

    Anglo-Iranian is the least relevant of the cases to MFN jurisdiction, providing support neither for nor against the notion of MFN jurisdiction. It warrants consideration, however, if only because it has featured so prominently in the MFN-jurisdiction awards.

    The case arose out of the now famous dispute between Iran and the Anglo-Iranian Oil Company regarding the latter's right to compensation upon Iran's nationalization of its oil fields in 1951. Iran challenged the jurisdiction of the ICJ over the matter, arguing that while it had accepted the ICJ's jurisdiction by a declaration ratified in 1932 (the 'Declaration'), the Declaration specifically 'limit[ed] the jurisdiction of the Court to disputes arising after . . . ratification.' The United Kingdom, litigating on the oil company's behalf, maintained, among a great many other things, that the court had jurisdiction by virtue of the combined force of MFN clauses in Iran-U.K. treaties of 1858 and 1903 and a 1934 (i.e., post-Declaration) treaty between Iran and Denmark. The ICJ declined jurisdiction, determining that the United Kingdom could not invoke a post-Declaration treaty through an MFN clause in a pre-Declaration treaty.

    Central to the court's holding was the principle of res inter alios acta, which expresses the rather unremarkable proposition that every third-party treaty is 'an act between others' and that persons and States neither party to nor a beneficiary of that treaty therefore lack standing to invoke its protections. The MFN standard was at one time viewed by some as an exception to the res inter alios acta rule, the idea being that an MFN beneficiary derived his right to superior treatment afforded by a third-party treaty from the third-party treaty itself. Anglo-Iranian put paid to that notion, holding that an MFN beneficiary looks not to the third-party treaty but rather to the basic treaty as the source of his rights. Because the basic treaties at issue-the 1858 and 1903 Iran-U.K. treaties-pre-dated the Declaration, their MFN clauses could not serve to reach a post-Declaration treaty:

    Without considering the meaning and scope of the most-favoured-nation clause, the Court confines itself to stating that this clause is contained in the [Iran-U.K.] Treaties, which are not subsequent to the ratification of the Declaration. While Iran is bound by her obligations under these treaties as long as they are in force, the United Kingdom is not entitled to rely upon them for the purpose of establishing the jurisdiction of the Court, since they are excluded by the terms of the Declaration.

    Plama discounted Anglo-Iranian on the basis of this passage's first clause, determining the case has no relevance because the ICJ declined to consider the scope of the basic treaties' MFN clauses. This largely misses the point. The court did not consider the scope of the MFN clause, and Anglo-Iranian is inapposite to the issues presented in the MFN-jurisdiction cases, because the clause's scope had no bearing on the question of its jurisdiction. The United Kingdom did not invoke the MFN clause in its treaties with Iran to enjoy a dispute-resolution right or indeed any right conferred by the Iran-Denmark treaty-which would have implicated the scope of the clause-but rather to enjoy a right conferred by Iran's 1932 Declaration. It was this which prevented the United Kingdom's recourse to the Iran-Denmark treaty, which rendered that treaty res inter alios acta.

    When taken together, the ICJ decisions and the Ambatielos Award provide important conceptual support for MFN jurisdiction. They establish, in sum, (1) that a broad and general MFN clause enjoys a broad and general scope, (2) that an MFN clause should be more readily understand as embracing a matter where other clauses in the basic treaty speak to that matter and where treaties of the same type generally address that matter, (3) that an MFN clause may operate to replace a precise and specifically-negotiated provision in a basic treaty with an equally precise provision in the third-party treaty, and (4) that an MFN clause may extend to jurisdictional matters.

    The Debate Over the Importance of the Importance of Arbitration to Investors and Investment

    Those on all sides of the debate over MFN jurisdiction agree, as they essentially must, that the ability of an investor to resolve a dispute with a host nation before a neutral arbitration tribunal rather than in the host State's courts counts amongst an investor's signal protections and, in turn, fosters foreign direct investment. Gas Natural, which arguably adopts the idea of MFN jurisdiction most wholeheartedly, states:

    As the Tribunal sees the history, first of the ICSID Convention, which created the institution of investor-state arbitration, and subsequently of the wave of bilateral investment treaties between developed and developing countries . . . a crucial element - indeed perhaps the most crucial element - has been the provision for independent international arbitration of disputes between investors and host states.

    Plama, perhaps the most vehement of MFN jurisdiction's detractors, reluctantly evinces a similar understanding, approving the proposition from Maffezini that 'dispute settlement arrangements are inextricably related to the protection of foreign investors' and accepting that the founding of ICSID represented 'a major step toward promoting an atmosphere of mutual confidence and thus stimulating a larger flow of private international capital in those countries which wish to attract it'.

    The disagreement centers on the role, if any, the conceded importance of arbitration to investors and investment should play in determining whether MFN jurisdiction lies. Maffezini sees it as taking lead billing; indeed it forms the crux of the tribunal's decision to accept jurisdiction. It states in the paragraphs leading up to its holding that 'today dispute settlement arrangements are inextricably related to the protection of foreign investors' and that '[i]nternational arbitration and other dispute settlement arrangements . . . are essential . . . to the protection of the rights envisaged under the pertinent treaties'.

    Plama disagrees, stating that protestations regarding the protective role of investor-State arbitration are 'undeniable in their generality' but are 'legally insufficient to conclude' that a basic treaty's MFN clause was intended by the contracting States to reach the treaty's dispute resolution provisions. Berschader goes somewhat further still, stating that arguments highlighting the importance of investor-State arbitration 'offer little or no guidance as to whether, in a specific case, the contracting parties to a treaty, which already provides for arbitration in certain types of disputes, actually intended the arbitration clause to be extended in the future to other kinds of disputes'.

    Plama and Berschader view Maffezini and its progeny as putting far too much weight on the combined force of, first, recitals in the basic treaties under consideration to the effect the treaties' purpose is to foster and encourage investment and, second, that portion of Vienna Convention Article 31(1) calling for a construction of a treaty 'in the light of its object and purpose.' In discounting investor-State arbitration's importance to the MFN-jurisdiction analysis, Plama states:

    Here, the Tribunal is mindful of Sir Ian Sinclair's warning of the "risk that the placing of undue emphasis on the 'object and purpose' of a treaty will encourage teleological methods of interpretation [which], in some of its more extreme forms, will even deny the relevance of the intentions of the parties<".

    Plama et al. have a point; one cannot have the tail wagging the dog. By the same token, they go too far, essentially denying the existence of the tail. They seem to overreact to statements in the cases accepting MFN jurisdiction concerning the importance of investor-State arbitration, ultimately giving the point no weight or even, appearances suggest, less than no weight in the analysis.

    This is unfortunate because the backlash amongst some tribunals against the importance-of-arbitration argument first espoused in Maffezini-and, indeed, the wholehearted endorsement of that argument by other tribunals-result from a misunderstanding of the Maffezini award. The BIT at issue in Maffezini contained a complex and rather unusual dispute-resolution regime, which provided, in relevant part, that an investor was required to litigate any dispute in the local courts for a period of eighteen months, at which point the investor could initiate arbitration, regardless of the progress of the judicial proceedings and, indeed, even if the proceedings had concluded with a ruling on the merits adverse to the investor.

    Tribunals declining MFN jurisdiction have understood Maffezini as little more than an attempt to rescue the investor-claimant from this 'non-sense' dispute-resolution provision. They read the decision as stretching the limits of logic and the law-as expounding hortatory, pie-in-the-sky principles about the importance of arbitration to investors-to find the Spain-Argentina MFN clause embraced the BIT's dispute resolution provisions and, in turn, to rescue the hapless claimant from eighteen months of expensive litigation. The Plama tribunal, essentially describing Maffezini as a case of 'bad facts making bad law', writes:

    The decision in Maffezini is perhaps understandable. The case concerned a curious requirement that during the first 18 months the dispute be tried in the local courts. The present Tribunal sympathizes with a tribunal that attempts to neutralize such a provision that is nonsensical from a practical point of view. However, such exceptional circumstances should not be treated as a statement of general principle guiding future tribunals in other cases where exceptional circumstances are not present.

    Yet Maffezini did not focus on the importance of arbitration in reaction to an odd dispute-resolution clause but rather in reaction to an odd MFN clause. That clause, what one commentator would describe as a 'conjoined' clause, links its promise of MFN treatment to another form of investor protection, specifically, the promise of fair-and-equitable treatment:

    1. Each Party shall guarantee in its territory fair and equitable treatment of investments made by investors of the other Party.

    2. In all matters governed by this Agreement, such treatment shall be no less favorable than that accorded by each Party to investment made in its territory by investors of a third country.

    One can reasonably understand this clause as expressly extending MFN treatment only to fair and equitable treatment and, by extension, not to any other protection accorded by the BIT, including the BIT's dispute-resolution provisions. Under this reading, statements regarding the role of investor-State arbitration in protecting investors are perfectly appropriate; they inform the sameness step in the ejusdem generis analysis, that is, the determination whether dispute-resolution mechanisms form a necessary part of, are ejusdem generis with, fair and equitable treatment. However, one can just as reasonably view the clause as an exceptionally broad one which embraces 'all matters governed' by the BIT, dispute resolution amongst them. Under this reading, sameness is obvious; dispute-resolution provisions in one treaty are of the same order as dispute-resolution provisions in another.

    Unfortunately, the Maffezini tribunal reads it both ways. It appears occasionally to accept that the basic treaty's MFN clause embraces only that treaty's promise of 'fair and equitable treatment' and that its job is therefore to determine whether fair-and-equitable treatment and the treatment afforded investors by the dispute-resolution provisions in the Spain-Chile BIT are sufficiently similar to satisfy the ejusdem generis rule. It writes:

    The second major issue concerns the question whether the provisions on dispute settlement contained in a third-party treaty can be considered to be reasonably related to the fair and equitable treatment to which the most favored nation clause applies under the basic treaties on commerce, navigation or investments and, hence, whether they can be regarded as a subject matter covered by the clause. This is the issue directly related to the ejusdem generis rule.

    At still other times, the tribunal suggests that MFN treatment is available to the claimant only because the MFN clause is best understood as embracing not merely the basic treaty's guarantee of fair-and-equitable treatment but also the treaty's dispute resolution provisions. It refers, for example, to MFN clauses in certain United Kingdom BITs that expressly encompass the treaties' dispute-resolution provisions, considers Spain's and Argentina's positions on dispute resolution during the negotiations preceding conclusion of the Spain-Argentina BIT, examines Spain's stance on dispute resolution in its historical treaty practice, and compares the breadth of the Spain-Argentina MFN clause with that of other clauses. It also concludes-in a formulation that, parenthetically, defies the rules of interpretation laid out in the Vienna Convention and that, despite the tribunal's statements to the contrary, enjoys no support from the Ambatielos Award:

    [Where a basic treaty does] not provide expressly that dispute settlement as such is covered by the [MFN] clause . . . it must be established whether the omission was intended by the parties or can reasonably be inferred from the practice followed by the parties in their treatment of foreign investors and their own investors.

    The reader walks away from Maffezini genuinely puzzled over which of the two approaches the tribunal has adopted, whether it found the ejusdem generis principle satisfied, on the one hand, because, while the MFN clause embraced only 'fair and equitable treatment', such treatment and the right to resort directly to arbitration in the Spain-Chile BIT are ejusdem generis or, on the other, because the MFN clause extended to the basic treaty's dispute resolution provisions, which would, of course, be ejusdem generis with the Spain-Chile arbitration provisions. Traces of both approaches, but particularly the former, can be found in those paragraphs of the award roughly constituting the holding:

    Notwithstanding the fact that the basic treaty containing the clause does not refer expressly to dispute settlement as covered by the most favored nation clause, the Tribunal considers that there are good reasons to conclude that today dispute settlement arrangements are inextricably related to the protection of foreign investors . . . .

    International arbitration and other dispute settlement arrangements have replaced . . . older and frequently abusive practices of the past. These modern developments are essential . . . to the protection of the rights envisaged under the pertinent treaties.

    From the above considerations it can be concluded that if a third-party treaty contains provisions for the settlement of disputes that are more favorable to the protection of the investor's rights and interests than those in the basic treaty, such provisions may be extended to the beneficiary of the most favored nation clause as they are fully compatible with the ejusdem generis principle.

    The mixed nature of Maffezini's analysis has been entirely lost on later tribunals and on commentators. Each has understood Maffezini as having determined that the MFN clause directly extended to the basic treaty's dispute-resolution provisions. They have understood as well, and by logical extension, that Maffezini relied heavily on the importance-of-arbitration argument in making that finding.

    Cases following Maffezini have, as a result, found the importance of arbitration to investment as essentially sufficient in itself to find that an MFN clause extends to a basic treaty's dispute resolution provisions. Gas Natural concludes, for example, that the 'critical issue [in gauging the existence of MFN jurisdiction] is whether or not the dispute settlement provisions of bilateral investment treaties constitute part of the bundle of protections granted to foreign investors by host states'.

    Plama and its kin-quite rightly in this author's opinion-reject that notion, concluding that the scope of an MFN clause cannot be gauged solely by reference to a BIT's stated purpose to encourage reciprocal investment. However, they allow their disapproval of what they perceive as an overly teleological approach to get the better of them. The pendulum swings too far in the other direction, with 'object and purpose' all but sacrificed, it seems, merely to make a point.

    The preferred approach, of course, would be to follow the rules of interpretation laid out in the Vienna Convention (i.e., the very rules which each tribunal purports to follow), which require 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 the light of its object and purpose.' This approach, which places words first and relegates 'object and purpose' to a backdrop against which those words should be considered, keeps teleology in its proper place.

    The Debate Over Evidence of Consent

    The proposition that arbitral jurisdiction exists only by the consent of the parties is black letter, fundamental, the basis of arbitration qua arbitration. As Redfern and Hunter put it, 'This element of consent is essential. Without there can be no valid arbitration.'

    Plama largely derives its presumption against MFN jurisdiction from this indisputable point. It begins by according special treatment to arbitration agreements, stating they create arbitral jurisdiction only if their intent in that regard is 'clear and unambiguous'. It then raises the bar still higher, determining that because 'doubts [regarding consent] can arise if the agreement to arbitrate is to be reached by incorporation by reference', MFN jurisdiction can exist only where an MFN clause 'leaves no doubt' of its intent to embrace dispute-resolution mechanisms. The tribunal never quite explains what an MFN clause must do to remove all doubt but strongly suggests that only an explicit reference to dispute resolution in the clause will suffice.

    Berschader later purports to refine the argument. It first questions Plama's 'clear and unambiguous' formulation, questioning, and with good reason, the proposition that an agreement to arbitrate should be construed more restrictively than other agreements. It observes in this regard that '[i]n many jurisdictions, there may in fact exist a tendency to interpret arbitration clauses rather widely.'

    Having thus set the record straight, the tribunal reverts to precisely the 'incorporation by reference' argument on which Plama's presumption largely rests. 'Particular care should be exercised', it warns, 'in ascertaining the intentions of the parties with regard to an arbitration agreement which is to be reached by incorporation by reference in an MFN clause'. It then goes on to adopt the Plama presumption wholesale, stating that MFN jurisdiction will exist only where the terms of the basic treaty 'clearly and unambiguously so provide or where it can otherwise be clearly inferred that this was the intention of the contracting parties.'

    Scott Vesel, the author of the leading article on MFN jurisdiction, Tangled Jurisprudence, perpetuates essentially the same presumption. He at one point states, 'As a general matter of interpretation . . . the advocates of including dispute settlement with the scope of MFN clauses have the stronger case, at least where the clause is broadly worded'. He then goes on, however, to put interpretation entirely aside. After citing certain authorities for the well-worn proposition that consent remains the sine qua non of arbitral jurisdiction, he writes:

    Against this backdrop, and in light of the vigorous debate over whether MFN clauses apply to dispute settlement provisions at all, it would be difficult-if not impossible-to establish a preponderance of evidence in favor of jurisdiction on the basis of an MFN clause combined with consent with regards to a third party.

    While consent must certainly remain the watchword in assessing arbitral jurisdiction, Plama and Berschader overstate the 'particular care' that must be taken in gauging consent where an arbitration agreement has been incorporated by reference. Fouchard et al. state that '[t]here is . . . no reason to take a hostile position towards arbitration clauses incorporated by reference' and that '[a]rbitration agreements incorporated by reference . . . should be interpreted using the general principles of interpretation of arbitration agreements, that is, neither extensively nor restrictively.'

    This pragmatic approach finds expression in the cases as well. Fouchard et al. discuss Bomar Oil, for example, a French case involving a sales transaction concluded by an exchange of telexes, one of which contained an arbitration clause. The Cour de cassation found the agreement valid, and in language very supportive of the notion of MFN jurisdiction. Indeed, if one replaced the word 'contract' with 'treaty' and 'document' with 'third-party treaty', it could serve as a contrary holding in Plama:

    In the field of international arbitration, an arbitration clause, if not mentioned in the main contract, may be validly stipulated by written reference to a document which contains it, for instance general conditions or a standard contract, when the party against which the clause invoked was aware of the contents of this documents at the moment of concluding the contract and when it has, albeit tacitly, accepted the incorporation of the document in the contract.

    Lew et al. discuss a case in which a Swiss court came to a similar conclusion when it held that 'a general reference to a separate document was sufficient to constitute' a binding arbitration agreement. They then cite to Bomar Oil and Spanish authority for the proposition that '[c]ourts in other countries have come to the same conclusion that no specific reference to the arbitration clause is required'.

    Amongst those other countries is the United States, with its courts consistently holding that an arbitration agreement can be incorporated by reference notwithstanding that the main document-that is, the document doing the incorporating-makes no mention of the arbitration provision to be incorporated. In Aceros Prefabricados, S.A. v. TradeArbed, Inc. , for example, the defendant provided to the plaintiff certain sales confirmation orders which incorporated by reference the defendant's General Conditions of Sale. The court found the plaintiff bound by an arbitration clause in the General Conditions of Sale notwithstanding that those General Conditions were not included with the confirmation orders and that the defendant had made no mention to the plaintiff of the arbitration clause.

    The English cases reveal a similar tack. In the Athena, the Commercial Court undertook a comprehensive review of the cases addressing incorporation-by reference to conclude that a contract referring to standard terms, and with no specific mention of dispute resolution, could incorporate an arbitration agreement contained within those terms. The court also noted, in a point relevant to the question of MFN jurisdiction, that terms are terms-be they 'substantive' performance terms or 'procedural' dispute-settlement clauses-and no special rules of incorporation apply to the latter:

    In principle, English law accepts incorporation of standard terms by the use of general words and, I would add , particularly so when the terms are readily available and the question arises in the context of dealings between established players in a well-known market. The principle . . . does not distinguish between a term which is an arbitration clause and one which addresses other issues.

    A word on the Arbitration Act 1996 warrants mention as well. Section 6(2) of the Act is largely modeled on section 7(2) of UNCITRAL Model Law on International Commercial Arbitration ('Model Law'), with the former providing that an arbitration clause may be incorporated into an agreement by reference 'if the reference is such as to make that clause part of the agreement', and the latter stating in almost identical terms that incorporation may be had if 'the reference is such as to make that clause part of the contract.' The Chartered Institute of Arbitrators' commentary to the Act states that section 6(2) 'should resolve that there need not be specific reference to the arbitration clause as a pre-requisite to incorporation.' This proves of interest because Plama read Model Law section 7(2) to precisely contrary effect. It concluded-from what might be described as a 'lemons are oranges' analysis-that the phrase '"if the reference is such as to make that clause part of the contract". . . is another way of saying that the reference must be such that the parties' intention to import the arbitration provision of the other agreement is clear and unambiguous.'

    If one accepts, as these authorities suggest, that an agreement to arbitrate is an agreement like any other, the Plama/Berschader presumption begins to flag. If one also accepts, again on the basis of the authorities, that an arbitration clause is as susceptible as any other term to incorporation by reference, the presumption fails. Plama acknowledges, as indeed it must, that an MFN clause acts to co-opt provisions from third-party treaties. If a dispute-resolution provision is, in the end, just another provision, it too can be co-opted.

    The Debate Over the Extent to Which MFN Jurisdiction Upsets the Balance of the Contracting States' Bargain

    Awards establishing a presumption against MFN jurisdiction, and even those establishing a presumption to the contrary, view MFN jurisdiction as a source of possible mischief. They fear that accepting jurisdiction might upset the 'benefit of the bargain' embodied in the BIT, that it stands to rob the respondent host State of 'the dispute resolution provisions in a specific treaty . . . negotiated with a view to resolving disputes under that treaty ', to 'subvert the intention of the parties to the basic treaty', and to occasion 'disruptive treaty-shopping that would play havoc with the policy objectives of underlying specific treaty provisions'.

    The "Specially Negotiated Provision" Objection to MFN Jurisdiction

    The Telenor tribunal sees this threat as essentially sufficient in itself to reject MFN jurisdiction as a conceptual matter. In listing its 'four compelling reasons' for declining jurisdiction, it states:

    [T]he effect of the wide interpretation of the MFN clause is to expose the host State to treaty shopping by the investor among an indeterminate number of treaties to find a dispute resolution clause wide enough to cover a dispute that would fall outside the dispute resolution clause in the base treaty . . . .

    [T]he wide interpretation also generates uncertainty and instability in that at one moment the limitation in the basic BIT is operative and at the next moment it is overridden by a wider dispute resolution clause in a new BIT entered into by the host State.

    This reasoning, it goes almost without saying, misunderstands the nature and effect of the MFN principle. An MFN clause continuously trolls for discriminatory treatment, existing for no other reason than to allow its beneficiaries to 'shop', to comb through third-party treaties in search of more favourable provisions. As one author put it:

    While the [Telenor] tribunal worried about '"treaty shopping' by the investor," that is arguably the very purpose of MFN clauses in that they ensure investors or investments the best treatment the host State has made available to any investor or investment.

    Moreover, having once found discrimination, the clause operates and is intended to operate so that 'one moment the limitation in the basic [treaty] is operative and at the next moment it is overridden by a [more favourable] clause in a new [third-party treaty]. This is the clause's sole effect; it serves to supplant basic-treaty provisions with kinder provisions in third-party treaties. The fact that the signatories to the basic treaty might have 'specifically negotiated' the term sought to be supplanted is irrelevant. Kevin Kennedy, addressing the operation of the MFN clause in the General Agreement on Tariffs and Trades, explains this elementary point, writing, '[T]he unconditional MFN obligation ensures that if Member A has negotiated specific tariff reductions on goods it exports to Member B in exchange for tariff cuts on goods that Member B exports to Member A, Member A can rest assured that it will not lose the benefit of its bargain if Member B thereafter negotiates an even lower tariff rate on those same goods with Member C.'

    The flaw in the argument that an MFN clause should not be allowed to displace specifically-negotiated provisions comes into stark relief when one takes it to its logical extreme. According to that argument, an MFN clause could reach out to a third-party treaty's provisions on, for example, tariff limits only so long as the basic treaty did not set specific tariff limits. Yet if the basic treaty did not set tariff limits, the MFN clause would very likely be prevented from looking to third-party provisions by sameness prong of the ejusdem generis principle. In short, the MFN clause would become all but a nullity, forbidden to operate, on the one hand, because it relates to tariffs and, on the other, because it does not.

    It warrants recalling on the same point that the Ambatielos Award stands for the proposition that MFN treatment is all the more likely available where the basic treaty and third-part treaty contain provisions addressing the same subject matter. The existence of a specifically-negotiated provision on a given matter in the basic treaty cuts in favor of and not against supplanting that provision with a provision governing the same matter in a third-party treaty.

    The 'Hurdle-Bar' Limitation on MFN Jurisdiction

    Many of the tribunals accepting MFN jurisdiction share Telenor's concerns about benefits of the bargain and treaty shopping but believe they can minimize the peril by exercising jurisdiction judiciously. An MFN clause, they conclude, will be allowed to disrupt the contracting States' bargain in respect of dispute resolution, but only a little. More specifically, they determine that MFN jurisdiction may be had where the MFN beneficiary seeks to displace only that portion of a basic-treaty dispute-resolution provision imposing a pre-arbitration waiting period or, as in the Spain-Argentina BIT, an 18-month pre-arbitration litigation period but not where the beneficiary seeks to displace the basic treaty's dispute-resolution procedures in toto. A beneficiary could not, for example, invoke an MFN clause to compel arbitration of a claim for breach of fair-and-equitable treatment where the basic treaty allows for arbitration only of claims involving expropriation.

    This 'hurdle-bar' approach confounds the MFN standard even more thoroughly than does Telenor's specifically-negotiated-provision argument. The tribunals adopting it had already run the MFN calculus, had measured the scope of the MFN clause and then gone on to determine that matters within its scope were sufficiently similar to matters addressed in a third-party treaty to meet the ejusdem generis rule. This was a legal question, a matter of treaty interpretation calling for a definitive construction one way or the other. Having once determined that an MFN clause reaches to third-party dispute-resolution provisions, a tribunal cannot, at least properly, seek shelter from that decision in an explanation of how softly the decision lands. As one commentator notes:

    If dispute settlement options are indeed part of the treatment guaranteed to investors by BITs generally and MFN clauses specifically, it is not entirely clear why MFN clauses should provide qualifying investors with equal treatment for preliminary matters only, nor why the MFN clauses should permit borrowing another treaty's dispute settlement provisions in part but not in their entirety as a complete and integrated package.

    Given its patent conceptual shortcomings, one might well wonder how a hurdle-bar distinction ever took hold. The distinction, like the debate over whether arbitration's importance to international investment should inform the MFN-jurisdiction analysis, has its genesis in and has been perpetuated by a misunderstanding of Maffezini.

    The case, as discussed earlier, involved a conjoined MFN clause, one which embraced 'all matters governed by' the BIT, on the one hand, and yet restricted itself to 'fair and equitable treatment,' on the other. Because the tribunal at times appears to read the clause as extending only to fair-and-equitable treatment, Maffezini can be understood, at least to this extent, as having accepted jurisdiction upon a determination that 'fair and equitable treatment' and dispute-resolution provisions are ejusdem generis.

    While the tribunal, as noted, accepted MFN jurisdiction, it took pains to limit the reach of that decisions. It stated that '[a]s a matter of principle, the beneficiary of the [MFN] clause should not be able to override public policy considerations that the contracting parties have envisaged as fundamental conditions for their acceptance of the [basic treaty]' and that 'a distinction has to be made between the legitimate extension of rights . . . by means of . . . [an MFN] clause . .. and disruptive treaty-shopping that would play havoc with the policy objectives of underlying specific treaty provisions . . . .' It then provided a non-exhaustive list of situations in which MFN jurisdiction should, in light of these public policy considerations, not lie, for example, where the basic treaty has an exhaustion-of-local remedies or "fork in the road" clause.

    The precise source of Maffezini's 'public policy considerations' has been the object of head-scratching by later tribunals and commentators. Yet Maffezini's reasoning adds up to the extent one views the case as finding dispute resolution and fair-and-equitable treatment ejusdem generis. If a BIT's MFN clause indisputably embraces the treaty's dispute resolution provisions-suppose, for example, that the MFN clause expressly reached those provisions-it makes little sense to worry about 'treaty shopping' or to question whether the BIT's dispute-resolution provisions may be replaced in toto. Indeed, as noted, that approach would eviscerate the MFN standard.

    Yet limits make a great deal more sense where a tribunal determines that an MFN clause extends to the basic treaty's guarantee of fair-and-equitable treatment but not to its dispute resolution provisions and then goes on to determine that dispute resolution mechanisms form a part of the fair-and-equitable standard. That standard, at least according to some, has no strictly defined limits, harkens to generalized notions of fairness, policy and principle. As one author explains, a BIT's fair-and-equitable-treatment provision is the 'catch-all', the clause used 'to fill gaps which may be left by the more specific standards, in order to obtain the level of investor protection intended by the treaties.'

    Unfortunately, Maffezini's limitations based on 'public policy considerations' have gained an independent existence, one not tied to the peculiar MFN clause at issue in the case. Plama, for example, approves Maffezini's warning about 'disruptive treaty shopping' on its way to declining MFN jurisdiction. National Grid, which accepts jurisdiction, states:

    The Tribunal concurs with Maffezini's balanced considerations in its interpretation of the MFN clause and with its concern that MFN clauses not be extended inappropriately. It is evident that some claimants may have tried to extend an MFN clause beyond appropriate limits. For example, the situation in Plama involving an attempt to create consent to ICSID arbitration when none existed was foreseen in the possible exceptions to the operation of the MFN clause in Maffezini.

    The perception, at least amongst some practitioners, has been that the hurdle-bar distinction describes the state of the law, and perhaps a rule of law. One author questions the logic of extending MFN treatment only to procedural hurdles but sees the limitation as a 'possible implication of the recent cases'. Another sees more than an implication, writing that '[t]he rule that appears to emerge [is that] MFN clauses can be used to overcome waiting periods and similar admissibility requirements, but not to replace, in whole or in part, the dispute resolution mechanism provided in the treaty upon which jurisdiction is based.' This rule, to the extent it exists, is unfortunate both because it is based on a misreading of Maffezini and, more important, because it does such violence to the MFN standard.

    One final aspect of the tribunals' mistreatment of the MFN standard warrants brief mention. Most of the presuming tribunals have relied on the historical treaty practice of the State parties to the BIT at issue. Berschader, for example, centered on the Belgium-USSR BIT, which made room for arbitration only of claims of compensation for expropriation. The Belgian claimant sought MFN jurisdiction over an expropriation claim (which would involve the additional question whether an expropriation occurred) on the basis of the Denmark-Russian Federation BIT, which allowed for arbitration of any dispute 'in connection with an investment'.

    The Russian Federation argued, and the tribunal accepted, that MFN jurisdiction should not lie because, among other things, the USSR consistently excluded expropriation claims from their BITs during the period of time in which the Belgium-USSR treaty was concluded. 'This consistent practice on the part of the Soviet Union', writes the tribunal, strongly suggests that the Soviet Party did not intend the MFN provision in Article 2 to extend to dispute resolution'.

    This cannot be right. Returning to the earlier hypothetical, this reasoning would have Fredonia properly refusing to reduce tariffs from the 20 kroner per bushel it exacted from its forty trading partners to the 10-kroner level it promised to Ruritania solely on the basis that that it had initially exacted a 20-kroner tariff from its forty trading partners. This reasoning leaves no room whatever for the MFN standard; it renders the MFN clause a dead letter.

    At times, the presuming tribunals' concern that MFN jurisdiction will rob the host state of its due seems the only concern. Other lines of argumentation-those relating to the meaning of earlier decisions, the importance of arbitration, and the evidence of consent-often appear to float on an undercurrent of suspicion of or outright dismay over an MFN clause's power to wipe away a treaty provision for which the host State might well have bargained fiercely or which it obtained only by granting considerable concessions. Time and again, the tribunals evince an unwillingness to view an MFN clause for what it is, to allow an MFN clause to do what it does. They evince marked antipathy towards the MFN standard, seeing it not as the great leveler-'generalizing concessions and . . . furthering and insuring international commercial equality'-but as a mischief maker out to disrupt the equilibrium, to upset the balance of the contracting States' bargain.

    They would do well in this regard to recall that an MFN clause is part of the bargain. A nation might from time to time rue its decision to include an MFN clause in a given treaty but will, in the main, be glad it is there. The clause, as Schwarzenberger explains, has its price, but one that virtually every nation is willing to pay:

    Though any State is well content to be treated by another on more favourable terms than any third State and does not object to discrimination practiced in its favour, everyone of them must be continuously on its guard against any more successful competitor. However, assuming that State are prepare to exchange a condition of unceasing vigilance and never-ending uneasiness for the safer and more dignified position in which anybody's advantage accrues to everybody's profit, the standard of m.f.n. treatment is the very means to this end.

    Conclusion

    One is almost expected in matters such as these to conclude with a rule, to muster the strength of the authorities under discussion and then boldly proclaim the way forward. In the case at hand, this would presumably involve some pronouncement on when MFN jurisdiction should lie and when it should not.

    Any such pronouncement would, of course, betray the idea resting at the dissertation's very heart. To the extent the dissertation formulates a rule, it is that there is no rule. Assessing MFN jurisdiction remains, at bottom, an exercise in treaty interpretation. Because every MFN clause is a thing unto itself, that interpretation must be de novo, undertaken unencumbered by predilection and predisposition.

    This is not to say that the authorities under discussion provide no assistance whatever in the interpretive exercise. The Vienna Convention's rules, for example, mandate a textualist approach to interpretation, which accords primacy to the text of a treaty itself. While a BIT's stated 'object and purpose'-its express desire 'to develop economic cooperation' or 'to create favourable conditions for investments'-must be taken into account, that object and purpose must inform rather than drive interpretation.

    The ICJ decisions and the Ambatielos Award provide somewhat more specific guidance. The Ambatielos Case and Award support the propositions that a broadly-worded MFN clause possesses a broad sweep and that a clause will more likely be viewed as bringing a matter within its scope if the basic treaty in which it resides specifically addresses that matter in its other provisions. Nationals in Morocco reinforces those propositions and establishes as well that an MFN clause may serve to displace very precise and comprehensive provisions in a basic treaty and may embrace basic-treaty provisions relating to jurisdiction.

    Secondary sources and case law from England, Europe and the United States combine to make clear that an arbitration agreement is an agreement like any other. Evidence of consent to arbitrate must exist, to be sure, but it need be no clearer or less ambiguous than is the evidence required to confirm the existence of any contractual undertaking. Neither must 'particular care' be taken where an arbitration agreement is claimed to have been incorporated by reference. Establishing effective incorporation of an arbitration clause is no harder, and no easier, than is establishing the incorporation of a 'regular' contract provision.

    Finally, myriad sources make abundantly evident that while the MFN principle can serve to startle, to disrupt existing bargains and otherwise to shake things up, those are, in the end, its intended functions. States bind themselves to the principle through MFN clauses precisely to enjoy those effects. That a State tumbling into the downward spiral of ever greater concessions will from time to time regret the decision is certainly understandable. It can be no reason, however, to deprive an MFN clause of its intended force.

    Bibliography

    Books

    R. Bernhardt (ed), Encyclopedia of Public International Law (Amsterdam: North-Holland Publishing Company, 1995)

    R. Dolzer and M. Stevens, Bilateral Investment Treaties (London: Martinus Nijhoff Publishers, 1995)

    L. Edmondson et al.,Domke on Commercial Arbitration (Egan, Minnesota: West Group, 2003)

    Fouchard et al., Fouchard, Gaillard, Goldman on International Commercial Arbitration (Cambridge, Massachusetts: Kluwer Law International, 1999)

    B. Harris et al., The Arbitration Act 1996: A Commentary 3rd ed (Oxford: Blackwell Publishing, 3rd ed, 2003)

    S.K. Hornbeck, (1911) 'The Most-Favored-Nation Clause in Commercial Treaties: Its Function in Theory and in Practice and its Relation to Tariff Policies' Bulletin of the University of Wisconsin 331

    J. Lew et al., Comparative International Commercial Arbitration (Cambridge, Massachusetts: Kluwer Law International, 2003)

    E. Ludwig, Consular Treaty Rights and Comments on the 'Most Favored Nation Clause ' (Akron, Ohio: New Werner Company, 1913)

    C. McLachlan et al., International Investment Arbitration: Substantive Principles (Oxford: Oxford University Press, 2007) § 7.162.

    A.D. McNair, The Law of Treaties (Oxford: Clarendon Press, 1962) 273 (emphasis in original).

    A. Redfern et al., Law and Practice of Commercial Arbitration (London: Sweet & Maxwell. 2004)

    L. Reed et al., A Guide to ICSID Arbitration (Cambridge, Massachusetts: Kluwer Law International, 2004)

    J. Paulsson, Denial of Justice in International Law (Cambridge: Cambridge University Press, 2005) 134.

    Articles, Conference Papers and Working Papers

    D.J. Bederman, 'Revivalist Canons and Treaty Interpretation', (1994) 41 University of California Law Review 953, 973.

    V.G. Curran, Romantic Common Law, Enlightened Civil Law: Legal Uniformity and the Homogenization of the European Union (2001) 7 Columbia Journal of European Law 63

    R. Dolzer, 'Fair and Equitable Treatment: A Key Standard in Investment Treaties' (2005) 39 International Lawyer 87, 90.

    G. Egli, 'Don't Get BIT: Addressing ICSID's Inconsistent Application of Most-Favored-Nation Clauses to Dispute-Resolution Provisions' (2007) 34 Pepperdine Law Review 1045

    P. Foy and R. Deane, 'Foreign Investment Protection Under Investment Treaties: Recent Developments under Chapter 11 of the North American Free Trade Agreement' (2001) 16 ICSID Review - Foreign Investment Law Journal 299

    E. Franckx (June 2000) 'Pacta Tertiis and the Agreement for the Implementation of the Provisions of the U.N. Convention on the Law of the Sea' FAO Legal Papers Online #8 (June 2000)

    D.H. Freyer and D. Herlihy, 'Most-Favored-Nation Treatment and Dispute Settlement in Investment Arbitration: Just How "Favored" is "Most-Favored?' (2005) 20 ICSID Review - Foreign Investment Law Journal 58, 82.

    M. Friedman et al., 'International Legal Developments in Review: 2006' (2007) 41 International Lawyer 251, 270.

    J.T. Gathii, 'Re-Characterizing the Social in the Constitutionalization of the WTO: A Preliminary Analysis' (2001) 7 Widener Law Symposium Journal 137, 159.

    A. Glashauser, 'What We Must Never Forget When it Is a Treaty We Are Expounding' (2005) University of Cincinnati Law Review 1243, 1260.

    C. Hamilton and P. Rochwerger, 'Trade and Investment: Foreign Direct Investment Through Bilateral and Multilateral Treaties' (2005) New York International Law Review 1

    M.F. Houde, 'Most-Favoured Nation Treatment in International Investment Law', (2004) Working Papers on International Investment Number 2004/2

    G. Kaufmann-Kohler, 'Arbitral Precedent: Dream, Necessity or Excuse' (2007) 3 Arbitration International 357, 370 (emphasis in original).

    K. Kennedy, 'The GATT-WTO System at Fifty' (1998) 16 Wisconsin International Law Journal 421, 429.

    P. Maki, 'Interpreting GATT Using the Vienna Convention' (2000) Minnesota Journal of Global Trade 343

    J.G. Merrills, 'Two Approaches to Treaty Interpretation' (1969) Yearbook of International Law 55, 58.

    Y. Radi, 'The Application of the Most-Favoured-Nation Clause to the Dispute Settlement Provisions of Bilateral Investment Treaties: Domesticating the "Trojan Horse' (2007) European Journal of International Law 757

    K.P. Sauvant et al., 'Most-Favoured Nation Treatment' (1999) UNCTAD Series on Issues in International Investment

    K.P. Sauvant et al., 'Dispute Settlement: Investor-State' (2003) UNCTAD Series on Issues in International Investment

    K. Scholze, 'Having Your Pie . . . and Eating it With one Chopstick - Most Favoured Nation Clauses and Procedural Rights' (2004) 5 Policy Papers on Transnational Economic Law 1

    G. Schwarzenberger, 'The Most-Favoured-Nation Standard in British State Practice' (1945) 22 British Yearbook of International Law 96

    B. Sheppard, International Legal Developments in Review: 2006 Disputes, International Arbitration, (2007) 41 International Lawyer 251

    J.C. Thomas, 'The Most-Favoured-Nation Clause: An Overview of the Principal Issues of Interpretation', Conference Paper presented at the International Bar Annual Conference in Singapore (2007)

    E. Ustor, 'The Most-Favoured-Nation Clause in the Law of Treaties' (1968) Yearbook of the International Law Commission Vol. II 165, 166.

    E. Ustor, Most-Favoured Nation Clause, Yearbook of the International Law Commission , (1968) Vol. II 165

    E. Ustor, 'First Report on the Most-Favoured-Nation Clause' (1969) Yearbook of the International Law Commission Vol. II 157

    E. Ustor, 'Second Report on the Most-Favoured-Nation Clause' (1970) Yearbook of the International Law Commission Vol. II 199

    E. Ustor, 'Third Report on Most-Favoured-Nation Clause' (1972) Yearbook of the International Law Commission Vol. II 161

    E. Ustor, 'Fourth Report on Most-Favoured-Nation Clause' (1973) Yearbook of the International Law Commission Vol. II 97

    S. Vesel, 'Clearing a Path Through a Tangled Jurisprudence: Most-Favored-Nation Clauses and Dispute Settlement Provisions in Bilateral Investment Treaties' (2007) 32 Yale Journal of International Law 125

    J. Wong, 'The Application of Most-Favored-Nation Clauses to Dispute Resolution Provisions in Bilateral Investment Treaties' (2008) 3 Asian Journal of WTO & International Health Law and Policy 171

    Publications of the International Law Commission

    ILC Draft Articles on Most-Favoured-Nation Clauses (1978)

    International Conventions

    Agreement the Government of the Kingdom of Norway and the Government of the Republic of Hungary on the Promotion and Reciprocal Protection of Investments

    Vienna Convention on the Law of Treaties

    English Legislation

    Arbitration Act 1996

    International Court of Justice Decisions

    Ambatielos Case (Merits: Obligation to Arbitrate) 1953 ICJ Reports 10

    Anglo-Iranian Oil Company Case 1952 ICJ Reports 93

    Case Concerning Rights of Nationals of the United States of America in Morocco 1952 ICJ Reports 176

    Conditions of Admission of a State to Membership in the United Nations , Advisory Opinion , 1948 ICJ Reports 57

    English Decisions

    The Athena [2007] 1 Lloyd's Rep. 280

    United States Decisions

    Aceros Prefabricados, S.A. v. TradeArbed, Inc, 282 F. 3rd 92 (2nd Cir. 2002)

    Energy Transport, Ltd. v. M.V. San Sebastian 348 F.Supp.2d 186 (S.D.N.Y. 2004)

    Moses H. Cone Memorial Hospital v. Mercury Construction Corp., 460 U.S. 1 (1983).

    Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc. 473 U.S. 614 (1985)

    Arbitral Awards

    The Ambatielos Claim , XII U.N. Reports of International Arbitration Awards 83 (6 March 1956) 73.

    Amco Asia Corp. v. Indonesia, ASEAN Case No. ARB/81/1, Award on Jurisdiction (25 September 1983), (1984) 23 International Legal Materials 351

    Berschader v. Russian Federation , SCC Case No. V080/2004, Award (21 April 2006)

    Camuzzi International S.A. v. Argentina , ICSID Case No. ARB/03/2, Decision on Objections to Jurisdiction (11 May 2005)

    Gas Natural SDG v. Argentina , ICSID Case No. ARB/03/10, Decision of the Tribunal on Preliminary Questions on Jurisdiction (17 June 2005)

    Maffezini v. Spain , ICSID Case No. ARB/97/7, Decision on Jurisdiction (25 January 2000)

    M.C.I. Power Group L.C. and New Turbine, Inc. v. Ecuador , ICSID Case No. ARB/03/6, Award (31 July 2007)

    National Grid plc v The Argentine Republic , UNCITRAL, Decision on Jurisdiction (20 June 2006)

    Plama Consortium v. Bulgaria , ICSID Case No. ARB/03/24, Decision on Jurisdiction (8 February 2005)

    RosInvestCo U.K. Ltd. v. Russian Federation , Stockholm Chamber of Commerce Case No. V079/2005, Award (October 2007)

     

    Salini Costruttori S.p.A. v. Jordan , ICSID Case No. ARB/02/13, Decision of the Tribunal on Jurisdiction (29 November 2004)

    Siemens A.G. v. Argentina , ICSID Case No. ARB/02/8, Decision on Jurisdiction (3 August 2004)

    Suez, Sociedad General de Aguas de Barcelona S.A. and Interagua Servicios Integrales de Agua S.A. v. Argentina , ICSID Case No. ARB/03/17 (16 May 2006)

    Suez, Sociedad General de Aguas de Barcelona S.A., and Vivendi Universal S.A. v. Argentina , ICSID Case No. ARB/03/19, Decision on Jurisdiction (3 August 2006)

    Tecnicas Medioambientales Tecmed, S.A. v. Mexico , ICSID Case No. ARB (AF)/00/2, Award, (May 29, 2003)

    Telenor Mobile Communications AS v. Republic of Hungary , ICSID Case No. ARB/04/15, Award (13 September 2006)

    Yaung Chi Oo Trading v. Myanmar Pte Ltd. v. Myanmar , ASEAN I.D. Case No. ARB/01/1, Award (31 March 2003)