I. Introduction
Most-favored-nation (MFN) clauses have been included in international commercial treaties for many centuries.Footnote 1 They also figure prominently as standard provisions in almost any international investment agreement (IIA).Footnote 2 Their longstanding and widespread use notwithstanding, investment law doctrine and arbitral practice continue to struggle with the clauses’ application and interpretation, in particular as regards their scope of application. What Stanley Hornbeck observed more than one hundred years ago in this Journal, that “there appear[s] constant disagreements and ever-recurring irritation over what is the meaning and what are the obligations attaching to this or that [MFN] clause,”Footnote 3 still characterizes the practice of investment tribunals and the literature on MFN clauses in IIAs today.
After focusing for a long time on their application to more favorable dispute settlement provisions,Footnote 4 the “constant disagreements and ever-recurring irritation” in the interpretation of MFN clauses, as Hornbeck called it, is now turning to whether investors can use MFN clauses to rely on more favorable substantive treatment contained in the host state's IIAs with third countries.Footnote 5 The award in İçkale v. Turkmenistan,Footnote 6 more restrictive drafting of MFN clauses in newer IIAs—including in the European Union-Canada Comprehensive Economic and Trade Agreement (CETA)Footnote 7 —and the position expressed by the contracting parties to the North American Free Trade Agreement (NAFTA)Footnote 8 mark the start of that debate. Building on these instances, Simon Batifort and J. Benton Heath go a step further and call for a fundamental reevaluation of the application of MFN clauses generally to better substantive standards of treatment absent from the basic treaty but included in other IIAs between the host state and a third country.
In their view, the prevailing approach in investment arbitration is based on “a conventional wisdom”Footnote 9 that not only results in unbalanced decisionsFootnote 10 but is also flawed in
rely[ing] on a particularly strong form of ‘top-down’ reasoning, which imposes presumptions as to the nature or essence of MFN clauses in general … [,] obscures the variation among MFN clauses in investment agreements and other treaties, and leads interpreters to adopt an unduly uniform approach to the function of MFN provisions.Footnote 11
To remedy these flaws, Batifort and Heath suggest, as their contribution to the “new MFN debate,” a “treaty-by-treaty, ‘bottom-up’ approach” to interpreting MFN clauses.Footnote 12 This approach questions the plausibility of allowing investors to benefit from more favorable standards of treatment in third-country IIAs. It also casts doubt on the proposition that the investment treaty regime is multilateralizing through the use of MFN clauses in bilateral treaties.Footnote 13
Batifort and Heath deserve credit for redirecting the debate away from the increasingly stale and stalemated issue of the application of MFN clauses to questions of procedure and jurisdiction and for directing attention to their effect on differential substantive standards of treatment granted under different IIAs—even if they do not develop their own positions on the issues they identity as crucial for the “new MFN debate.”Footnote 14 Their analysis contains a number of valid points. I agree with them, for example, that the reasoning of many arbitral tribunals, in particular in earlier cases, is weak and unsatisfying, and therefore constitutes a brittle basis for constructing a jurisprudence constante on the issue in question. I also agree with them that the interpretation of MFN clauses in IIAs has to start with a treaty-by-treaty approach that is attentive to the text, context, and object and purpose of the clause in question. Consequently, there is no question for me that the interpretation of MFN clauses in newer treaties, such as that in Article 8.7 of CETA, will result in declining attempts by investors to rely on better standards of treatment granted under the host state's third-country IIAs.Footnote 15
But I disagree with Batifort and Heath's call for a general reevaluation of the application and interpretation of MFN clauses to better substantive standards of treatment. While partly poorly reasoned, the existing jurisprudence constante on the issue—the “conventional wisdom”—is not, as Batifort and Heath claim, based on a “top-down” approach to interpretation; did not emerge in reaction to the post-Maffezini debate on whether MFN clauses applied to questions of arbitral procedure and jurisdiction; and has not resulted in incorrect applications allowing covered investors to rely on better substantive standards of treatment. On the contrary, the prevailing approach in investment jurisprudence starts, as it should, from a treaty-by-treaty, bottom-up approach that accords principal weight to the interpretation of the individual MFN clauses in the applicable investment treaty.
At the same time, the prevailing approach does not confine the interpretative exercise to the bilateral treaty in question. It builds on the foundations of MFN clauses in the general international law of treaties. Under general international law, as accepted by other international courts and tribunals, most importantly the International Court of Justice (ICJ), and as reflected in the work of the International Law Commission (ILC), MFN clauses are to be interpreted against a multilateral background. Moreover, their wording permitting, they apply to more favorable treatment that the granting state extends to third states and their subjects, including investors, in international agreements. This practice backs the prevailing interpretation of MFN clauses in traditional IIAs and the resulting multilateralization of bilateral treaty commitments. This approach is also unaffected by the more restrictive wording of MFN clauses in newer IIAs, such as Article 8.7 of CETA. These treaties, as will become clear, mark a departure from the traditional approach to MFN treatment in IIAs and can therefore not be used as an indication of how contracting states understood MFN clauses in older IIAs.Footnote 16
In my view, Batifort and Heath misunderstand the resulting “MFN paradox,” namely that MFN clauses, including in traditional IIAs, are bilateral commitments to multilateralism. They are multilateralization devices cast in bilateral form that prevent the states granting MFN treatment from shielding more favorable bilateral bargains contained in international treaties with third states from multilateralization. This is not a question of understanding MFN clauses from a top-down perspective and imposing preconceived notions on the treaty-interpreter about the function and effect of the clauses. Instead, this commitment in bilateral treaties to multilateralism is what states intended to make by including MFN clauses in their investment treaty relations.
For this reason, I disagree with Batifort and Heath that developments in the drafting of MFN clauses in more recent IIAs, together with the isolated award in İçkale v. Turkmenistan and the position of the contracting parties on the MFN clause in NAFTA, should lead to a reevaluation of the interpretation and application of MFN clauses in IIAs more generally. This claim is not only doctrinally flawed. It is a politically and ideologically questionable assault on the commitment of contracting states under MFN clauses to the multilateral structures in international investment law, which have been built over the past decades through treaty-making and dispute settlement, and which are currently being reformed in newer treaties.
To develop my critique, I will first address how Batifort and Heath mischaracterize the development and content of the “conventional wisdom” in investment treaty jurisprudence on the application of MFN clauses to better substantive standards of treatment (Part II). Second, I will turn to the general international law background on the interpretation of MFN clauses and show that the treaty-by-treaty approach mandated under that framework supports the result reached by the “conventional wisdom” in investment arbitration (Part III). Third, I will focus on the practice on Article 1103 of NAFTA and the Award in İçkale v. Turkmenistan and demonstrate that these instances do not suggest a different approach (Part IV). I will close with remarks on the ideological foundations that Batifort and Heath build on, which leads to undermining treaty commitments that constitute part of the multilateral foundations of the investment treaty regime.
II. Lifting the Blinders: Batifort and Heath's Misrepresentation of the Prevailing Approach to MFN Clauses in Investment Jurisprudence
Batifort and Heath present their “bottom-up” approach to the interpretation of MFN clauses as an innovative contribution to the debate. However, in doing so, they misrepresent the prevailing approach in investment law to the interpretation and application of MFN clauses in two important respects.Footnote 17 This approach is neither based on “top-down” interpretation that gives short shrift to the text of the MFN clauses in question, nor has it developed in reaction to the post-Maffezini debate. On the contrary, the prevailing approach existed before Maffezini and starts from a treaty-by-treaty analysis of the individual MFN clause in question.
The Prevailing Approach Is Not a Product of Post-Maffezini Debates
First of all, Batifort and Heath are incorrect in attributing the origin of the “conventional wisdom” on the application of MFN clauses to more favorable substantive standards of treatment to the post-Maffezini debate.Footnote 18 Neither was Maffezini v. Spain “the first published investment treaty case to permit the importation of treaty provisions, whether substantive or procedural, via MFN,”Footnote 19 nor was the view that MFN clauses in IIAs could allow an investor to attract benefits granted by the host state in IIAs with third states a product or projection of the post-Maffezini debate.Footnote 20
Instead, the view that MFN clauses covered better substantive standards of treatment in the host state's third-country IIAs has been firmly entrenched since the beginning of investment arbitration. Thus, in 1990, ten years prior to Maffezini, the Tribunal in Asian Agricultural Products Ltd. v. Sri Lanka, the first known investment treaty arbitration ever, accepted in principle that an investor covered by the MFN clause in the basic treaty, the U.K.-Sri Lanka Bilateral Investment Treaty (BIT), could rely on more favorable substantive treatment granted under other Sri Lankan BITs.Footnote 21 The Tribunal, however, also found that the investor could not show that the Switzerland-Sri Lanka BIT (the treaty alleged to be more beneficial) provided for a stricter liability standard than the U.K.-Sri Lanka BIT.Footnote 22 Certainly, the reasoning of the Tribunal in AAPL was anything but elaborate, but that was likely due to the fact that the application of the MFN clause to cover better treatment from Sri Lanka's other BITs was so obvious to the Tribunal as to render further discussion unnecessary.Footnote 23
Batifort and Heath also fail to mention or acknowledge the relevance of other early decisions accepting reliance on better substantive treatment granted under the host state's third-country IIAs. For example, in CME v. Czech Republic, the Tribunal held that an MFN clause allowed an investor to rely on more favorable standards of compensation for expropriation granted under the host state's third-country IIAs.Footnote 24 Similarly, ADF Group v. United States, an early NAFTA case, can be read as accepting, in principle, that MFN clauses in IIAs apply to better substantive standards of treatment.Footnote 25
More importantly, Batifort and Heath reduce the relevance of the decisions in Pope & Talbot to the interpretation of the fair and equitable treatment (FET) provision in Article 1105 of NAFTA .Footnote 26 Pope & Talbot, however, is crucial as an early reflection in investment arbitration on the application of MFN clauses to substantive standards of treatment and on the lack of controversy about the issue at the time. Thus, when discussing the scope of NAFTA's FET standard, the Tribunal faced two possibilities. The first and more restrictive position, invoked by Canada, asserted that the standard was equivalent to the customary international law minimum standard expressed in the 1920s Neer case.Footnote 27 The second position supported a free-standing and arguably broader interpretation of FET.Footnote 28
The Tribunal concluded initially that preference should be given to the latter interpretation, inter alia, because of the presence of the MFN clause in Article 1103 of NAFTA, which would entitle investors to the broader interpretation of FET adopted in the respondent's BITs with third countries.Footnote 29 While technically not entering a holding on Article 1103 of NAFTA, the Tribunal's understanding that this provision applied to benefits granted under third-country IIAs is manifest. What is more, after NAFTA's Free Trade Commission had interpreted Article 1105 as an expression of the international minimum standard,Footnote 30 the Pope & Talbot Tribunal reaffirmed its position that NAFTA's MFN clause would allow reliance on a more favorable FET standard granted under other Canadian BITs,Footnote 31 even though the decision itself did not turn on that issue anymore.Footnote 32
As these cases show, the view that MFN clauses in IIAs, in principle, can apply to more favorable standards of treatment in a host state's third-country IIAs has been firmly established pre-Maffezini and was not, as Batifort and Heath claim, a product of post-Maffezini debates.
The Prevailing Approach Does Not Follow a “Top-Down” Approach to Interpretation
Second, and perhaps more importantly, the prevailing doctrine on the interpretation of MFN clauses in IIAs by no means follows a “top-down” approach, which “imposes presumptions as to the nature or essence of MFN clauses in general”Footnote 33 that override the intention of contracting states. Instead, the prevailing approach to applying and interpreting MFN clauses in IIAs is based on the same treaty-by-treaty approach Batifort and Heath plead allegiance to.Footnote 34 This becomes particularly clear in the various arbitral decisions dealing with the application of MFN clauses to questions of dispute settlement, where the wording of the clauses plays a key role.Footnote 35
That the prevailing doctrinal position consists in a treaty-by-treaty approach to interpreting MFN clauses can also be gleaned from two particularly authoritative restatements. To start with, the Institut de Droit International, one of the most revered learned societies in international law—whose membership comprises most of the world's most highly regarded public international lawyers, several of whom also regularly sit as arbitrators in investment treaty arbitrationsFootnote 36 —restated the approach to the interpretation of MFN clauses in investment treaties in its 2013 Tokyo Resolution on “Legal Aspects of Recourse to Arbitration by an Investor Against the Authorities of the Host State Under Inter-State Treaties” as follows: “Most favoured nation treatment requires interpretation of the specific wording of the clause of the treaty in which it is inserted, in order to respect the intentions of the States parties.”Footnote 37
The ILC's Study Group on the Most-Favored-Nation Clause, which addressed the interpretation of MFN clauses in IIAs, reached the same result in its 2015 Final Report:
The interpretation of MFN clauses is to be undertaken on the basis of the rules for the interpretation of treaties set out by the VCLT [i.e. Vienna Convention on the Law of Treaties, articles 31–32]. The central interpretative issue in respect of the MFN clause relates to the scope of the clause and the application of the ejusdem generis principle. That is, the scope and nature of the benefit that can be obtained under an MFN provision depends on the interpretation of the MFN provision itself.Footnote 38
The prevailing doctrine on the interpretation of MFN clauses in IIAs, as these examples show, does not endorse a “top-down” approach to interpretation, but starts from, and accords principal weight to, determining the scope of the individual MFN clause in question. Contrary positions in the scholarship on international investment law are the exception.Footnote 39
The reason why investment jurisprudence has so consistently allowed investors to rely on better substantive standards of treatment on the host state's third-country IIAs then is not the existence of a top-down approach to interpretation that would override the specificities of MFN clauses. The jurisprudence constante is principally due to the limited relevance of variations in the wording of MFN clauses that came up in actual disputes. The results will (have to) be different, however, once MFN clauses of the sort included in CETA are in question. But for now, MFN clauses—in particular those that refer simply to “better” treatment or “all” treatment, but possibly also those applying to “treatment related to the management, maintenance, use, or disposal of investment” with or without a qualifier clarifying that investors must be “in similar situations”—can faithfully be interpreted as allowing covered investors to rely on better substantive standards of treatment granted in one of the host state's third-country IIAs.Footnote 40
This result is also supported by the general international law framework governing the interpretation of MFN clauses in any bilateral treaty, including an IIA, addressed next.
III. MFN Clauses as Multilateralization Devices in General International Law
Batifort and Heath not only misrepresent the state-of-the-art in arbitral practice and doctrine on the interpretation of MFN clauses in IIAs. They also ignore the role that general international law plays for the interpretation and application of MFN clauses. A consideration of this background shows not only that the rules for interpreting MFN clauses themselves are multilateral, but also that the objective of MFN clauses is to multilateralize, for beneficiary states and their investors, commitments the granting state makes in international agreements with third countries in respect of substantive standards of treatment.
The Multilateral Rules on Interpreting MFN Clauses
Batifort and Heath present MFN clauses in IIAs as treaty provisions in solitary confinement. Under their treaty-by-treaty approach, the MFN clause in the treaty in question is the sole normative material relevant for interpretation. Most importantly, they aim to isolate each MFN clause not only from the treaty practice on MFN clauses in other IIAs and their interpretation in arbitral jurisprudence, but also from the normative background that general international law provides for the interpretation of such clauses. Taking into account this normative background shows that the interpretation of MFN clauses is not limited to the clause and the treaty in question but itself follows multilateral rationales.
The most authoritative expression of the general international law rules on the interpretation of MFN clauses are still the Draft Articles on Most-Favoured-Nation Clauses adopted in 1978 by the ILC,Footnote 41 which Batifort and Heath wrongly dismiss as “provid[ing] little insight into the questions raised in contemporary arbitrations.”Footnote 42 After all, with respect to their core provisions on the functioning, effect, and interpretation of MFN clauses, the Draft Articles enshrine what must be considered customary international law rules concerning the interpretation of MFN clauses contained in any international treaty. IIAs are no exceptions. In fact, the ILC itself recognized in its 2015 Final Report that “MFN clauses remain unchanged in character from the time the 1978 draft articles were concluded. The core provisions of the 1978 draft articles continue to be the basis for the interpretation and application of MFN clauses today.”Footnote 43
The 1978 ILC Draft Articles, of course, do not define the specific benefits covered by a given MFN clause in an IIA. But they reflect the states parties’ intentions when including such clauses in their international treaty relations to create common definitions and objectives of MFN clauses (see Articles 4 and 5 of the 1978 ILC Draft Articles) and to lay down multilateral rules for determining the scope of application of each singular MFN clause, in particular the ejusdem generis rule (see Article 9 of the 1978 ILC Draft Articles). Consequently, even though interpretation has to proceed treaty-by-treaty, the exercise is neither isolated from the IIA practice of the two contracting states, nor from the IIA practice of states more generally.
On the contrary, in order to understand what states intended when including a specific MFN clause in a concrete IIA, it will be necessary to compare that clause to other MFN clauses and to ask whether it diverges in a relevant way from, or if it follows, general IIA practice. In the first case, the scope of the MFN clause may differ from that in other IIAs; in the latter case, generalizations and interpretative presumptions about what states intended when using standardized MFN language are called for. Either way, the interpretation of MFN clauses in IIAs is never isolated from general international law and IIA practice more generally.Footnote 44 Instead, interpretation of MFN clauses is based on multilateral rules, includes multilaterally accepted interpretative presumptions, and cannot be limited to an analysis of the singular clause in question.
Application of MFN Clauses to Better Treatment Granted Under Third-Country IIAs
The customary rules on interpretation as reflected in both the 1978 ILC Draft Articles and the jurisprudence of other international courts and tribunals, including most importantly that of the ICJ and under the World Trade Organization (WTO) Dispute Settlement Understanding, lead to the well-established result that benefits under an international treaty between the granting state and a third state are generally covered by an MFN clause, unless the clause makes clear otherwise. The caution Batifort and Heath call for in this respect is therefore misplaced.
To start with, the 1978 ILC Draft Articles and the accompanying Commentary could not be clearer about the presumption that MFN clauses apply, within their scope of application, to more favorable treatment extended by the granting state to a third state or its nationals through an international treaty. Both Articles 5 and 8(2) of the ILC Draft Articles generically use the term “treatment” as the relevant object of multilateralization under an MFN obligation without distinguishing between benefits granted under domestic law and practice or international agreements. Article 17 of the ILC Draft Articles further clarifies that it is irrelevant whether the treatment in question “has been extended under an international agreement, whether bilateral or multilateral.” Clearly, therefore, the term “treatment” as used by the 1978 ILC Articles encompasses better treatment extended on the basis of international agreements.
This is also reiterated in the ILC's Commentary to Article 5, which states:
It is not necessary … that the treatment actually extended to the third State, with respect to itself or the persons or things concerned, be based on a formal treaty or agreement. The mere fact of favourable treatment is enough to set in motion the operation of the clause. However, the fact of favourable treatment may consist also in the conclusion or existence of an agreement between the granting State and the third State by which the latter is entitled to certain benefits. The beneficiary State, on the strength of the clause, may also demand the same benefits as were extended by the agreement in question to the third State. The mere fact that the third State has not availed itself of the benefits which are due to it under the agreement concluded with the granting State cannot absolve the granting State from its obligation under the clause.Footnote 45
Two important points derive from this statement. First, unless a contrary intention of the states parties to the MFN clause can be shown, the form in which better treatment is extended to the third state—through an international agreement, based on domestic law, or de facto treatment—is generally irrelevant for the operation of the MFN clause.Footnote 46 Second, it is equally irrelevant whether the third state, or its nationals, have in fact availed themselves of the benefit in question, or whether they have only been given the right or possibility to avail themselves of it. In either case, the beneficiary state, or its nationals, can invoke the better treatment through the MFN clause in the basic treaty. For the context of IIAs this means that MFN clauses allow covered investors to rely on better substantive standards of treatment granted in the host state's third-country IIAs, unless the clause's wording suggests otherwise, and to do so independently of whether a concrete third-country investor has actually benefitted from the better treatment in question or not.
The same result, namely that MFN clauses apply to better treatment under the granting state's international agreements with third states, has also been consistently accepted by international courts and tribunals outside the investment context. Thus, in all ICJ cases that gave rise to MFN issues, allegedly better treatment was extended on the basis of an international treaty.Footnote 47 It was for this reason that the Court in the Anglo-Iranian Oil Company case introduced its well-known conceptual distinction between the different treaties involved:
The treaty containing the most-favoured-nation clause is the basic treaty. … It is this treaty which establishes the juridical link between the [beneficiary State] and a third party treaty and confers upon that State the rights enjoyed by the third party. A third-party treaty, independent of and isolated from the basic treaty, cannot produce any legal effect as between the [beneficiary State] and [the granting State]: it is res inter alios actos.Footnote 48
Similarly, in Rights of Nationals of the United States of America in Morocco, the better treatment at issue was allegedly contained in treaties between Morocco (the granting state), on the one hand, and the United Kingdom and Spain (the third states), on the other. Yet, since the third-country treaties meanwhile had come to an end, the better treatment extended under these treaties equally had ceased to exist.Footnote 49 ICJ jurisprudence therefore leaves no question that MFN clauses in general cover benefits extended on the basis of international agreements.
Likewise, MFN clauses in international trade law, including under the WTO, without doubt apply to better treatment extended by the granting state through international agreements.Footnote 50 The same, finally, holds true when looking at the application of MFN clauses by national courts. Here too, it is well-accepted that more favorable treatment can be extended by the granting state on the basis of international treaties.Footnote 51
As these examples show, the standard case for the operation of MFN clauses was the extension of benefits through an international treaty concluded between the state granting MFN treatment and a third state. For this reason, the function of MFN clauses has been designated as a multilateralizing device for bilateral treaty commitments.Footnote 52 To claim now, as Batifort and Heath do, that by including MFN clauses in their traditional IIAs, states “did not envision that these provisions would apply to import standards of treatment but rather that they would forbid actual discrimination through conduct by the host state,”Footnote 53 disregards the extensive state practice to the contrary and the resulting expectation of negotiating states.Footnote 54
The multilateralizing function of MFN clauses, wording permitting, also applies to different levels of substantive investment protection granted under different IIAs of the home state. Most importantly, generally worded MFN clauses in IIAs “multilateralize the bilateral inter-state treaty relationships and harmonize the protection of foreign investments in a specific host state. MFN clauses thus level differences in the standard of protection offered by varying investment treaties.”Footnote 55 They prevent states from shielding those bilateral bargains from multilateralization in relation to the beneficiary state and its investors that fall within the scope of application of an MFN clause. This applies particularly to better substantive standards of treatment granted under third-country IIAs, as differences here would result in discriminations between investors from different home states that are no different from differences in treatment under domestic law. The better “treatment” that MFN clauses in IIAs apply to will thus regularly encompass better “standards of treatment” in the host state's third-country IIAsFootnote 56 and not be limited to benefits extended under domestic law.Footnote 57
All in all, the general international law background governing the application and interpretation of MFN clauses confirms the prevailing practice of investment tribunals to allow reliance on better standards of treatment granted in the host state's third-country IIAs. Batifort and Heath's suggestion that this practice should be reevaluated contradicts the well-established place that the interpretation of MFN clauses has in general international law. Their suggestion therefore should not be followed.
IV. The Lack of Systemic Implications of NAFTA Practice, Rebalanced IIAs, and the Award in İçkale
As support for their reevaluation of MFN clauses, Batifort and Heath do not rely only on a critique of the reasoning of arbitral tribunals when applying MFN clauses in IIAs to better substantive standards of treatment. They also invoke three instances where such an effect of MFN clauses is denied: in İçkale v. Turkmenistan, in the submissions of contracting parties on the MFN clause in NAFTA's investment chapter, and in the drafting of MFN clauses in newer IIAs. For Batifort and Heath, these instances hold a larger systemic relevance: they “discourage the application of broad presumptions as to the nature and effect of ‘the’ MFN clause.”Footnote 58
I disagree: NAFTA practices and MFN clauses in newer IIAs must be seen as treaty-specific opt-outs that confirm, rather than question, the multilateral background of MFN clauses and its interpretative presumptions. İçkale v. Turkmenistan could be treated similarly, but doubts as to the correctness of the Award prevail. None of these instances therefore should prompt the reevaluation of the prevailing practice on applying MFN clauses, their wording permitting, to better substantive standards of treatment extended under the host state's third-country IIAs.
NAFTA Article 1103 and Recalibrated MFN Clauses as Leges Speciales
Express statements by contracting parties indicating that a specific MFN clause should allow investors to rely on better substantive standards of treatment rarely appear. This is different with NAFTA and the MFN clauses in more recent IIAs. Indeed, as Batifort and Heath show in detail, Canada, Mexico, and the United States have put their opposition to applying NAFTA's MFN clause to more favorable standards of treatment in the host state's third-country IIAs on record, either as respondents in investment arbitration or in submissions as non-disputing parties.Footnote 59
One should be careful, however, in building a general theory on MFN clauses based on Article 1103 of NAFTA . First of all, it is an unresolved question whether this provision indeed excludes an investor's reliance on better substantive standards of treatment granted under one of the host state's third-country IIAs. So far, no Chapter 11 tribunal has held accordingly; on the contrary, several tribunals have indicated their openness to applying Article 1103 NAFTA to “import” better substantive standards of treatment.Footnote 60 In the words of the Tribunal in Apotex Holdings v. United States, “whether the NAFTA Parties are correct [regarding their interpretation of Article 1103 of NAFTA] will have to await the decision of another NAFTA tribunal.”Footnote 61
Second, it is questionable whether the concordant submissions of NAFTA parties as respondents or as non-disputing parties have any binding effect in interpreting Article 1103 of NAFTA as “subsequent practice” or as a “subsequent agreement” under the ordinary rules of treaty interpretation enshrined in Article 31(3)(a) and (b) VCLT.Footnote 62 The reason for this is that it is difficult to infer from a state's pleading as respondent the intention to be bound by such a statement in cases beyond the concrete dispute. After all, pleadings as respondents in investor-state arbitration are usually, if not always, made not in order to express an abstract and disinterested position on the correct interpretation of a treaty provision with binding effect for any existing or future dispute, but with the concrete purpose of prevailing in a concrete dispute as party.Footnote 63
The same holds true for many non-disputing party submissions pursuant to Article 1128 of NAFTA, which equally may serve the purpose of winning in a concrete dispute, rather than establishing binding interpretations once and for all. This becomes particularly clear when considering that the United States submitted, as a third-party submission on the interpretation of NAFTA Article 1103 in the Pope & Talbot case, part of its pleadings on exactly that issue as Respondent in the Methanex arbitration, which was pending in parallel.Footnote 64 The motive behind its non-disputing party submission in Pope & Talbot—to prevail in Methanex—appears rather obvious, but may equally cast doubt on the proposition that non-disputing party submissions are intended to be given effect beyond the individual dispute in which they are presented.
Finally, an institutional argument calls for caution in drawing broader inferences from submissions in NAFTA arbitrations. If the parties wanted to enter a binding interpretation of Article 1103 of NAFTA, they could have easily done so, as they have on the interpretation of Article 1105,Footnote 65 by issuing a joint interpretation through NAFTA's Free Trade Commission, a treaty organ established, inter alia, for that purpose.Footnote 66
More importantly, however, and independently of what the correct interpretation of Article 1103 actually is, there are several reasons why NAFTA's MFN clause is not representative of MFN clauses in other IIAs. First, Article 1103 is worded differently from MFN clauses in many other IIAs. It does not apply to more favorable “treatment” tout court, but to “treatment … in like circumstances … with respect to the establishment, acquisition, expansion, management, conduct, operation, and sale or other disposition of investments.” Second, the term “treatment” in Article 1103 may need to be read in context with other NAFTA-specific provisions, such as Article 1101(1), which provides that NAFTA Chapter 11 applies to “measures adopted or maintained by a Party,” and Article 201, which defines a “measure” as including a “law, regulation, procedure, requirement or practice.”Footnote 67 Finally, generalizing from NAFTA is problematic insofar as that agreement is developing into a regime of economic integration distinct from the rest of the investment treaty regime. The special nature of the NAFTA regime is suggested by the emphasis of NAFTA tribunals on NAFTA-precedents in Chapter 11 arbitrationsFootnote 68 as well as the increasingly NAFTA-specific nature of doctrinal analyses and scholarship.Footnote 69
All of these specificities may account for differences between Article 1103 of NAFTA and MFN clauses in other IIAs. Generalizations based on Article 1103 for the interpretation of MFN clauses in other IIAs therefore have to be treated with caution. Instead, NAFTA practices—to the extent they were to diverge from the prevailing approach to MFN clauses in other IIAs—can be viewed as a treaty-specific approach that is entirely consistent with the treaty-by-treaty approach to interpretation under the general international law framework on MFN clauses.
For the same reason, recalibrated MFN clauses in newer IIAs that expressly exclude their application to more favorable standards of treatment, such as Article 8.7 of CETA, cannot be used to question the interpretation of generally worded MFN clauses in older IIAs. Instead, they should be understood as diverging from earlier MFN practice. That divergence is necessary in order for the contracting parties to ensure that the reform efforts they undertake under these new IIAs by, inter alia, granting additional policy space and reducing overly broad investment protection, are not undermined through the operation of the MFN clause in the same, recalibrated IIA.Footnote 70
That MFN clauses in newer IIAs diverge from earlier practice can also not be questioned by pointing to the qualifier that new MFN clauses, such as Article 8.7 of CETA, regularly contain, which clarifies “[f]or greater certainty” that better substantive standards of treatment under third-country IIAs treaties are not covered. This qualifier certainly clarifies the intention of the parties to the treaty in question but cannot be used under a treaty-by-treaty approach to impact the interpretation of MFN clauses in unrelated IIAs. Thus, whereas newer IIAs can be used to interpret older treaties when these contain the usual language reflecting the traditional view of MFN under general international law,Footnote 71 recalibrated MFN provisions in newer IIAs purposely diverge from earlier treaty practice and can therefore not be used to reinterpret commitments to multilateralization under traditional MFN clauses.
Errors in Treaty Interpretation in ?çkale v. Turkmenistan
The other central pillar on which Batifort and Heath rest their claim for a systemic reevaluation of the application of traditional MFN clauses is the Award in İçkale v. Turkmenistan, which declined to interpret the MFN clauses in Article II(2) of the Turkey-Turkmenistan BIT to permit the investor's reliance on more favorable standards of treatment offered under Turkmenistan's third-country IIAs.Footnote 72 The relevance of this Award could—and under Batifort and Heath's BIT-by-BIT approach in fact should—be dismissed as limited to the MFN clause in the Turkey-Turkmenistan BIT.
This is also consistent with the Tribunal's own approach. Its reasoning on the MFN clause is entirely sealed off (1) from how other tribunals, hitherto consistently, have approached the application of MFN clauses to better substantive standards of treatment, (2) from how the parties framed their arguments in reliance on precedent,Footnote 73 and (3) from how other authorities have discussed the interpretation of MFN clauses.Footnote 74 Instead, the Tribunal treats the issue as concerning a singular MFN clause in a singular IIA.
More importantly, however, the attempt to assign systemic importance to İçkale for the interpretation of MFN clauses in IIAs is questionable if one critically analyses that Award and its reasoningFootnote 75 on the interpretation of Article II(2) of the Turkey-Turkmenistan BIT, which provides:
Each Party shall accord to these investments, once established, treatment no less favourable than that accorded in similar situations to investments of its investors or to investors of any third country, whichever is the most favourable.
The Tribunal concluded that “the legal effect of the MFN clause, properly interpreted, is to prohibit discriminatory treatment”Footnote 76 between covered investors and third-country investors, but held that the qualifier “in similar situations” restricted the scope of the clause to prohibiting what the Tribunal termed “de facto discriminations,” that is discriminations not with respect to the scope of “legal rights” of investors covered under different IIAs, but those arising from “a comparison of the factual situation” of different investors.Footnote 77 This, the Tribunal reasoned, was the only way to give effect (effet utile) to the qualifier “in similar situations,” thereby rejecting the claimant's attempt to rely on the better substantive standards of treatment Turkmenistan granted under its IIAs with the United Kingdom, Egypt, and Bahrain.
Yet, in my reading of the reasons given in the Award, the Tribunal committed two errors of interpretation, creating doubt about the correctness of the result reached and invalidating it as a good decision that queries the “conventional wisdom” on the application of MFN clauses to substantive standards of treatment. The first error of the Tribunal consists in a non sequitur when concluding that Article II(2) of the BIT only applies to de facto discriminations. In the Tribunal's view, this followed from the limitation of the prohibition of discrimination to investments “in similar situations.” To reach this conclusion, the Tribunal reasoned as follows:
328. … this obligation [i.e. the prohibition of discrimination contained in Article II(2) of the BIT] exists only insofar as the investments of the investors of the home State and those of the investors of the third State can be said to be in “a similar situation.” Conversely, the MFN treatment obligation does not exist if and when an investment of an investor of the home State is not in a “similar situation” to that of the investments of investors of third States; in such a situation, there is de facto no discrimination.
329. The terms “treatment accorded in similar situations” therefore suggest that the MFN treatment obligation requires a comparison of the factual situation of the investment of the investors of the home State and that of the investment of the investors of third States … .Footnote 78
The non sequitur occurs between the last sentence of paragraph 328 and the first sentence of paragraph 329. The Tribunal first correctly concludes that there is no discrimination when the third-country investors and the investor covered by Article II(2) of the BIT are “not in a ‘similar situation.’” It adds that “in such a situation, there is de facto no discrimination.” Note here the syntactical position and semantic use of the word “de facto.” The Tribunal does not say “there is no de facto discrimination”; it says, “there is de facto no discrimination.” Thus, the Tribunal does not use “de facto” as an adjective to qualify the type of discrimination prohibited under Article II(2) (as in de facto versus de jure discrimination), but as an adverbial construction that emphasizes that no prohibited discrimination had “in fact” occurred in the situation of a third-country investor not in a similar situation.
However, while it is correct to conclude that Article II(2) of the BIT is not violated in a case where de facto no discrimination exists, this does not mean that the treaty provision in question only protects against de facto discriminations. This conclusion is an impermissible non sequitur. It builds not only on an equation of two grammatically fundamentally different constructions (“de facto no discrimination” versus “no de facto discrimination”), it is also logically defective to derive the positive scope of application of a legal norm—here Article II(2) of the BIT—from a situation in which that legal norm has not been breached. Yet, the Tribunal in İçkale does exactly that.
Instead, the Tribunal should have asked, and then provide reasons for or against, whether Article II(2) of the BIT covers only concrete discriminations of foreign investors, that is, requiring, as argued by the respondent, the existence of a concrete third-country investor in a similar situation who received more favorable treatment than the claimant, or whether Article II(2) also applies when the host state accords more favorable treatment in the abstract by concluding IIAs with third countries that include more favorable substantive standards of treatment. The latter is the conclusion suggested by the general international law framework on the interpretation and application of MFN clauses.Footnote 79 Absent any specific circumstances surrounding the Turkey-Turkmenistan BIT, it seems likely that this conclusion should have been adopted by the Tribunal as well. What one would have expected, at a minimum, however, is a reading of Article II(2) of the Turkey-Turkmenistan BIT that does not violate the rules of grammar and the laws of logic.
Apart from this non sequitur, the Tribunal's second error in interpretation occurs with its finding that “the only way” to read the qualifier in Article II(2) of the BIT that a third-country investor needs to be “in a similar situation” in good faith and in a way that gives it effet utile is to exclude the application of the MFN clause to more favorable substantive standards of treatment.Footnote 80 This reading is one possible construction of Article II(2) of the BIT, but it is not the only one, and certainly not the most plausible one, given the general international law background discussed above. A more convincing reading would have been to understand the qualifier (“in similar situations”) to indicate that better treatment could only be relied on if, and only if, the (concrete or hypothetical) third-country investor were to enjoy the benefits under its IIA in question if “in a similar situation” to the investor under the basic treaty. By contrast, if the third-country investor were not to enjoy the benefits in question, even in a similar situation, the MFN clause in question would not extend to the better substantive standards of treatment granted under the third-country IIA.
The qualifier would then require more than asking if the (concrete or hypothetical) third-country investor invested in the same territory as the investor under the basic treaty;Footnote 81 it would require comparisons in other respects, for example concerning the industry sector the investor operates in. The qualifier would then ensure, for example, that Article II(2) of the BIT would not allow an investor in the construction sector, such as the claimant in İçkale, to rely on better substantive treatment granted by Turkmenistan to investors in the energy industry under the Energy Charter Treaty (ECT). ECT benefits could not be relied on by the claimant in İçkale because a (real or hypothetical) investor covered by the ECT would not be “in similar situations” because that investor operates in a different industry sector. Consequently, the qualifier “in similar situations” can be given meaning, and effet utile, without the need to exclude more favorable substantive standards of treatment from the scope of application of Article II(2) of the BIT.
Finally, the Tribunal's construction makes Article II(4) of the BIT superfluous, in violation of the effet utile-principle. This provision, which establishes exceptions from MFN treatment, inter alia, for “agreements entered into by either of the Parties (a) relating to any existing or future customs union, regional economic organization or similar international agreements, (b) relating wholly or mainly to taxation” would not be needed at all, if the qualifier in Article II(2) of the BIT already had the effect of excluding benefits granted under the host state's third-country treaties.
For all of these reasons, the Award in İçkale, and its interpretation of the MFN clause in the Turkey-Turkmenistan BIT, is highly problematic and should not be used to query the effect of MFN clauses in IIAs as commitments to multilateralize benefits granted under IIAs to third-countries and to harmonize investment protection in a given host state. Instead, the general international law background, which supports the application of MFN clauses in IIAs to substantive standards of treatment, should be given preference in applying IIAs, as long as the clauses’ wording and other relevant context so permit.
V. MFN Clauses and Political Ideology
As is apparent from the preceding analysis, I second Batifort and Heath's call for close attention to the wording of MFN clauses in IIAs in order to determine whether they were intended to allow investors to rely on better substantive standards of treatment granted under the host state's third-country IIAs. In fact, a treaty-by-treaty approach to interpreting MFN clauses is mandated under general international law and has been accepted by international courts and tribunals in a variety of fields, including in investment treaty arbitration. Close attention to the wording of MFN clauses is also necessary to appreciate the changes that states are making to their MFN commitments in newer IIAs. In fact, MFN clauses in newer IIAs are often worded to exclude the multilateralization of substantive standards of treatment granted in the host state's third-country IIAs.
The same will not hold true with respect to traditionally worded MFN clauses in older IIAs that simply grant most favored “treatment” to foreign investors or their investments. Under a treaty-by-treaty approach, the interpretation of such clauses not only has to remain unaffected by more recent changes in IIA practice, it also has to respect the rules developed in general international law specifically for the interpretation and application of MFN commitments. These rules have been accepted in the practice of states and international courts and tribunals and have been restated in the ILC's work on MFN clauses. Unless their wording or other relevant context suggest otherwise, MFN clauses, under these rules, are generally understood as ensuring equal treatment among foreign investors independently of whether differences in treatment result from domestic laws or from variations in the substantive standards of treatment granted under different IIAs of the host state.
Under this view, MFN clauses are generally recognized as multilateralizing benefits granted under a host state's third-country IIAs. They ensure equal treatment of the state benefitting from MFN treatment and prevent the granting state from crafting preferential investment rules in their IIAs that are limited to certain third-country investors. In that sense, MFN clauses constitute commitments to multilateralism in the form of bilateral treaties. Together with other structural elements of the investment treaty regime—namely the possibilities for investors to make use of treaty-shopping or nationality-planning, the coordination of IIA-making through multilateral policy processes, and the use of multilateral approaches to treaty interpretation by arbitral tribunals, including through the widespread use of arbitral precedent even across treaties—MFN clauses are essential pieces of the multilateral structures that underlie international investment law.Footnote 82
These multilateral structures do not seek to maximize investment protection, but ensure the application of (relatively) uniform rules for investor-state relations that are independent of the sources and targets of foreign investment flows. Such multilateral structures are the prerequisite for equal competition that allows capital to flow to where it is allocated most efficiently and national economies to specialize in areas where they enjoy a comparative advantage. Furthermore, these structures contribute to international peace and security because they discourage, or even make impossible, any bloc-building behavior by states that would discriminate based on the nationality of investors, as their net effect is to increase economic interdependence worldwide.Footnote 83
Against this background, Batifort and Heath's claim that MFN clauses in IIAs generally should be reevaluated as regards their application to better substantive standards of treatment is not convincing. Their claim is doctrinally flawed not only because they disregard the well-established general international law background for the application and interpretation of MFN clauses. Their claim is also normatively unattractive as it undermines (1) the certainty these rules provide for the interpretation of general MFN commitments in traditional IIAs to encompass better substantive standards of treatment, (2) the resulting harmonization of investment protection rules in a given host state and the value of equal treatment of foreign investors independently of their nationality, and (3) the contribution MFN clauses make toward multilateralism in international investment governance.
What is more, Batifort and Heath's strict treaty-by-treaty approach to interpreting MFN clauses is anything but ideologically neutral. A narrow focus on the text of individual clauses would not, as they suggest, help overcome the existing “deep ideological contestation” and “fre[e] the interpreter from having to rely on presumptions and received wisdom, without necessarily forcing him or her to embrace a competing ideology.”Footnote 84 On the contrary, Batifort and Heath's approach would replace one ideology of, or vision for, international investment governance, namely that of ordering international investment relations on the basis of multilateral considerations and the principle of non-discrimination, with another, namely that of basing investment relations on preferential, quid pro quo bilateral bargains that discriminate on the basis of nationality.
This latter ideology is not only in line with the current trend in U.S. foreign policy toward protectionism, nationalism, and isolationism, which has translated into the withdrawal from multilateral regimes, such as the Paris Agreement or the Trans-Pacific Partnership. It also parallels changes in international economic policy that occurred after the 1929 world economic crisis, when countries like the United Kingdom, the United States, or Germany turned their back on the principle of non-discrimination in international economic relations to endorse preferential bargains that foreshadowed military alliances during World War II.Footnote 85 Just as MFN commitments were incompatible with the newly emerging ideology of the 1930s,Footnote 86 today they do not suit those who elevate short-term domestic political benefits over long-term gains from multilateral cooperation. Yet today, just as in the 1930s, a switch from multilateral to bilateral modes of governing international economic relations could have dramatic effects on economic and geopolitical stability.
If states want to depart from existing multilateral structures in international investment governance, they can do so by recalibrating IIAs, by limiting possibilities of treaty-shopping, and by restricting the scope of application of MFN clauses—or even by withdrawing from IIAs altogether. But such changes should be brought about through treaty-making, as newer developments in IIA-making illustrate, not through the arbitral reinterpretations of traditional IIAs as suggested by Batifort and Heath. With respect to traditional MFN clauses, it is not time to “engage in a serious debate regarding the use of MFN clauses to import standards of treatment from one investment treaty to another.”Footnote 87 Rather, this is precisely the time to strengthen multilateral reform efforts and to defend the multilateral structures underlying the current investment treaty regime against erosion through the reinterpretation of existing commitments to MFN treatment.Footnote 88
