There has been a steady increase in the number of bilateral investment treaties (BITs—treaties between two countries aimed at protecting investments made by investors of both countriesFootnote 1 ) across the world—from 500 in the 1990s to more than 3,324 by the end of 2016.Footnote 2 A BIT allows a foreign investor to directly bring a claim against a host state using the investor-state dispute settlement [ISDS], if the latter has taken a measure that allegedly violates the BIT.Footnote 3 The increasing mass of BITs has led to a significant increase in investor-state disputes in international investment law—from a negligible number in the early 1990s, the total number of known treaty-based ISDS cases has risen to 767 as of 1 January 2017.Footnote 4 Foreign investors have used the ISDS provisions in BITs to challenge a wide range of state regulatory measures such as environmental measures,Footnote 5 and monetaryFootnote 6 and taxationFootnote 7 measures, as well as public health protection measures.Footnote 8
Adjudication of disputes on potential breaches of BITs covering a wide breadth of sovereign regulatory measures by ISDS tribunals has triggered a debate on whether BITs are encroaching upon a host state’s right to regulate.Footnote 9 It has also triggered a backlash against BITs and ISDS, as evident from some states walking out of the system by denouncing the International Centre for the Settlement of Investment Disputes [ICSID] ConventionFootnote 10 (that provides for the ISDS mechanism).Footnote 11 Additionally, some countries have terminated their BITs.Footnote 12
Proposals which aim to amend the existing ISDS system by either making it more transparentFootnote 13 or by bringing about other kinds of reforms, such as introducing an appellate mechanismFootnote 14 or even developing a world investment court system, have also been advanced.Footnote 15 It has also been persuasively argued that the frictions between the international investment law regime with competing public interests can be attenuated using various tools such as interpretative techniques and the recalibration of investment treaties.Footnote 16 Indeed, some countries have started recalibrating their investment treaties to reconcile investment protection with a host state’s right to regulate, thus avoiding the extreme step of denouncing BITs.Footnote 17
Against this backdrop, the recent decision of an ISDS tribunal in Philip Morris Brands Sàrl v. Oriental Republic of Uruguay Footnote 18 (Philip Morris v. Uruguay) is of considerable importance.Footnote 19 This case, involving the competing interests of Philip Morris, a giant tobacco company, and the public health interests of Uruguay, came to be seen as a litmus test to determine whether BITs and ISDS unduly encroach upon a host state’s right to regulate. Philip Morris challenged Uruguay’s two public health measures aimed at restricting the marketing of tobacco products.Footnote 20 The two public health measures are: (1) the “Single Presentation Requirement” [SPR], which requires that cigarette brands sell only under a single package or variant, and (2) the “80/80 Regulation”, which mandates that the health warnings on cigarette packages increase from fifty percent to eighty percent of the surface of the packages, leaving only twenty percent of the space on the packages for display of trademarks, logos, and other information.Footnote 21 Philip Morris challenged these regulations as breaching Uruguay’s obligations under the Switzerland-Uruguay BIT.Footnote 22 Specifically, Philip Morris argued that the two measures adopted by Uruguay violated the following Articles of the Switzerland-Uruguay BIT: Article 3(1) (impairment of use and enjoyment of investments), Article 3(2) (fair and equitable treatment [FET] and denial of justice), Article 5 (expropriation), and Article 11 (observance of commitments entered into with respect to investments of investors).Footnote 23 The tribunal, by a majority, upheld the legality of the two Uruguayan regulatory measuresFootnote 24 and dismissed Philip Morris’s claims.Footnote 25 The tribunal found that the two Uruguayan regulatory measures neither violated the FET provision nor the provision on expropriation in Article 5 of the Switzerland-Uruguay BIT.
The outcome of this case is extremely significant, not just for Uruguay, but also for the international investment law community at large. The tribunal’s decision of upholding Uruguay’s right to adopt measures for the protection of public health as consistent with the BIT could be seen as a measure to shore up the eroding confidence of countries in BITs and ISDS. Defenders of the regime would use this case to argue that the fear of BITs and ISDS unduly encroaching upon a host state’s regulatory autonomy are exaggerated. Arguably, if the tribunal had ruled in favour of Philip Morris, it might have given political impetus to many states to withdraw from investment treaties. However, the fact that one of the arbitrators dissented in part and held that the SPR violated the FET and denial of justice provision in the Switzerland-Uruguay BITFootnote 26 arguably reveals the continuing uncertainties in defending sovereign health-related regulatory measures before ISDS tribunals.Footnote 27
I. PURPOSE AND SCOPE OF THE PAPER
Generally speaking, the acceptability of the outcome of an ISDS case is considered more important than the rigour of the tribunal’s reasoning.Footnote 28 This paper, while recognizing that the outcome of the Philip Morris award is important, focuses critically on the reasoning adopted by the tribunal. The purpose is to examine the clarity and rigour in the tribunal’s reasoning because the quality of arbitral reasoning, not just the outcome, plays an important role in building the legitimacy of the ISDS system.Footnote 29 ISDS awards that are not judiciously and thoroughly reasoned lead to incoherent case-law,Footnote 30 and augment the anxiety and resentment of states towards international arbitration.Footnote 31 Such awards dampen the effectiveness of the system because both states and foreign investors are not able to mould their conduct in accordance with the standards articulated by the tribunals.Footnote 32 Ortino has identified three types of egregious failures in the reasoning of the ISDS tribunals: misuse of precedent, lack of internal consistency, and minimalism.Footnote 33
Given the significance of clarity and rigour in legal reasoning, this paper critically examines the reasoning of the Philip Morris tribunal in dealing with the question of whether Uruguay’s regulatory measures resulted in the expropriation of Philip Morris’s investment. Specifically, the paper discusses how the tribunal used Article 31(3)(c) of the Vienna Convention on the Law of Treaties [VCLT] to invoke the police powers rule in interpreting the expropriation provision of the Switzerland-Uruguay BIT under which the dispute arose. Given the paucity of space, the paper does not discuss other important aspects of the award, such as the interpretation of the FET provision.
There are two important reasons to focus on the tribunal’s reasoning and interpretation of the expropriation provision. First, one of the most common grounds for challenging a host state’s regulatory measures is that the measures breach the expropriation provision in the BIT.Footnote 34 As direct expropriations—state actions which deprive investors of legal titleFootnote 35 —have become rare,Footnote 36 the focus has shifted to determining what constitutes indirect expropriation—deprivation of the substantial benefits flowing from the investment without any formal “taking” of the property.Footnote 37 Determining what constitutes indirect expropriation is not easy. ISDS tribunals have developed different tests to do so, such as the sole effects test where the focus is only on the severity of the effect of the regulatory measure on foreign investment.Footnote 38 According to this test, measures that do not constitute direct expropriation may nevertheless constitute indirect expropriation if the effect of the regulatory measure causes a substantial deprivation of foreign investment.Footnote 39 In addition, some tribunals have developed what is described as the police powers test, borrowed from customary international law [CIL],Footnote 40 where state measures that are prima facie lawful exercises of the government’s powers (such as adopting a measure pursuing a legitimate public welfare objective) may affect foreign interests considerably without amounting to expropriation.Footnote 41 Apart from these two tests, there is a third test: proportionality analysis, which requires balancing the public purpose behind the regulatory measure with the effect that the measure has on foreign investment.Footnote 42
The Philip Morris tribunal focused on both the sole effect (or substantial deprivation) test and the police powers test to determine whether foreign investment had been indirectly expropriated. As the paper will discuss, the use of both tests muddies the water and does not bring much clarity to the question of what constitutes indirect expropriation. Also, the use of the police powers test raises critical questions regarding the definition of police powers and its application, which has implications far beyond the current case.Footnote 43
The second important reason is the tribunal’s use of Article 31(3)(c) of the VCLT to invoke the police powers rule in interpreting the expropriation provision of the BIT. Article 31(3)(c) of the VCLT provides: “There shall be taken into account, together with the context: (c) any relevant rules of international law applicable in the relations between the parties.”Footnote 44 As Simma and Kill argue, tribunals routinely refer to international law rules that derive their normative legitimacy from sources outside the treaty which is the subject matter of interpretation.Footnote 45 Article 31(3)(c), which has attracted considerable attention,Footnote 46 reflects the proposition that no treaty originates outside the international legal system and that this system will continue to be relevant for the purposes of interpreting an international treaty.Footnote 47 Article 31(3)(c) has been widely hailed as an instrument for bringing about the systemic integration of international law.Footnote 48 ISDS arbitral tribunals have started referring to Article 31(3)(c) of the VCLT as a tool for interpreting the investment treaty in question.Footnote 49 For the international investment law community, an important question is how well the tribunals have reasoned the use of Article 31(3)(c) when referring to treaty or customary norms outside the investment treaty. This question is extremely pertinent because the ISDS system’s engagement with external legal norms could have implications for the legitimacy of the system.Footnote 50
It is important to bear in mind that Article 31(3)(c) of the VCLT is not the only gateway for bringing extraneous rules, like a customary rule of police powers, into the interpretation of investment treaties. If a treaty term has a meaning recognized in customary law, the customary law could arguably be incorporated as “ordinary” meaning under Article 31(1), or as “special” meaning under Article 31(4) of the VCLT.Footnote 51 However, since the Philip Morris tribunal focused only on Article 31(3)(c) to deal with extraneous rules, this paper restricts its analysis to Article 31(3)(c). While the focus of this paper is the issues of Article 31(3)(c) and the police powers rule through the prism of the Philip Morris case, the issues it discusses have wider implications, and are thus of interest to the international investment law community as a whole.
The Philip Morris tribunal’s reasoning on indirect expropriation is at two levels: first, whether the regulatory measures, i.e. the SPR and 80/80 Regulation, deprived Abal (Philip Morris Brand Sarl owned 100 percent of Abal—an entity constituted under the laws of UruguayFootnote 52 ) of the value of its business or caused a “substantial deprivation” of the value, use, or enjoyment of its investment; and second, whether the challenged measures were a valid exercise of Uruguay’s police powers, which would defeat the expropriation claim under Article 5(1). Therefore, Section II of this paper deals with the tribunal’s application of the substantial deprivation test. In Section III, the paper deals with the tribunal’s application of the police powers test using Article 31(3)(c) of the VCLT. Section IV concludes by discussing what the Philip Morris tribunal should have done and by raising larger conceptual issues on the relationship between the police powers rule and determination of indirect expropriation.
II. THE TEST OF SUBSTANTIAL DEPRIVATION TO DETERMINE WHAT CONSTITUTES INDIRECT EXPROPRIATION
In this section, we discuss the tribunal’s focus on the substantial deprivation test to determine whether expropriation had taken place. The tribunal’s task was to interpret Article 5(1) of the Switzerland-Uruguay BIT, which states:
Neither of the Contracting Parties shall take, either directly or indirectly, measures of expropriation, nationalization or any other measure having the same nature or the same effect against investments belonging to investors of the other Contracting Party, unless the measures are taken for the public benefit as established by law, on a non-discriminatory basis, and under due process of law, and provided that provisions be made for effective and adequate compensation. The amount of compensation, interest included, shall be settled in the currency of the country of origin of the investment and paid without delay to the person entitled thereto.
Thus, Article 5(1) of the BIT clearly prohibits countries from adopting expropriatory measures, either directly or indirectly, unless the measures are taken for public benefit on a non-discriminatory basis, following due process, and provisions are made for effective and adequate compensation. Since this case involved determining what constitutes indirect expropriation, the important words in Article 5(1) are “any other measure having the same nature or the same effect against investments”.
The tribunal began its analysis in paragraph 191 by recognizing that the claim relates to indirect expropriation. The question that the tribunal asked itself was to determine the threshold for finding indirect expropriation.Footnote 53 The tribunal started answering this question in paragraph 192 at two levels. First, to determine what constitutes indirect expropriation, the tribunal recognized the central role of the “effect” of the regulatory measure on investment. In other words, the tribunal recognized the “sole effects” doctrineFootnote 54 to determine what constitutes indirect expropriation, whereby the crucial factor in determining whether an indirect expropriation has occurred is solely the effect of the governmental measure on the property. Thus, the purpose behind the regulatory measure is irrelevant under the sole effects test. The justification which many ISDS tribunals have offered for relying on the sole effects test is textual, i.e. the expropriation provision in the BIT contains only the word “effect”, as is the case with Article 5(1) of the Switzerland-Uruguay BIT. For example, in AWG v. Argentina,Footnote 55 where the expropriation provision in the UK-Argentina BIT contains the phrase “subjected to measures having effect equivalent to nationalization or expropriation”, the tribunal held that specific reference to “effects” in the BIT “affirms the importance of evaluating ‘effects’ of the measure on the investment in determining whether an expropriation has taken place”.Footnote 56
Second, in paragraph 192, the tribunal recognized that the focus on the “effect” of the regulatory measure on foreign investment to determine what constitutes indirect expropriation fails to answer the key question: i.e. how severe should the effect be? The tribunal stated that “in order to be considered an indirect expropriation, the government’s measures interference with the investor’s rights must have a major adverse impact on the claimant’s investments”.Footnote 57 The tribunal further clarified that, for regulatory measures to have a major adverse impact, the regulatory measures should lead to “substantial deprivation” of the value, use, or enjoyment of the claimant’s investment.Footnote 58 The effect can certainly be more than substantial deprivation, such as cases where the deprivation is complete or total. For example, the tribunal in Total SA v. Argentina Footnote 59 held that, under international law, those measures that do not constitute direct expropriation may nevertheless result in indirect expropriation “if an effective deprivation of the investment is thereby caused”.Footnote 60 Furthermore, the Philip Morris tribunal said that substantial deprivation would be determined by taking into account the “intensity” and “duration” of the economic deprivation suffered by the investor.Footnote 61
On the basis of these principles, and having examined the claims of the investor, the tribunal held that the regulatory measures adopted by Uruguay did not result in substantial deprivation of Abal’s investment.Footnote 62 The tribunal held that, since sufficient value of the investment remained after the implementation of the challenged regulatory measures, the impugned measures did not amount to indirect expropriation.Footnote 63 The tribunal found that, despite the adoption of the regulatory measures by Uruguay, Abal’s profits had increased, though the profits would have increased to a greater extent if the impugned regulations had not been adopted.Footnote 64 Consequently, the tribunal concluded that there was no indirect expropriation of Abal’s investment.
Up to this point, it is difficult to find fault with the reasoning of the tribunal. The tribunal carefully interpreted Article 5(1) and correctly applied the law to the facts at hand—an adverse economic effect on foreign investment such as deteriorating profits, or the incurring of losses, short of total or at least substantial deprivation, shall not amount to indirect expropriation.Footnote 65 This interpretation gives ample space to host states to exercise their regulatory powers without worrying about foreign investors challenging such actions as expropriations, unless the high threshold of substantial deprivation is breached.
III. ARTICLE 31(3)(C) OF VCLT AND POLICE POWERS
The finding that Uruguay did not indirectly expropriate foreign investment obviated the need to deal with the issue of expropriation any further. The tribunal itself tacitly accepted this in paragraph 287: “the tribunal’s analysis [of the indirect expropriation claim] might end here”.Footnote 66 However, the tribunal did not stop there, and went on to offer an “additional” reason to support its conclusion. The tribunal held that the SPR and 80/80 Regulation adopted by Uruguay was a valid exercise of its police powers, “with the consequence of defeating the claim for expropriation under Article 5(1) of the BIT”.Footnote 67 The tribunal held that for this reason “also”, the claim regarding indirect expropriation must be rejected.Footnote 68 The usage of words like “additional” and “also” is curious. The use of these words suggests that, according to the tribunal, the defence of police powers always operates as an extra argument or reason to defeat the claim of indirect expropriation once it is established that the regulatory measures did not result in a “substantial deprivation” of foreign investment. But, what if the initial analysis had sufficed to show that the challenged regulatory measures led to a “substantial deprivation” of Abal’s investment? If the police powers argument is an “additional” reason to defeat expropriation claims, it would be of no significance if regulatory measures did result in substantial deprivation of foreign investment.
This approach adopted by the Philip Morris tribunal somewhat mirrors the approach adopted in Chemtura v. Canada.Footnote 69 The Chemtura tribunal, while examining whether the regulatory measure amounted to indirect expropriation, first considered whether the regulatory measures resulted in substantial deprivation of investment.Footnote 70 Once the tribunal came to the conclusion that there was no “substantial deprivation”,Footnote 71 it was said that “irrespective of contractual deprivations”, the measures challenged were part of the state’s police powers, and thus did not constitute expropriation.Footnote 72
The tribunal in AWG v. Argentina also followed a similar route. As pointed out earlier, the tribunal first affirmed the importance of determining the “effect” of the regulatory measures on foreign investment and whether the regulatory measures had resulted in a “substantial deprivation” of foreign investment.Footnote 73 It then held that, in evaluating a claim of expropriation, “it is important to recognize a State’s legitimate right to regulate and to exercise its police power in the interests of public welfare and not to confuse measures of that nature with expropriation”.Footnote 74 Finally, the tribunal concluded that the measures adopted by Argentina were within the general police powers of the Argentine state, and that they did not constitute a permanent and substantial deprivation of foreign investment, and thus did not amount to indirect expropriation.Footnote 75 The conclusion reached by the AWG tribunal makes one wonder what the outcome would have been if the regulatory measures had led to a permanent and substantial deprivation of foreign investment. Would it have resulted in a finding of expropriation or would the application of the police powers rule mean a finding of no expropriation despite substantial deprivation? Be that as it may, we now turn to critically examine the Philip Morris tribunal’s reasoning on the application of the police powers rule using Article 31(3)(c) of the VCLT.
A. How did the Tribunal Use Article 31(3)(c)?
The tribunal reasoned that Article 5(1) of the Switzerland-Uruguay BIT should be interpreted in accordance with Article 31(3)(c) of the VCLT. The tribunal used Article 31(3)(c) to incorporate the police powers rule in CIL to interpret Article 5(1) of the BIT.
The critical question is whether the tribunal used Article 31(3)(c) correctly to interpret Article 5(1) of the Switzerland-Uruguay BIT. Before we look at how the tribunal used Article 31(3)(c) of the VCLT, it will be useful to first have a quick look at the legal architecture of Article 31(3)(c). As indicated before, according to the wording of Article 31(3)(c), the treaty interpreter is under an obligation to take into account any relevant rules of international law that are applicable in the relations between the parties. Analytically speaking, the interpretative framework that Article 31(3)(c) provides to the treaty interpreter comprises the following:Footnote 76 first, one has to determine whether there is a “rule of international law”;Footnote 77 second, whether such a rule is “applicable in the relations between the parties”;Footnote 78 third, this applicable rule should also be “relevant”;Footnote 79 and fourth, if a rule satisfies the three conditions mentioned above, it is admissible in the process of interpretation,Footnote 80 though one still has to determine the weight that should be accorded to this admissible rule in the interpretation of the treaty norm.Footnote 81
In paragraph 290, the tribunal referred to Article 31(3)(c) and held that the provision mandates the tribunal to refer to rules of CIL.Footnote 82 Hence, the tribunal referred to the police powers rule, which is part of CIL.Footnote 83 The tribunal also held that protecting public health has “long been recognized as an essential manifestation of the State’s police power”.Footnote 84 The tribunal, like some tribunals in the past,Footnote 85 located the authority for the police powers rule in the 1961 Harvard Draft Convention on the International Responsibility of States for Injuries to Aliens,Footnote 86 specifically in Article 10(5), and in the Third Restatement of the Foreign Relations Law of the United States (US) 1987.Footnote 87
The tribunal, drawing support from the Organization for Economic Co-operation and Development’s [OECD] paper on indirect expropriation,Footnote 88 laid down the police powers rule in CIL as follows: “State’s reasonable bona fide exercise of police powers in such matters as the maintenance of public order, health or morality, excludes compensation even when it causes economic damage to an investor and that measures taken for that purpose should not be considered expropriatory.”Footnote 89 The tribunal further justified the police powers rule by citing past ISDS tribunals that made mention of it,Footnote 90 and from the fact that the rule “has found confirmation in recent trade and investment treaties” such as the 2012 US Model BITFootnote 91 and the EU-Canada Comprehensive Economic and Trade Agreement [CETA].Footnote 92 On this basis, the tribunal concluded that, irrespective of whether the police powers rule is introduced in the treaty, it reflects the position under general international law,Footnote 93 i.e. it is significant in interpreting Article 5(1).
After reaching this conclusion, the tribunal devoted the next three paragraphs to examining whether the SPR and the 80/80 Regulation were adopted by Uruguay in fulfilment of its national and international legal obligations.Footnote 94 The tribunal referred to Uruguay’s domestic law provisions that imposed an obligation on Uruguay to adopt measures for the fulfilment of public health objectives.Footnote 95 The tribunal also referred to international legal instruments like the Framework Convention on Tobacco Control [FCTC] and held that, since the SPR and 80/80 Regulation specifically concerned regulating the use of tobacco, they were in accordance with the FCTC.Footnote 96
In the next paragraph, the tribunal returned to the police powers doctrine, though this time it was stated differently from that in paragraph 295. The tribunal stated that a state’s exercise of regulatory powers does not constitute indirect expropriation if such action is bona fide for the purpose of protecting public welfare, is non-discriminatory, and is proportionate.Footnote 97 The tribunal concluded that Uruguay’s regulatory measures were proportionate to the objective “they meant to achieve, quite apart from their limited adverse impact on Abal’s business”.Footnote 98 In footnote 405, the tribunal explained that Uruguay’s regulatory measures had a limited impact on Abal’s investment because it merely limited the use of Abal’s trademark, unlike other cases where public health regulatory measures resulted in the banning of the production and sale of the subject matter.Footnote 99 On this basis, the tribunal thereby concluded that the challenged measures were a valid exercise of Uruguay’s police powers for the protection of public health, and thus did not constitute an expropriation of the claimant’s investment.Footnote 100
Now let us evaluate the tribunal’s analysis on each of the four components of Article 31(3)(c) mentioned before.
B. Did the Tribunal Precisely Identify the “Rule”?
“[R]ules of international law” in Article 31(3)(c) refers to rules that can be derived from CIL, treaties, and the general principles of law. Inherent in this derivation is the fact that the rule(s) should be precisely identified. The rule in question here is the police powers rule. It is one thing to say that the police powers rule is part of CIL, quite another to precisely lay down the rule. The Philip Morris tribunal dealt with the police powers rule in two instances. First, in paragraph 295, the tribunal laid down the police powers rule in CIL as follows: reasonable bona fide regulatory measures adopted by a state for the purpose of public welfare such as public health does not constitute expropriation, and thus absolves the state from paying compensation to the foreign investor, notwithstanding the economic damage caused to the investor (Rule 1). In other words, as Newcombe and Paradell have argued, according to Rule 1, a legitimate and bona fide exercise of the state’s police powers does not amount to indirect expropriation, subject inter alia to an analysis of reasonableness.Footnote 101 Second, in paragraphs 305 and 306, the tribunal introduced the notion of “proportionality” and defined the police powers rule as follows: a bona fide, non-discriminatory regulatory measure adopted for public welfare does not amount to indirect expropriation, provided the regulatory measure is proportionate (Rule 2).
Is Rule 1 the same as Rule 2? If the answer is “yes”, one can say that the Philip Morris tribunal identified the police powers rule precisely. If the answer is “no”, the tribunal did not lay down the police powers rule precisely. Let us examine this.
The core distinction between Rule 1 and Rule 2 is that, while the former talks of the test of reasonableness, the latter talks of the test of proportionality. Is the test of reasonableness different from test of proportionality? While a full analysis of the tests of reasonablenessFootnote 102 and proportionalityFootnote 103 is beyond the scope of this paper, it will be useful to briefly discuss how ISDS tribunals have dealt with these terms. While some ISDS tribunals refer to reasonableness without defining the nature of their review,Footnote 104 some say that a measure will be reasonable if there is an “appropriate correlation between the State’s public policy objective and the measure adopted to achieve it”.Footnote 105 Some even go one step further and say that, in addition to an appropriate correlation existing between the public policy objective and the measure adopted, the measure’s impact on the investor should be proportionate to the policy objective sought.Footnote 106 In other words, the meaning of the word “reasonable” could vary from a measure making a (significant) contribution to the public policy objective to a more strict interpretation involving weighing and balancing the impact of the measure on foreign investment with the public policy it seeks to achieve.Footnote 107 The strict interpretation of “reasonable” has the same meaning as “proportionality”, which consists of three stepsFootnote 108 that must be assessed cumulatively.Footnote 109 First, whether the measure is suitable for the legitimate public purpose—this requires a causal link between the measure and its object.Footnote 110 If the first step is fulfilled, the second step asks whether the measure is necessary, i.e. whether there is a less restrictive alternative measure that will achieve the same objective.Footnote 111 If the measure is found to be “necessary”, the third step (also known as proportionality stricto sensu) will involve balancing the effects of the measure on the right that has been affected with the public benefit the measure seeks to achieve.Footnote 112
The Philips Morris tribunal did not explain the meaning of “reasonable” under Rule 1 and “proportional” under Rule 2. If the tribunal intended the stricter interpretation of “reasonable” as laid down in Eletrabel v. Hungary, one can conclude that Rule 1 and Rule 2 laid down the same police powers rule. However, since the tribunal did not give much indication about this, it is plausible to conclude that the words “reasonable” and “proportional” do not have the same meaning, i.e. “reasonable” under Rule 1 has a broader meaning than “proportional”. In fact, the tribunal commented that the impugned regulatory measures were directed at and capable of achieving the regulatory objective of reducing instances of smoking,Footnote 113 which seems to point to the broader meaning of “reasonable”. If we accept the broader meaning of “reasonable”, Rule 1 can be elaborated as follows: bona fide and non-discriminatory measures adopted by a state for the purpose of public welfare, such as public health, does not constitute expropriation if the measure has a rational relationship with the public policy goal it seeks to achieve, and thus absolves the state from paying compensation to the foreign investor, notwithstanding the economic damage caused to investment. Consequently, Rule 2 can also be elaborated as follows: a bona fide, non-discriminatory regulatory measure adopted for public welfare does not amount to indirect expropriation, provided the regulatory measure is proportionate, i.e. the measure is suitable for the regulatory goal, is necessary, and balances the effects with the benefits that it seeks to achieve. When the notion of proportionality is brought in, it means there is an admission of the fact that bona fide non-discriminatory regulatory measures adopted for public welfare (such as public health) could amount to expropriation if the effect of the regulatory measure on foreign investment is disproportionate to the benefits of the measures.
Laying down two different conceptions of the police powers rule and the failure to identify the police powers rule precisely show a lack of internal consistency in the tribunal’s reasoning. The fact that the Philip Morris tribunal failed to identify the police powers rule preciselyFootnote 114 is buttressed when we undertake a close study of the different ISDS tribunals that it cited to support the police powers doctrine.Footnote 115 The tribunal cited Methanex v. USA, Saluka v. Czech Republic, Tecmed v. Mexico, and Chemtura v. Canada. The choice of cases cited by the tribunal is interesting. It gives an impression that all these cases laid down a common police powers rule. While it is true that all these tribunals recognized the police powers doctrine in claims of indirect expropriation, it is misleading to present them as if there were no differences in the manner in which these tribunals laid down the police powers rule. These differences, if ignored, could result in conceptual errors. Let us briefly discuss the police powers rule laid down by each tribunal. Since these cases have already been discussed in detail in the literature, we will focus only on the parts relevant to laying down the police powers rule.
1. Methanex v. USA (or the Methanex rule)
Methanex v. USA, the oft-cited case whenever any tribunal talks of the police powers rule, laid down the rule as follows:
[A]s a matter of general international law, a non-discriminatory regulation for a public purpose, which is enacted in accordance with due process and, which affects, inter alios, a foreign investor or investment is not deemed expropriatory and compensable unless specific commitments had been given by the regulating government to the then putative foreign investor contemplating investment that the government would refrain from such regulation.Footnote 116
According to the Methanex tribunal, the primary test for determining whether a measure amounts to expropriation or lawful, non-compensable regulations depends on whether it is taken for a public purpose in a non-discriminatory manner, through a law enacted with due process. There can only be indirect expropriation if the state reneges on specific commitments given to the investor that it would refrain from adopting a regulatory measure. There is neither a mention of regulatory measures being “reasonable” or “proportional”. Thus, it is different from both Rule 1 and Rule 2 laid down by the Philip Morris tribunal. There are other noticeable differences. First, Rule 1 specifically includes “economic damage” in the police powers rule, which is missing in the Methanex rule. Second, Rule 1 does not include “specific commitments” as part of the police powers rule, which is present in the Methanex rule. Another important issue not specifically captured in the rule laid down by the Methanex tribunal is the role of the quantum of economic harm caused to foreign investment as a factor to determine indirect expropriation.Footnote 117 Thus, the critical question is: assuming that a regulatory measure satisfies all the above conditions but results in the total or substantial deprivation of foreign investment, will it still be non-compensable? The Saluka and the Chemtura cases discussed in the following sections dealt with this issue.
2. Saluka v. Czech Republic (or the Saluka rule)
In this case, the tribunal had to interpret the expropriation provision in Article 5 of the Czech Republic-Netherlands BIT (1991) titled “deprivation”. The tribunal, citing Article 31(3)(c), held that the term “deprivation” given in Article 5 imports into the BIT the CIL notion that the exercise of regulatory actions aimed at maintaining public order justify deprivation.Footnote 118 The tribunal went on to lay down the police powers rule as follows: It is now established in international law that states are not liable to pay compensation to a foreign investor when, in the normal exercise of their regulatory powers, they adopt bona fide regulations that are aimed at general welfare in a non-discriminatory manner. The tribunal then held that regulatory measures falling under the police powers doctrine do not constitute expropriation, notwithstanding the fact that “the measure had the effect of eviscerating” foreign investment.Footnote 119 If evisceration is to be understood as “substantial deprivation” of foreign investment, the rule can be expressed as follows: states are not liable to pay compensation to a foreign investor when, in the normal exercise of their regulatory powers, they adopt non-discriminatory, bona fide regulations aimed at general welfare, notwithstanding the quantum of harm such measures might cause to foreign investment.Footnote 120 Although the tribunal did mention that the police powers exception is not absolute,Footnote 121 it defined the boundary of the police powers exception only in terms of the measure not being discriminatory and being bona fide (i.e. adopted for a rational public objective). There is no mention, at least not clearly, of the measure being either reasonable or proportionate. Thus, the Saluka rule differs from both the rules laid down by the Philip Morris tribunal. The Saluka rule also does not talk of “specific commitments” mentioned in the Methanex rule, and thus differs from the Methanex rule as well. The Saluka rule can be articulated as follows: a bona fide, non-discriminatory measure adopted for public welfare objective is not expropriatory, notwithstanding the economic impact on foreign investment.Footnote 122
3. Chemtura v. Canada (or the Chemtura rule)
In Chemtura v. Canada, Footnote 123 the ban imposed by the Canadian Pesticide Management Regulation Agency [PMRA] on “lindane”, a pesticide used in canola farming and considered to have an adverse effect on human health, was challenged by the claimant, Chemtura, a US company manufacturing “lindane”, as amounting to expropriation under Article 1110 of the North American Free Trade Agreement [NAFTA]. The tribunal laid down the police powers rule as follows:
Irrespective of the existence of a contractual deprivation, the Tribunal considers in any event that the measures challenged by the Claimant constituted a valid exercise of the Respondent’s police powers … [The measures were adopted] in a non-discriminatory manner, motivated by the increasing awareness of the dangers presented by lindane for human health and the environment. A measure adopted under such circumstances is a valid exercise of the State’s police powers and, as a result, does not constitute an expropriation.Footnote 124
The Chemtura rule comes close to Rule 1 laid down by the Philip Morris tribunal. It talks of deprivation, which presumably is the same as economic damage mentioned in Rule 1. It also imposes a requirement that measures be non-discriminatory, as is the case with measures in Rule 1. Furthermore, stating that the measures adopted were motivated by increasing health risks is tantamount to stating that the measures were reasonable (made a contribution to the public policy objective) as given in Rule 1, but not in the Saluka rule. However, the Chemtura rule is not the same as Rule 2, laid down by the Philip Morris tribunal, because it does not talk of measures being proportional.
4. Tecmed v. Mexico (or the Tecmed rule)
Another case cited by the Philip Morris tribunal in support of the police powers rule is Tecmed v. Mexico. The tribunal cited paragraph 119Footnote 125 of Tecmed v. Mexico: “The principle that the State’s exercise of its sovereign power within the framework of its police powers may cause economic damage to those subject to its powers as administrator without entitling them to any compensation whatsoever is undisputable.”Footnote 126 However, the tribunal’s analysis did not end here, and thus by merely quoting the above statement, the Philip Morris tribunal could be accused of selective reading of Tecmed v. Mexico. The tribunal in Tecmed v. Mexico went on to state that
we find no principle stating that regulatory administrative actions are per se excluded from the scope of the Agreement [Spain-Mexico BITFootnote 127 ], even if they are beneficial to the society as a whole—such as environmental protection—particularly if the negative economic impact of such actions on the financial position of the investor is sufficient to neutralize in full the value, or economic or commercial use of its investment without receiving any compensation whatsoever.Footnote 128
In other words, there is a clear recognition that regulatory measures that fall within the police powers of the state may nevertheless be expropriatory if the impact of such measures on foreign investment effectively neutralizes the value of investment. Finally, the tribunal, by taking into account both the purpose behind the regulatory measure and the effect on foreign investment, stated that in order to determine if a regulatory measure is expropriatory or not, one needs to find out “whether such actions or measures are proportional to the public interest presumably protected thereby and to the protection legally granted to investments, taking into account that the significance of such impact has a key role upon deciding the proportionality”.Footnote 129 Thus, according to the tribunal, a regulatory measure shall not be expropriatory if a “reasonable relationship of proportionality” could be established between the charge or weight imposed by the measure on foreign investment and the aim that the impugned measure seeks to achieve.Footnote 130
This articulation of the police powers rule, requiring a reasonable relationship of proportionality, comes close to Rule 2 given by the Philip Morris tribunal, but is very different from the Methanex, Saluka, and Chemtura rules. In sum, one can safely conclude that the Philip Morris tribunal did not identify the police powers rule precisely. The four cases it cited in support of the police powers rule, while having common elements, differ quite a bit from each other, and thus amounted to a clear misuse of precedent,Footnote 131 one of the reasons for inconsistent and badly reasoned ISDS awards.
C. Is the Rule Applicable in the Relations Between the Parties?
The rule of international law identified should be “applicable in the relations between the parties”,Footnote 132 unless one of the states has persistently objected to the rule.Footnote 133 Whether a rule of international law is “applicable” between the parties to the treaty is composed of many facets, such as how to address the issue of inter-temporal law in question (i.e. whether the applicability is to be restricted to those rules that were in force when the treaty in question was concludedFootnote 134 or those that have evolved up to the time of interpretation of the treatyFootnote 135 ), and whether “applicable” is restricted to rules that are “binding” or “in force” between the parties or whether the term “applicable” should be interpreted more flexibly such that it goes beyond the strict meaning of “in force” or “binding”.Footnote 136
Answering these questions is beyond the scope of this paper, and has in fact been dealt with elsewhere.Footnote 137 It is, however, sufficient to say that since the police powers rule, notwithstanding the lack of consistency in defining it, owes its origins to CIL, it is binding on all countries, and therefore is clearly “applicable” “in relations between the parties” to the treaty as mandated by Article 31(3)(c).
Another doctrinal question is what is the meaning of “parties” in Article 31(3)(c)—does it refer to parties to the dispute (i.e. rule is only applicable on parties to the dispute) or some/all parties to the treaty being interpreted (i.e. rule is applicable on some/all countries that are parties to the treaty)? This issue has been adequately debated, and is thus not dealt with in this paper.Footnote 138 It is sufficient to say that since the Switzerland-Uruguay BIT is a bilateral treaty, the issue is irrelevant in the current case.
The Philip Morris tribunal did not discuss the issue of the applicability of the police powers rule between Uruguay and Switzerland. It presumed that the rule was applicable. However, once a tribunal invokes Article 31(3)(c), it is incumbent on the tribunal to deal with all of the major elements of the said provision to ensure that the method of treaty interpretation is robust.
D. Is the Rule “Relevant”?
Is the police powers rule “relevant” under Article 31(3)(c)? The question of relevance of a “rule of international law” can be approached in two ways.Footnote 139 The first approach is a narrow one, where the determination of whether a rule is “relevant” is made by “examining criteria such as the subject of the dispute and the content (i.e. the subject matter) of the rules under consideration”.Footnote 140 The second approach is a broader one, where “relevant” has a wider meaning and is not restricted to the subject matter of the rule under consideration.Footnote 141 Those who argue in favour of the second approach contend that, since Article 30 of the VCLT deals with “successive treaties relating to the same subject-matter”, the word “relevant” in Article 31(3)(c) “should not be interpreted as requiring unity of subject-matter”.Footnote 142 In other words, under the broader approach, any rule of international law will be a “relevant” rule,Footnote 143 although bringing in the chapeau of Article 31(3)(c), which requires the “relevant” rule to be taken into account together with the “the context” of the treaty, could narrow the broader interpretation of the word “relevant”.Footnote 144
With regard to the police powers rule, even if we include both Rule 1 and Rule 2 given by the Philip Morris tribunal, it will pass the test of “relevant” even under the narrower approach because of the similarity in terms of subject matter. Moreover, under the broader approach, it is surely also a “relevant” rule applicable in the relations between Switzerland and Uruguay. However, as with the issue of “applicable”, the tribunal did not discuss the issue of relevance while invoking Article 31(3)(c) of the VCLT when considering the police powers rule.
On the basis of the above discussion, one can conclude that the police powers rule satisfies the three limbs of paragraph (c). We now turn to the chapeau of Article 31(3)(c).
E. The Chapeau of Article 31(3)(c)
The chapeau of Article 31(3) ensures that international rules admitted as interpretative materials, relying upon paragraph (c), are “taken into account” “together with the context” in interpreting the treaty in question. The significance of the word “context” in Article 31(3)(c) was highlighted by Judge Higgins in her separate opinion in the Oil Platforms case, where she said that, while considering relevant rules, one cannot “ignore that Article 31, paragraph 3, requires ‘the context’ [of the treaty] to be taken into account”.Footnote 145 Keeping the context in mind, the rule so admitted will have to be given an appropriate interpretative weight.Footnote 146 In other words, it is one thing to state that a rule of CIL is a relevant rule applicable in the relations between the parties, and it is quite another to decide what bearing it will have on the treaty provision that is being interpreted. Given the analytical structure of Article 31(3)(c), one should be mindful of the distinction between using relevant rules to interpret the treaty and applying the same rules directly to the facts at hand, resulting in the displacement of the applicable law.Footnote 147 The “rule” found admissible in the treaty interpretative process should be used to clarify the content of the provision being interpreted, and not to limit the treaty provision to the scope and content of the admissible rule.Footnote 148
The Philip Morris tribunal did not take the police powers rule into account together with the context of the Switzerland-Uruguay BIT in interpreting Article 5(1) of the treaty. There is no discussion of what interpretative weight was attached to the police powers rule or how the police powers rule assisted the tribunal in finding the meaning of the expropriation provision in Article 5(1). The tribunal merely cited Article 31(3)(c), before directly applying the police powers rule to the facts at hand by judging the consistency of the SPR and the 80/80 Regulation with the police powers rule.Footnote 149 The tribunal concluded that, since these regulations meet the police powers rule, they are not expropriatory. The tribunal missed the critical distinction between the admissibility of an extraneous rule as treaty interpretative material and the interpretative weight that is to be attached to this interpretative material.Footnote 150 This represents a classic example of the minimalism that Ortino has rightly identified as one of the egregious failures of reasoning by ISDS tribunals.
The approach of the tribunal reduces Article 31(3)(c) to a mere licence to incorporate an extraneous rule into the treaty and apply it directly to the facts at hand, rather than contextualizing it according to the treaty. In fact, while discussing the police powers rule, the tribunal did not refer to the expropriation provision in the BIT at all.Footnote 151 Consequently, the tribunal erringly ended up displacing Article 5(1)—the very provision it was mandated to interpret—due to its reliance on the CIL police powers rule.
The approach of the Philip Morris tribunal in using Article 31(3)(c) is not very different from the approach the International Court of Justice [ICJ] adopted in the Oil Platforms case.Footnote 152 This case involved the destruction of Iranian oil platforms by the US, at the time of Iran-Iraq war when the US was defending its shipping in the Gulf.Footnote 153 Iran alleged that the US, inter alia, had violated Article X(1) of the 1955 Treaty of Amity, Economic Relations and Consular Rights between the US and IranFootnote 154 that provides for freedom of commerce.Footnote 155 The ICJ found that Article XX (1)(d)Footnote 156 of the 1955 treaty does not preclude the Contracting Parties from adopting “measures … necessary to protect its essential security interests”, as providing the parties a possible defence on the merits.Footnote 157 At the time of considering the merits, the majority judgment took the general international law defence of self-defence as the starting point by making reference to Article 31(3)(c) of the VCLT.Footnote 158 This approach meant that the Court, as Judge Higgins in her separate opinion said, did not interpret Article XX(1)(d) of the 1955 treaty using treaty interpretation rules, but invoked Article 31(3)(c) to displace Article XX(1)(d) (the applicable law) with CIL on the use of force.Footnote 159
The Saluka tribunal made a similar conceptual error. As mentioned before, the mandated task of the tribunal was to interpret the term “deprivation”, which provided the provision on expropriation. The tribunal invoked CIL on expropriation using Article 31(3)(c) and thereby brought in the police powers rule.Footnote 160 It then, just as the Philip Morris tribunal did, directly applied the police powers rule, although the mandate of the tribunal was to interpret and apply Article 5 of the Netherland-Czech Republic BIT. Through the direct application of the police powers rule, which led to the displacement of the applicable law, the tribunal concluded that the state had not expropriated foreign investment, i.e. had not violated Article 5 of the BIT.Footnote 161
IV. CONCLUSION: WHAT SHOULD THE TRIBUNAL HAVE DONE AND WHAT SHOULD OTHER TRIBUNALS DO IN THE FUTURE?
From the discussion above, the Philip Morris tribunal’s reasoning in using Article 31(3)(c) to interpret the expropriation provision in the Switzerland-Uruguay BIT was arguably flawed. The reasoning was internally inconsistent, based on a misuse of arbitral precedents and suffered from minimalism. The Philip Morris tribunal should not have divided its analysis of the expropriation provision into two parts as it is doctrinally and conceptually problematic. Given these errors, the core question is: What should the Philip Morris tribunal have done?
To answer this, let us briefly discuss Article 5(1) of the Switzerland-Uruguay BIT. As mentioned before, Article 5(1) prohibits countries from expropriating foreign investment either directly or indirectly, i.e. taking measures that would have the same effect on foreign investment as direct expropriation, unless the following conditions are met: (1) the investment has been expropriated for public benefit as established by law; (2) the investment has been expropriated on a non-discriminatory basis following due process of law; and (3) effective and adequate compensation is paid.
Thus, to prove that the above-mentioned provision has been breached, two conditions have to be satisfied. First, it must be proved that the state has expropriated foreign investment. Second, this expropriation must have taken place without satisfying the three criteria mentioned above. If the host state expropriates foreign investment, directly or indirectly, satisfying all these three conditions, it would amount to lawful expropriation. Otherwise, it would amount to unlawful expropriation. In Article 5(1), “public benefit” exists as a criterion to determine the legality of expropriation. Article 5(1) does not allow “public benefit” to be used to determine whether expropriation has taken place.Footnote 162 According to Article 5(1), the criterion to determine expropriation is whether the regulatory measure resulted in substantial deprivation of foreign investment. In other words, under Article 5(1) of the Switzerland-Uruguay BIT, borrowing from Paparinskis, “expropriation is only permitted for bonafide regulation if compensation is paid” and not otherwise.Footnote 163 In short, the rule given in Article 5(1) is “compensation in all cases”,Footnote 164 i.e. even when regulation is non-discriminatory and bona fide aimed at achieving a public purpose, provided it results in substantial deprivation of foreign investment. The criterion to distinguish legitimate regulation from expropriation is the quantum and degree of deprivation.Footnote 165
While the Philip Morris tribunal clearly identified the above principles and applied them to the facts, its discussion and application of the police powers rule is not convincing. It neither identified the police powers rule clearly nor did it rigorously discuss how the police powers rule satisfied the four components of Article 31(3)(c).
In order to complete the argument, let us, borrowing again from Paparinskis, identify the customary police powers rule as follows: “no expropriation for bonafide regulation despite substantial interference with investment” or in other words, “no compensation in some cases”.Footnote 166 Now, the question is how this rule will be applied to interpret Article 5(1) of the Switzerland-Uruguay BIT using Article 31(3)(c) of the VCLT. It will first have to be established that the rule is a “relevant” rule of international law “applicable in the relations between the parties”, and thus admissible as interpretative material under Article 31(3)(c) to interpret Article 5(1). The rule that we have identified would pass the three tests of Article 31(3)(c), as has been discussed in Section III. The next question is what interpretative weight should be given to this customary rule. The customary rule of “no compensation in some cases” should be compared and contrasted with the treaty norm of “compensation in all cases”, and thereafter rejected.Footnote 167 Since the treaty rule, as discussed earlier, does not permit substantial interference with investment without compensation, it cannot be read to allow substantial interference with investment in certain cases (such as when the measure is bona fide and for public purpose) without compensation.
This case is also significant because it raises the larger issue of the relationship between the police powers rule and the expropriation provision in the BIT. It is critical to bear in mind that invoking the police powers rule in assessing expropriation claims would mean relying upon, inter alia, “public welfare”, “public benefit”, or “public interest”Footnote 168 as the criteria to determine expropriation. This, in turn, could result in a strange contradiction where, despite substantial deprivation of investment, there would not be any expropriation because the regulatory measure would have been adopted to achieve a “public purpose”, subject to it being nondiscriminatory and enacted under due process. Such an interpretation would defeat the very purpose of having expropriation provisions like Article 5(1) in the Switzerland-Uruguay BIT.Footnote 169 Indeed, the law of expropriation as codified in numerous BITs clearly recognizes that a nondiscrimnatory measure that deprives the investor of her investment is expropriatory even if enacted under due process to achieve a public benefit.Footnote 170 If presence of public benefit becomes the criterion to distinguish between compensable and non-compensable regulation, it would allow host states to shift the burden of achieving public benefit onto foreign investors even for regulatory measures that lead to deprivation of investment.Footnote 171 As it has been argued, “historically, police powers have never been meant to cover regulations amounting to expropriations, except perhaps in situations where there is a state of emergency or a state of necessity”.Footnote 172
In other words, it is one thing to state that a host state has the right to adopt nondiscriminatory regulatory measures for a public purpose, and it is another thing to decide how this will be applied in the light of the fact that this very host state has accepted restrictions on its right by entering into a BIT containing the aforementioned expropriation provision. The tribunal in ADC v. Hungary Footnote 173 held that, while a sovereign nation possesses the inherent right to regulate its domestic affairs, the exercise of this right must have its boundaries.Footnote 174 The tribunal recognized that the relevant BIT provided such boundaries.Footnote 175 Similarly, the Azurix tribunal found the criterion that the “host state is not liable for economic injury that is the consequence of bonafide regulation within the accepted police powers of the state” was insufficient to determine what constitutes indirect expropriation, and recognized that a legitimate measure serving public purpose could give rise to a compensation claim.Footnote 176
An apprehension often expressed is that focusing solely on the effect of the regulatory measure, not the purpose, to determine indirect expropriation will reduce the regulatory space available to host countries to adopt measures in the public interest. However, this apprehension is lessened if one adopts the “substantial deprivation” test. This test will ensure that an adverse effect on foreign investment will not constitute expropriation, unless that effect results in a “substantial deprivation” of foreign investment.Footnote 177 This high threshold gives the host state ample space to adopt a number of regulatory measures for the public purpose without worrying about expropriation. The analysis of the tribunal on the effect of Uruguay’s regulatory measures on Philip Morris’s investment is testament to this fact.Footnote 178 As discussed in this paper, the tribunal held that simply because regulatory measures lead to some adverse effect, such as deteriorating profits, this is not sufficient to establish “substantial deprivation”.
The Philip Morris tribunal’s analysis of the effect of the regulatory measure on investment clearly showed that Uruguay’s tobacco regulations did not lead to expropriation. There was no need for the tribunal to engage in a discussion of the police powers rule. The reasoning of the tribunal on Article 31(3)(c) of the VCLT to deal with extraneous norms such as the police powers rules raises a number of conceptual questions. This paper has tried to address some of these issues. Future ISDS tribunals should show greater doctrinal clarity in dealing with these complex issues. Such clarity in legal reasoning will go a long way in increasing the legitimacy of the system for all stakeholders, including states and foreign investors.
