I. INTRODUCTION
International law has, until recently, played a marginal role in the management of financial crises. Critical attention has instead focused on the lending practices of international financial institutions and their contribution to outbreaks such as the debt crises of the 1980s, the Mexican peso crisis of 1994 and the Asian crises of 1997–1998.Footnote 1 The financial crisis that enveloped Argentina in late 2001 has revealed a new legal subject with the serious potential to constrain State autonomy in mitigating adverse effects of such crises. This is the network of international investment treaties that grant foreign investors the right to challenge signatory State laws for breach of treaty commitments in a range of arbitral institutions.Footnote 2 Foreign investors began with increasing urgency to invoke these broad treaty disciplines to challenge regulatory measures implemented by Argentina in the aftermath of its financial crisis.Footnote 3 The arbitral cases to date have largely ruled against Argentina and awarded significant monetary compensation to the claimant investors.Footnote 4
These cases engage a treaty exception rarely adjudicated upon at international law that entitles a signatory State to pass emergency measures ‘necessary’ for the maintenance of ‘public order’ or the protection of ‘essential security interests’.Footnote 5 This type of exception is not confined to international investment treaties and can be found in other fora at international law.Footnote 6 Despite attracting academic comment in these other fields,Footnote 7 there has been little direct jurisprudence surrounding this exception.Footnote 8 The cases to emerge from the Argentine financial crisis thus offer a unique and timely opportunity to assess the scope and operation of this important treaty defence.Footnote 9
The arbitral tribunals that have ruled against Argentina's invocation of the exception adopt a remarkable interpretative approach. They draw extensively on customary principles of necessity (taken to be represented by the work of the International Law Commission) to guide their interpretation and application of the treaty exception. The stringent conditions of the plea of necessity in the ILC Articles on State ResponsibilityFootnote 10 are applied as operative components of the treaty defence. The outcome in all but a few cases has been to refuse Argentina's claim to abrogation of liability.
On first view, these cases might simply be taken as part of the contemporary project of managing fragmentation of international law.Footnote 11 By weaving the treaty exception and customary plea together, the cases offer a visible rejoinder of the criticism that international investment treaties are a self-contained regime at international law,Footnote 12 divorced from a broader normative framework. After all, the resistance in early cases to using custom as a defining component of other treaty provisions—such as the obligation to accord fair and equitable treatment—has proven contentiousFootnote 13 and even triggered the intervention by State parties in certain settings.Footnote 14 This favourable, impressionistic view of the emerging jurisprudence on the treaty exception though is mistaken. As a recent Annulment Committee constituted in this area has noted, the arbitral tribunals to date fail to rigorously examine the precise relationship between the treaty defence and customary plea.Footnote 15
My principal aim in this article is to uncover and analyse the key methodological possibilities of interpreting the treaty exception and its relationship to customary law. To do so, I adopt a particular methodology of my own distinct from existing approaches in the field. My analysis focuses on objective expressions of the function and purpose of the investment treaty regime and its connection to, and at times departure from, the customary law of State responsibility. Those objective expressions—reflected in the immediate treaty text, contextual indicators and a broader historical narrative—are prioritized in my analytical structure over evidence of the subjective intentions of the State parties and their negotiators. In this respect, my approach intentionally tracks the ordering of particular interpretative techniques as codified in the Vienna Convention on the Law of Treaties 1969.Footnote 16 I do so as my theses are both descriptive (identifying a flaw in the use of these rules in the dominant methodology to date) and normative (presenting a framework as to what tribunals should do in future cases). This is not to deny that in certain treaty settings, differentiated hermeneutics (such as a search for the subjective intention of the parties) may offer critical guidance to an adjudicator. On the whole, the network of Bilateral Investment Treaties (BITs) seems, perhaps counter-intuitively, ill-equipped for delving into originalist interpretation. We are dealing with a device not unlike a standard-form contract employed in a domestic setting; BITs typically follow a model form offered by a (strong) regulation maker to an (at times, weak) regulation taker.Footnote 17 This translates into a surprising lack of evidence on the part of the regulation taker as to their intentions in entering into the treaty device.Footnote 18 The temptation is then to draw on the extensive evidence of the strong player as entirely conclusive of the common intention of the parties.Footnote 19 This inevitably cashes out into a deterministic claim of such treaties as devoted solely to investment protection where in fact the silence of the regulation taker may mask multiple, even conflicting goals.Footnote 20
I begin the next Part II by tracing the factual matrix surrounding Argentina's entry into this unique area of international law, the outbreak of the financial crises and the abrogation of contractual rights that triggered investor claims for compensation. Readers familiar with this complicated but essential fact-set may wish to proceed directly to Part II(C). There I begin my substantive analysis of the text and history of the operative legal standards examined in the jurisprudence respectively, article XI of the 1991 US–Argentina Bilateral Investment Treaty and the project of development (and perhaps even reflection) of the customary plea of necessity in ILC article 25.
Turning then to the cases, there are at least three methodologies open to an adjudicator in understanding the relationship between the treaty exception and the customary plea of necessity. These can be termed respectively methodologies I (confluence), II (lex specialis) and III (primary—secondary applications). Method I is clearly the dominant and most restrictive approach in the jurisprudence to date, whereby tribunals expressly conflate the treaty defence with the customary plea of necessity. This, I argue, is mistaken both on careful interpretation of these legal standards but also on an investigation of the complex and nuanced history of the shift from customary to treaty protections for foreign investors.
The choice for future adjudicators comes down to an election between methods II (lex specialis) and III (primary–secondary applications). Both have a distinct advantage of presenting a plausible account of the relationship between the treaty defence and customary plea. There are however critical and open interpretative questions of the scope of priority to be accorded to the treaty defence under the lex specialis reading, which have largely been ignored by those sympathetic to this account.Footnote 21 On the other hand, the taxonomy of method III (primary–secondary applications) clearly separates findings of wrongfulness and State responsibility at international law. It characterizes all of the treaty provisions at issue—both forms of obligation and exception—as primary legal standards that determine whether a State has committed a wrongful act at international law. Under this approach, it is only if breach is determined by the composite application of these treaty rules, that an adjudicator should examine the secondary possibility of the customary plea of necessity to preclude wrongfulness as a matter of State responsibility. The key implication of method III is to require an adjudicator to attend to interpretation of key aspects of the treaty defence without simply borrowing or transplanting from the customary plea. This has important doctrinal and normative advantages, not least as a means of responding to the concern that investment treaties represent a classic embodiment of incomplete contracting. The logical implication though of method III is to raise a host of difficult interpretative questions surrounding the treaty exception and the article concludes by offering an interpretative framework to future tribunals charged with that challenge.
II. THE ARGENTINE FINANCIAL CRISIS AND INVESTOR–STATE LITIGATION
To assess Argentina's claim to defence at international law, it is essential that we have an accurate indication of its response to the crisis as well as the causes and magnitude of that event. Many of the underlying legal categories at issue—particularly the ‘necessity’ of chosen means to effect permitted regulatory ends—require an adjudicator to examine and test the rationality of the regulatory record as a whole.
A. The Crisis and Argentina's Regulatory Response
Our starting point in uncovering the factual record should begin before the onset of the financial crisis in late 2001. The advantages enjoyed by foreign investors and rendered inutile by Argentina's response to the crisis have their origins in a broader macroeconomic change. In 1991, Argentina passed a Convertibility Law that fixed the exchange rate of its currency with the US dollar.Footnote 22 The goal was to constrain the spiral of hyperinflation and relative stagnation that had plagued the Argentine economy since the 1960s.Footnote 23 The strategy was one of intentional ‘hands tying’ on the part of the Argentine Government. Convertibility prevented monetary authorities from simply increasing monetary supply depending on shifts in economic conditions, with resulting inflationary effects.
In the early 1990s, Argentina initiated an extensive privatization program. By 1994, over 90 per cent of State enterprises had been privatized including the telephone, electricity, gas and water utilities.Footnote 24 The policy shift to privatization was linked to the convertibility strategy. As this required continuing reserves of foreign capital, the funds garnered through privatization were used to bolster convertibility of the Argentine peso.Footnote 25 Argentina though faced a reputational overhang in its project of attracting the entry of foreign investment into these sectors.Footnote 26 There were two mechanisms used by Argentina to signal its break from past practices. First, it offered a range of specific incentives to private companies who acquired ownership interests in various utility sectors.Footnote 27 Tariffs were to be calculated in US dollars, conversion to pesos occurred at the time of billing and most crucially, tariffs were to be adjusted every six months under a key US inflation index (the US Producer Price Index).Footnote 28 Secondly, Argentina entered into a range of Bilateral Investment Treaties (BITs) throughout the 1990s.Footnote 29
BITs offer a range of specific guarantees of protection to foreign investors to mitigate the risk of operation in a host State.Footnote 30 These include key obligations of non-discrimination,Footnote 31 a commitment on the part of the signatory State to accord ‘fair and equitable treatment’ to a foreign investorFootnote 32 and a guarantee that the State compensate the foreign investor for takings of foreign assets.Footnote 33 These treaties displace the customary rule that a home State of an investor must elect to exercise diplomatic protection on behalf of the investor.Footnote 34 Instead, a foreign investor is granted the unique right to bring a claim for breach of a treaty directly against a signatory host State.Footnote 35 There are various systems of adjudication presented in a typical BIT. One of the most active of these—and later invoked against Argentina—is the World Bank's Convention on the Settlement of Investment Disputes Between States and Nationals of Other States (ICSID).Footnote 36
Up until 1995, the Argentine policy shift had generated significant positive outcomes. There was significant entry of foreign investment and inflation had subsided to levels Argentines had not seen in decadesFootnote 37 combined with sizeable rates of economic growth.Footnote 38 In the aftermath of the Russian financial crisis, a tipping point was reached in late 1998 due to a range of exogenous factors.Footnote 39 Confidence in the economy began to weaken with investors electing to withdraw capital. In late 2000 and early 2001, the Argentine government sought to buttress convertibility by arranging for emergency loan finance from the IMF.Footnote 40 It offered external lenders the possibility of a voluntary debt swap in an attempt to hold off devaluation of the peso.Footnote 41 By July 2001, Argentina had moved to a drastic zero deficit position; it reduced government spending—including cuts of 13 per cent in government salaries and pensions—in an attempt to prevent the need to further borrow external funds.Footnote 42
These efforts to stave off devaluation proved futile. There was a run on the domestic banking system in November 2001 as depositors lined up to withdraw 6 per cent of total bank deposits over three days.Footnote 43 The Government responded with a freeze on bank withdrawals. Account holders could only withdraw the equivalent of US$250 in cash per week and domestic economic activity began to grind to a halt, particularly in vulnerable sectors that rely on cash transactions (including the self-employed).Footnote 44 These progressively harsh measures generated significant popular unrest with outbreaks of riots and looting leading to a score of fatalities. From December 2001 to January 2002, five different presidents were appointed in a ten-day period.Footnote 45 By January 2002, 25 per cent of the urban workforce was unemployed with a majority of the population below the official poverty line.Footnote 46 The Government responded to this dire situation by officially abrogating convertibility through an Emergency Law whereby the peso immediately declined in value by almost 70 per cent.Footnote 47
The Emergency Law also targeted licensees in various utility sectors, giving rise to the later legal claims of foreign investors in those sectors. These measures comprised two, related elements. Tariffs would now be assessed in devalued pesos rather than in US dollars and the contractual right to adjust those tariffs in line with US inflation indices was terminated.Footnote 48
B. The Onset of Litigation
The financial losses to flow from Argentina's Emergency Law have triggered dozens of claims by foreign investors alleging breach of Argentina's investment treaty obligations. Argentina has invoked defined treaty exceptions and customary law in its defence to those claims. The invocation of the exceptions is the focus of our inquiry (rather than the question of breach of operative standards of liability).
There have been five ‘first instance’ arbitral awards that have directly considered the relationship between the customary and treaty exceptions invoked by Argentina. These cases were all initiated within the ICSID system and concern the same subject treaty, the 1991 US–Argentina BIT.Footnote 49 There is an initial cluster of four awards; CMS v Argentina (May 2005),Footnote 50LG&E v Argentina (October 2006),Footnote 51Enron v Argentina (May 2007)Footnote 52 and Sempra v Argentina (September 2007).Footnote 53 The CMS, Enron and Sempra Tribunals rule against Argentina and adopt similar methods of reasoning. The LG&E Tribunal on the other hand, partially finds in favour of Argentina on its invocation of the exceptions. What follows is the release of a ruling by an ICSID Annulment Committee on the CMS award. There is no formal system of appellate review on questions of law within the ICSID but an unsuccessful party can apply for annulment on narrow grounds essentially reflecting due process failures.Footnote 54 While largely finding against Argentina, the Annulment Committee—adopting the language and gravitas of an appellate judicial bodyFootnote 55—offers a range of targeted criticisms of the legal reasoning adopted by the CMS Tribunal.Footnote 56 The final and fifth instalment (to date) is the Continental v Argentina (September 2008) award ruling largely in favour of Argentina, with a distinct method of its own.
These five rulings engage fundamentally different and at times conflicting methods of interpreting the relationship between the operative treaty exception and relevant customary law. Before entering this complex jurisprudential mix, it is essential that we have a clear understanding of the legal terrain before us.
C. The Operative Legal Standards
1. The plea of ‘necessity’ at customary international law
The ICSID tribunals constituted in the Argentine cases draw heavily on customary international law in determining Argentina's liability. The question then is whether custom might somehow excuse State liability in exceptional circumstances such as financial crises. This raises an initial methodological question of where an adjudicator should look in determining the existence of such a customary rule. Custom comprises State practice coupled with a conviction (opinio juris) that the practice reflects a legal obligation.Footnote 57 The identification of these elements of custom has, in certain settings, proven to be a difficult conceptual exercise.Footnote 58 None the less, one might expect the tribunals as a starting point to at least attempt to identify primary sources in assessing the existence of such an exceptional defence.Footnote 59
The tribunals do not engage in such a search but instead accept the International Law Commission's (ILC) Articles on the Responsibility of States for Internationally Wrongful Acts as constituting customary international law.Footnote 60 On first view, this is perhaps not surprising. The ILC was established in 1947 by the United Nations General Assembly with the explicit objective of the progressive development of international law and its codification.Footnote 61 The topic of State responsibility was selected as an early candidate in the work of the ILC and it has produced multiple reports under the auspices of different rapporteurs in this difficult area. That work has itself been accepted as codifying custom by a range of international tribunals including the ICJ.Footnote 62
There remains though a legitimate question of whether the work of the ILC should be taken as a final expression of customary law, at least in this area. For one thing, there has been no attempt to transform the ILC Articles into a treaty through a diplomatic law-making conference.Footnote 63 David Caron has also argued that several of the broad (and often controversial) ILC articles are better viewed as part of the ILC's separate mission of the ‘progressive development’ of international law.Footnote 64 This would see them as ‘an authoritative scholarly statement of the law that would provide some guidance and clarity, yet also grow and change as it confronted the tests that the world presents’.Footnote 65 Investing them with the status of a formal source of law, on the other hand, would obviate this flexibility as adjudicators would confer excessive authority to the ILC articles. Caron's concern, expressed in 2002, is prescient given the inflexible application of the ILC Articles in the method of confluence adopted in many of the Argentine cases.
If we accept however, along with the ICSID tribunals, that the work of the ILC codifies customary international law, then article 25 is our first potential legal standard. Given the extensive reliance on that Article by many of these tribunals, it is worth extracting it in its entirety:
1. Necessity may not be invoked by a State as a ground for precluding the wrongfulness of an act not in conformity with an international obligation of that State unless the act:
(a) is the only way for the State to safeguard an essential interest against a grave and imminent peril; and
(b) does not seriously impair an essential interest of the State or States towards which the obligation exists, or of the international community as a whole.
2. In any case, necessity may not be invoked by a State as a ground for precluding wrongfulness if:
(a) the international obligation in question excludes the possibility of invoking necessity; or
(b) the State has contributed to the situation of necessity.Footnote 66
It is important to give careful consideration to the precise scope of operation of this provision. Along with a number of other provisions in Chapter V of the ILC articles, this article sets out particular circumstances that will preclude a wrongful act at international law.Footnote 67 Those clauses are thus logically predicated on the initial finding of a wrongful act at international law.
The current ILC Articles do not themselves define what will constitute a wrongful act. This is not accidental but instead represents an intentional departure from earlier ILC efforts to codify substantive principles of State responsibility. ILC Rapporteur García Amador produced a series of six reports from 1956 to 1961 that attempted the ambitious and difficult task of codifying the primary, substantive rules that would trigger State responsibility for injuries to aliens.Footnote 68 These included the politically charged issue of whether States were required to compensate foreign actors for nationalization of private property even when linked to decolonization processes. The difficulty of reaching agreement in this highly contested area drove a distinct shift in perspective and strategy in the operation of the ILC in the early 1960s. A new ILC Rapporteur, Roberto Ago, began work on secondary rules of State responsibility that would only apply on the prior finding of a wrongful act at international law.Footnote 69 These rules did not themselves identify what would constitute the wrongful act in the sense of operative rights and obligations. They instead provide the framework for determining whether primary obligations have been breached (such as the key issue of attribution of conduct to a State) together with the consequences that will flow at international law generally from a wrongful act.Footnote 70
The current ILC rules—finalized by Rapporteur James Crawford in 2001—also adopt this taxonomy. The question of wrongfulness is assessed by the primary rules of international law applied to a given fact-set, an essential precondition to the invocation of the ILC rules.Footnote 71 It is also exclusively a question for the substantive legal regime at issue, whether customary or conventional in origin. A breach of these primary obligations constitutes a wrongful act that in turn, triggers the operation of the articles in ILC Chapter 5.Footnote 72 Those articles, including article 25 on necessity, offer the secondary possibility of precluding wrongfulness as a matter of international law. ILC article 25 then acts a general defence that can apply to breach of any international legal obligation (other than a peremptory rule of international law).
Given this expansive operation, it is entirely understandable that ILC article 25 sets out a range of highly stringent conditions for the secondary plea of necessity to apply. This is also driven by the concern of potential abuse common to any form of derogation.Footnote 73 The greatest limiter on the customary plea of necessity is ILC article 25(1)(a) which requires that the chosen governmental measure (‘act’) be the ‘only way’ for a State to meet that objective.
There is no inherent, natural meaning of when a particular measure will be ‘necessary’ to achieve a given end. Consider the jurisprudence on a similar sort of exception in the law of the World Trade Organization (WTO). Articles XX(a), (b) and (d) of the General Agreement on Tariffs and Trade (GATT) require an otherwise GATT-inconsistent measure to be ‘necessary’ for certain ends in order to avoid liability.Footnote 74 The WTO Appellate Body in the Korea-Beef case presented the interpretative options on the term ‘necessary’ in GATT Article XX(d) across a spectrum:
We believe that, as used in the context of Article XX(d), the reach of the word ‘necessary’ is not limited to that which is ‘indispensable or ‘of absolute necessity’ or ‘inevitable’. Measures which are indispensable or of absolute necessity or inevitable to secure compliance certainly fulfil the requirements of Article XX(d). But other measures, too, may fall within the ambit of this exception. As used in Article XX(d), the term ‘necessary’ refers, in our view, to a range of degrees of necessity. At one end of this continuum lies ‘necessary’ understood as ‘indispensable’; at the other end, is ‘necessary’ taken to a mean as ‘making a contribution to’.Footnote 75
Unlike GATT article XX(d), there is no continuum of meanings in the application of the customary plea of necessity. ILC article 25(1)(a) supplies the operative test; the chosen means must be indispensable as the ‘only way’ for the State to safeguard its essential interests. Where there are multiple options open to a State, the customary defence is simply precluded. In the few occasions where the customary plea has been raised at international law, it is this strict test of means-end scrutiny that has been applied to defeat invocation of the plea.Footnote 76 The origins of this exceedingly stringent test lie in the claim of necessity surrounding the use of force in self-defence.Footnote 77 In the famous 1841 Caroline controversy between the United States and Great Britain, US Secretary of State Daniel Webster observed that in order for a State to act in anticipatory self-defense, there must be a ‘necessity of self-defense, instant, overwhelming, leaving no choice of means and no moment for deliberation’.Footnote 78 Here we have in a nutshell the danger of formalizing or applying a general legal rule without careful attention to the context of its emergence. It is perhaps sensible to advocate for such a stringent test to the use of force between States. It is questionable though whether the same sort of strict means-end scrutiny should apply to all other areas of extraordinary State action.
2. The treaty exception: article XI of the US–Argentina Bilateral Investment Treaty
The five cases that are the subject of this article engage the same treaty exception, article XI of the 1991 Argentine–US BIT. That exception is nested within an overall treaty regime that imposes significant obligations on a State party in their dealings with foreign investors of the other signatory State.Footnote 79 The treaty as a whole then prescribes the sort of primary rules whose application might lead to a finding of a wrongful act at international law. This is the predicate for the invocation of the customary plea of necessity in the taxonomy of the ILC.
Article XI of the US–Argentina BIT is remarkably brief and simply states:
This Treaty shall not preclude the application by either Party of measures necessary for the maintenance of public order, the fulfilment of its obligations with respect to the maintenance or restoration of international peace or security, or the protection of its own essential security interests.Footnote 80
The treaty provision thus encompasses three permitted regulatory objectives: (1) public order; (2) obligations with respect to the maintenance or restoration of international peace or security; and (3) essential security interests. A chosen measure of a signatory State must also be ‘necessary’ to achieve those goals, in order to fall within the scope of the exception.
There are various formulations of this sort of necessity exception in different treaty instruments. GATT articles XXI(b) and (c) offer a useful comparison with the BIT exception:
Nothing in this Agreement shall be construed:
…
(b) to prevent any contracting party from taking any action which it considers necessary for the protection of its essential security interests
(i) relating to fissionable materials or the materials from which they are derived;
(ii) relating to the traffic in arms, ammunition and implements of war and to such traffic in other goods and materials as is carried on directly or indirectly for the purpose of supplying a military establishment;
(iii) taken in time of war or other emergency in international relations; or
(c) to prevent any contracting party from taking any action in pursuance of its obligations under the United Nations Charter for the maintenance of international peace and security.Footnote 81
Three key points can be drawn from a comparison between GATT article XXI and the US–Argentine BIT exception. First, the text of GATT article XXI(b) expressly reserves authority for a signatory State to determine what ‘it considers necessary’ for the protection of its essential security interests. There is no equivalent language within the BIT exception. This critical absence in the text makes it difficult to convincingly claim that the BIT exception is of an entirely self-judging nature.Footnote 82 Such a claim is especially marginal given the ruling of the ICJ on an almost identical treaty provision to that of the BIT exception in the Nicaragua case.Footnote 83 This is not to deny that there are serious competence concerns surrounding the capacity of an arbitral tribunal to resolve particular questions, especially the scope of a State's ‘essential security interests’. It appears clear though that the BIT exception contemplates some role for an adjudicator in the application of the exception.Footnote 84
Second, unlike the BIT exception, GATT article XXI(c) identifies the source of the obligations ‘for the maintenance of international peace and security’ as located within the United Nations Charter. This would be a logical candidate for triggering the BIT exception even in the absence of such express wording. Under Chapter VII of the Charter, the Security Council has express authority ‘to maintain or restore international peace and security’.Footnote 85 The implications of this branch of the BIT exception are often under-appreciated.Footnote 86 It deals with an objective that falls outside the customary arena and offers crucial textual evidence that the State parties did not simply seek to displace or incorporate custom where this would result only in higher standards of investment protection. The multiple branches of the exception show that they also contemplated the priority of State and even supra-national action (such as the institutional operation of Chapter VII of the UN Charter) in defined settings.
The third key difference between the BIT exception and GATT article XXI goes to the notion of ‘essential security interests’. The GATT provision expressly limits that concept to traditional, inter-State security objectives including military and defence matters. Although the BIT exception also speaks of ‘essential security interests’, there is no indicative list of examples of that concept, leaving open the possibility that security be understood by an adjudicator in dynamic rather than static terms. Moreover, the BIT exception offers a new ground of exemption for the ‘maintenance of public order’ that is entirely absent from the GATT article XXI. As with the branch of the treaty exception for ‘maintenance of international peace and security’, this is again important evidence of contemplation of priority of State action over and above the permissible limits at customary law.Footnote 87 Finally, there is the charged question of when a given measure will be ‘necessary’ to achieve the permissible objectives listed in the BIT exception. There is no supplied test of means–end inquiry in the BIT exception, a critical departure from the secondary plea of necessity in ILC article 25.
III. METHODOLOGY I: CONFLUENCE
There have been five arbitral tribunals that have examined Argentina's invocation of the BIT exception. Three of these awards—CMS, Enron and Sempra—expressly conflate the treaty defence with the customary plea of necessity, which is taken to be represented by ILC article 25 (‘method I’).
A. The Cases: CMS, Enron and Sempra Awards
There are subtle differences between the three tribunals in their adoption of this dominant methodology. The CMS Tribunal makes a show of first analysing the customary exception and then separately considering the treaty clause.Footnote 88 Yet in substance, its analysis of the constituent components of the treaty draws heavily and is largely inseparable from the customary principle. The Sempra Tribunal is more explicit in conflating the two standards:
This Tribunal believes … that the Treaty provision is inseparable from the customary law standard insofar as the definition of necessity and the conditions for its operation are concerned given that it is under customary law that such conditions have been defined.Footnote 89
Moreover, the Sempra Tribunal gives an indication of its motivation in choosing this approach. The ‘problem’ for the Tribunal is that ‘the Treaty itself did not deal with the legal elements necessary for the invocation of a state of necessity … the rule governing such questions will thus be found under customary law’.Footnote 90 As discussed earlier, the ILC articles supply the operative legal test for particular elements of the customary defence, most notably for when a chosen measure will be necessary. The stringency of these operative tests is linked to the generality of operation of the customary exception. This Tribunal seems simply to crave the same sort of guidance in relation to the treaty exception, without even considering the construction and relationship between the two legal standards.Footnote 91
An initial consequence of this approach is a singular focus on the treaty exception for ‘essential security interests’. There is, on the other hand, no substantive analysis of the alternate treaty ground for exemption of measures necessary to maintain ‘public order’, a logical if regrettable consequence of the method chosen by these tribunals. Explicit attention to the alternate treaty exemption for ‘public order’ would raise an obvious, key difference between the customary and treaty standards at odds with the elected approach.
The method I tribunals then present what might on first view, seem to be an expansive reading of the scope of a State's ‘essential security interests’. The CMS Tribunal for example, rejects the idea that these interests are limited to national security concerns of an international character as ‘there is nothing in the context of customary international law … that could on its own exclude major economic crises’.Footnote 92 Yet, the ILC articles are soon used to significantly narrow the operation of the treaty exception. The tribunals draw on the requirement that a State safeguard an essential interest against a ‘grave and imminent peril’ under ILC article 25(1)(a) to find that even Argentina's severe economic crisis would not be sufficient to fall within the treaty exception. The CMS Tribunal, for example, presents ‘a major breakdown with all its social and political implications’ and ‘total economic and social collapse’ as operative standards.Footnote 93 For the Sempra Tribunal, Argentina's economic crisis would have had to compromise the ‘very existence of the State and its independence’ to attract the treaty exception.Footnote 94 Once again, it is important to clearly draw out the implications of the methodology chosen by the Tribunal. There is no textual equivalent of the ILC standard of ‘grave and imminent peril’ in the treaty exception; these tribunals are simply importing from the customary norm an exceedingly stringent standard of operation.
Critically, the tribunals then draw on the ILC articles to determine whether Argentina's chosen measures were ‘necessary’ to protect its essential security interests. As we have seen, ILC article 25(1)(a) requires that the means chosen must be the ‘only way’ for the State to safeguard its interests. This highly stringent test of necessity is presented as a second ground for rejecting Argentina's defence. According to the CMS Tribunal, Argentina's regulatory response was not the only way to deal with the financial crisis as it possessed ‘a variety of alternatives, including dollarization of the economy, granting of direct subsidies to the affected population or industries and many others’.Footnote 95 The problem with this restrictive test is that it is always possible to conceive of multiple responses to a given financial crisis rendering the treaty exception largely ineffectual in this context.Footnote 96
There are two further uses of the ILC articles to narrow the operation of the treaty exception. The preclusion of the customary defence where a State has ‘contributed to the situation of necessity’ under ILC article 25(2)(b) is introduced as an operative element of the treaty exception.Footnote 97 Again, there is no reflection within the treaty text of such a limiter. Argentina's contribution to the crisis is characterized rather bluntly as ‘substantial’ and offered as yet another ground for rejecting its treaty defence.Footnote 98 Finally, the tribunals draw on the ILC articles on the question of compensation where the defending State otherwise meets the operative components of the treaty exception. The confluence methodology applied to this final issue is not entirely benign in the sense of the simple if mistaken transport of elements of the customary plea. On this point, the tribunals arguably engage in misrepresentation of the position at customary law. This goes to the critical issue of whether a State is under a duty to compensate even if entitled to invoke a circumstance (such as the plea of necessity) to preclude wrongfulness. ILC article 27(b) is clear on this point; the otherwise successful plea of necessity is ‘without prejudice to the question of compensation for any material loss incurred by the act in question’.Footnote 99 In other words, the ILC Articles offer a reservation to the question of compensation, presenting this as a case-by-case issue to be decided between the parties concerned.Footnote 100 The CMS Tribunal though endorses article 27 as ‘[establishing] the appropriate rule of international law on this issue’Footnote 101 but mistakenly characterizes it as imposing a ‘duty to compensate’.Footnote 102
B. Problems
There are two fundamental problems with the dominant method of confluence adopted in the jurisprudence to date.
1. The interpretative problem
The first flaw is clearly one of interpretation (and more subtly, path dependence given the sequencing and commonality of language in the CMS, Enron and Sempra awards). This method ignores the express taxonomy endorsed by the ILC. If we accept—as each of these tribunals do—that the ILC articles codify custom in this area, then there is a logical obligation to follow its presented analytical structure. The ILC articles distinctly present the customary defence of necessity as applicable to the issue of State responsibility. The elements of that defence only become operative if we have first identified a wrongful act at international law. The test for wrongfulness in turn is purely a question for the international legal regime at issue. This would involve an assessment of (i) whether there is breach of the legal obligations of the investment treaty and then (ii) whether the treaty exception applies to save that breach. It is only if we have an answer of ‘yes’ to (i) and ‘no’ to (ii) that we have a wrongful act at international law that would entitle an adjudicator to consider the customary defence. In short, the treaty and customary defences operate on different legal planes; the treaty defence comprises the set of primary legal rules that must be adjudicated upon before possibly attracting the secondary, customary defence. The blunt legal error in this interpretative method of confluence is especially surprising as other investor-State tribunals have proven themselves capable of identifying the key difference between primary and secondary rules in the ILC taxonomy.Footnote 103
2. The contextual problem
There may be some who regard this first point of criticism (and the distinction between primary–secondary planes) as excessively formalistic, a type of lawyerly trick devoid of any real substance.Footnote 104 There is, though, a second substantive flaw in the easy conflation of the customary and treaty standards. The method I tribunals ignore the complex and nuanced history of the shift from customary to treaty protections for foreign investors. This is a narrative that raises serious doubts as to the likelihood that the State parties would simply seek to crystallize the customary plea in the treaty exception. There are multiple factors that led State parties to build a network of investment treaties in the decades following the Second World War. The most fundamental of these—especially from the perspective of capital exporting States—was a conviction that custom was increasingly ill-equipped to deal with particular challenges faced by foreign economic actors.Footnote 105 The result is a project of tailoring treaty standards to regulate the conduct of States vis-à-vis foreign investors, largely but not entirely distinct from the position at customary international law.
To begin with, the immediate decades following the Second World War witnessed a dramatic rise in expropriation of private property by capital importing States.Footnote 106 Although customary international law had devised protections for certain forms of governmental expropriation,Footnote 107 capital-importing States began to invoke justifications for newer forms of taking.Footnote 108 The most notable of these was the practice of nationalization where particular industries were targeted as part of a State-led development program. This gave rise to an open question in the period as to whether nationalization would comprise a novel category of taking distinct from the existing customary rules on expropriation.Footnote 109 Custom, as an inherently evolutionary device, began to slowly shift in a direction favourable to capital-importing States on these questions.Footnote 110 The treaty provisions to emerge from this period counter this development and mitigate the new risk faced by foreign economic actors; nationalization along with expropriation is expressly included as a category of taking that requires full compensation by a signatory State.Footnote 111
The strategic concern as to the limitations of customary law in the contemporary period is also borne out when we consider the changes to dispute settlement processes. The customary rules on diplomatic protection entitle a State of an injured national to bring action against another State for injuries caused to that national by an internationally wrongful act.Footnote 112 Where the injured national is an artificial legal person such as a corporation, there is the added question of whether the ability to initiate diplomatic protection is limited to the State of the corporation or may in certain situations, extend to the State of individual shareholders (where these differ). This issue arose before the ICJ in the Barcelona Traction case which ruled that it is only the State of incorporation of the corporate entity, rather than the State of its controlling shareholders, that can invoke a claim for diplomatic protection.Footnote 113 The modern treaty provisions again counter this gap in customary international law. They offer the possibility of direct investor to State dispute settlement and as such are distinct from the State-to-State dynamic of customary international law. The ability to initiate this new form of dispute settlement extends beyond the foreign company that is operating in a signatory State. The typical broad, BIT definition of ‘investment’ gives shareholders in a company the ability to bring a claim on their own behalf.Footnote 114
To sum up, we have two key examples of a desire not simply to displace customary norms but to tailor the resultant treaty protections to specific concerns surrounding the entry and operation of foreign investors in host States. This is not to say that all the background elements of custom are displaced in the newer treaty instruments. Certain treaty instruments expressly link the obligation to accord fair and equitable treatment to the broader corpus of international law.Footnote 115 This textual link evidences an intention to incorporate customary international law (on the minimum standard of treatment to be afforded to aliens) in the treaty obligation. State parties though are adopting a ‘pick and choose’ strategy when it comes to incorporating customary law into the new treaty protections. The choice to do so appears largely an express one. Absent this express link, other treaty components (such as expropriation and dispute settlement rules) clearly oppose customary law and offer tailored standards of operation.
With this context in mind, there are good reasons to suggest that the treaty exception also offers a tailored defence distinct from the customary norm. To put this differently, the process of tailoring was not only designed to offer commitments to restrain State conduct but, in certain limited circumstances, to allow for newer flexibilities on the part of the State.Footnote 116 First, there is no clear textual link to customary law in the treaty exception. Where State parties had intended for such a linkage to occur they had proven themselves perfectly capable of doing so expressly in other parts of the treaty. More fundamentally, the structure of the new treaty exception contemplates the priority of the post-war institutional structure of the Charter of the United Nations. The State parties envisioned a new ground for excuse that falls entirely outside the scope of the customary defence. Similarly, the treaty exception includes a ‘public order’ component, which does not find direct reflection in customary law. These factors are strongly probative of an objective desire to expand the scope of the treaty exception over and above the customary plea. Perhaps most tellingly, the customary provision supplies an operative test for when a measure will be ‘necessary’ while the treaty standard is silent on the choice of means–end inquiry. This precise difference between a customary and treaty standard was relied upon by the ICJ in the Nicaragua case in finding that the two sources of law in that case did not overlap and continued to exist alongside each other.Footnote 117
C. Motivations: Insights into the Sociology of Investor–State Arbitration?
Given the pointed differences between the treaty exception and the customary norm, there is the obvious question as to why the arbitrators would choose to conflate the two. This is a difficult question to answer, as the pleadings and transcript of proceedings in these cases are not publicly available.Footnote 118 A close analysis of the awards themselves though reveals a set of intriguing clues as to the possible motivations at play in the adoption of this flawed methodology. These offer in turn crucial insight into the unique sociology of investor–State arbitration,Footnote 119 a further factor which should inform our choice on other interpretative methods of reading the treaty standard vis-à-vis the customary norm.
1. Argentina's litigation strategy: the claim of auto-interpretation and its implications
First, the preparedness of the tribunals to accept confluence may have been shaped by Argentina's litigation strategy. Argentina's defence, as revealed in the awards, places enormous emphasis on the marginal claim that the treaty exception is self-judging.Footnote 120 This is an argument that would require an adjudicator to show significant deference to the State invoking the exception. To put this slightly differently, the argument would largely exclude any appreciable role for an adjudicatory tribunal. It falls logically at an extreme end of a spectrum of the role for an adjudicator in assessing whether the constituent elements of the treaty exception have been proven.
The risk of exclusion implicit in Argentina's case should be seen in light of the likely personal dimensions of arbitral adjudication. Appointment to an ICSID arbitral tribunal remains a prestigious, even crowning achievement for practitioners and others working in the field. Once appointed, there is a natural desire to show one's professional and analytical wares, so to speak.Footnote 121 This is by no means confined to investment treaty arbitration and may even offer important systemic checks in certain settings. Consider the prevalent practice of the individual justices of the ICJ issuing separate rather than concurring judgments.
Argentina's strategy of claiming auto-interpretation would, at its most fundamental, relegate the role of arbitral members to that of bystanders in the adjudicatory process. The unappealing nature of this argument may have made the individual members of the tribunals more receptive to arguments raised by the investor at the other end of the spectrum of possible readings on the treaty exception. The investor will have logically emphasized the claim for confluence, as the restrictive conditions of the customary norm make it highly likely than the treaty exception will not apply. The tribunals are then effectively presented with a stark choice; simple deference to the State or a distinct and public role for the adjudicator in assessing defined components of the customary-treaty exception. Viewed in this light, the choice of the latter is perhaps unsurprising.Footnote 122
2. Internal and external dynamics of arbitral decision-making
The loose preference accorded to the customary exception may also reflect the particular dynamics of investment treaty arbitration. There are two characteristics of this system that deserve close attention. First, there is an internal dimension in that all three tribunals were instituted under the ICSID Convention. The ICSID rules offer a specialized forum for resolution of investment disputes, but this remains a system of arbitration with all the unique practices, habits and language that characterize that form of adjudication.Footnote 123 At its most fundamental, arbitration is a dispute resolution system that prioritizes and values party autonomy, speed and finality over correctness in adjudication.Footnote 124 The crafting of an outcome (and its likely acceptance among the disputing parties) has traditionally been regarded as far more important than the rigour of the process of legal reasoning and justification. This emphasis alone may go some way in accounting for the simple reasoning adopted by the method I tribunals.
There is a second, external dimension which merits consideration. ICSID rules do not provide for a standing body to adjudicate in investment disputes. Instead, a three member ad hoc tribunal is established with the parties each choosing a member and with a tribunal's President appointed by agreement between the parties.Footnote 125 This ad hoc structure offers no guarantee that the arbitrators have worked together in the past or indeed, are necessarily experts in international law. This can itself translate into a striking, even understandable tendency to defer to ‘neutral’ and authoritative sources such as the ILC articles (as in the confluence approach) where adjudicating on an untested treaty norm.Footnote 126 That deference can be even more pronounced when we consider that an arbitrator's behaviour may influence whether or not they are chosen for future arbitrations.Footnote 127 A tribunal that appears to ‘throw deep’, offering a new unexpected reading of a treaty provision, could significantly harm its prospects in the external market for future appointments. Anchoring a reading to the respected and cited work of the ILC offers a counter to an impression of adjudicatory adventurism, thereby preserving reputational credibility in the appointments market.Footnote 128
C. Arbitrator Assumption as to the Telos of the Subject Treaty
Finally, and most problematically, the stringent reading implicit in the confluence methodology seems to have been chosen as representing a personal sense of the arbitrators as to how the regime should work. Here we have clear evidence from the text of the awards as revealed in this statement of the Enron Tribunal:
[T]he object and purpose of the Treaty is, as a general proposition, to apply in situations of economic difficulty and hardship that require the protection of the internationally guaranteed rights of its beneficiaries. To this extent, any interpretation, resulting in an escape route from the obligations defined cannot be easily reconciled with that object and purpose. Accordingly, a restrictive interpretation of any such alternative is mandatory.Footnote 129
The order of this reasoning reveals a fundamental flaw. The Enron Tribunal begins by asserting its own claim as to the dominant purpose of the treaty, being one of protection of the rights of investors. This remains a key, untested assumption in the jurisprudence of investor–State arbitral tribunals.Footnote 130 The Tribunal then fashions a default, restrictive interpretative preference to comply with that claimed purpose.
This method ignores the rules and sequencing of steps on treaty interpretation at international law. Under article 31(1) of the Vienna Convention on the Law of Treaties, an adjudicator must begin with the text of the treaty provision in question.Footnote 131 A given textual reading should then be considered in light of context and the object and purpose of the treaty under review.Footnote 132 The primary role of the treaty text in this setting has an important disciplinary function as it prevents the adjudicator from preferencing their own intuitive sense of how the system should work over objective expressions in the treaty text.Footnote 133 The Enron Tribunal subverts this order of interpretation, bypassing the text as the start-point in the interpretative process. Moreover, the stubborn claim in the jurisprudence to a dominant purpose of investment protection is itself open to legitimate contestation. On close examination, the objective expressions of purpose in the preambular recitals of a typical investment treaty reveal a concern for the interests of multiple stakeholders and values beyond an exclusive focus on the protection of foreign investors.Footnote 134 Indeed, other cases within the system have begun to adopt more rigorous approaches to treaty interpretation and have identified and grappled with the multiple values referenced in the preambular recitals of investment treaties.Footnote 135
IV. METHODOLOGY II: THE TREATY EXCEPTION AS LEX SPECIALIS
As we have now seen, the dominant methodology employed in the case law conflates the status and scope of operation of the customary and treaty standards. The tribunals employing this first methodology do not, as the CMS Annulment Committee later criticizes, take a clear and rigorous position on the relationship between the two legal standards.Footnote 136
There is another method touched upon in the jurisprudence that offers a plausible account of that relationship. This second methodology would begin from the premise that the customary standard (formalized in ILC article 25) is a primary legal rule that goes to the determination of breach, operating on the same legal plane as the treaty exception. Unlike methodology I however, it offers an approach by which an adjudicator can give effect to the clear differences between the two legal standards. Method II does this by prioritizing the treaty exception as an expression of lex specialis, constituting a specific elaboration or updating of the general customary norm.
A. Lex Specialis and ILC Article 55
The lex specialis principle finds reflection in the ILC ArticlesFootnote 137 and seems to be the favoured approach of a number of commentators in the field.Footnote 138 For it to apply however, it is not enough that the same subject matter is covered by the treaty and customary standards. There must be some actual inconsistency, the identification of which remains a question of interpretation.Footnote 139 If we accept the premise of the customary standard as primary legal rule, there are key differences between the two norms, especially the inclusion of the public order exception in the treaty provision. We have then the sort of inconsistency that might allow for the application of the lex specialis principle.
B. Scope of Priority
The identification of the treaty defence as lex specialis is only a starting point. There is still the question of the scope of priority to be accorded to the treaty defence, a critical issue largely ignored by those sympathetic to this reading.Footnote 140 Here two alternative approaches may be taken, with vastly different implications for the operation of the treaty exception.Footnote 141
1. The residual operation of custom: the Oil Platforms approach
The first is that endorsed by the ILC articles, which provide that the customary or general norm will continue to operate in a residual fashion.Footnote 142 This would see the general norm displaced only to the extent of any inconsistency, leaving other aspects of the general law still applicable.Footnote 143 The ILC Study Group on the Fragmentation of International Law presents the interpretative link in the following terms:
[T]he point of the lex specialis rule is to indicate which rule should be applied …[T]he special, as it were, steps in to become applicable instead of the general. Such replacement remains, however, always only partial. The more general rule remains in the background providing interpretative direction to the special one.Footnote 144
The ICJ applied this approach in the Oil Platforms caseFootnote 145 which has also found endorsement in a recent ICSID award.Footnote 146 In Oil Platforms, the US sought to rely on an exception for necessary ‘essential security interests’ as a defence to Iran's claim that US military attacks on Iranian oil platforms breached the operative provisions of the treaty.Footnote 147 Reminiscent of Argentina's plea for auto-interpretation, the US suggested that the Court should afford a ‘measure of discretion’ to a party's good faith application of measures to protect essential security interests.Footnote 148 The ICJ rejected the idea that the treaty exception was intended to operate independently of relevant rules of international law on the use of force.Footnote 149 In particular, it applied the customary rules on self-defence to give content to the treaty standard that a given measure is ‘necessary’ for essential security interests.Footnote 150 As the test for necessity in those customary rules is both ‘strict and objective’, it left no room for the US's proposed measure of discretion.Footnote 151
The Oil Platforms approach to the lex specialis principle has important implications for arbitral tribunals constituted to cases involving financial crisis. The site of inconsistency between the customary and treaty standards at issue here goes to the articulation of permissible ends. There is however a degree of overlap between the standards that cuts across this difference; both endorse ‘necessity’ as the legal standard for when a measure should be taken to have achieved a particular end. With this overlap in mind, the stringent customary test of ‘necessity’—that a measure be the ‘only way’ to effect a given end—could apply in a residual fashion to give content to the treaty provision.Footnote 152
2. The displacement of custom in its entirety: the UPS v Canada approach
As an alternative to the Oil Platforms approach, there is the possibility that the instrument of lex specialis could be taken to exclude the general law in its entirety. This was the approach of the NAFTA Chapter 11 Tribunal in UPS v Canada.Footnote 153 In UPS, the investor had argued that particular conduct of Canada Post, a State entity created by statute with a monopoly on the delivery of certain letter mail, should be attributed to Canada using the test for attribution in ILC articles 4 and 5. The UPS Tribunal rejected the idea that the ILC articles could be used in this way and instead concluded that specific provisions elsewhere in the NAFTA—Chapter 15 that regulates the conduct of monopolies and State enterprises—constituted lex specialis and applied to the exclusion of the ILC articles.Footnote 154 The correctness of the substantive reasoning adopted by the UPS Tribunal is not without question. One might expect an adjudicator to find some clear evidence of intent to fully displace the more general norm. After all, contracting out of custom should be clear and unambiguous.Footnote 155 The UPS Tribunal instead looks at an entirely separate chapter of the NAFTA and different (although at times similar) commitments to suggest that anything to do with attribution of State enterprises is cabined in Chapter 15. The UPS ruling though remains a possible (although far less convincing) approach to the scope of priority to be accorded to the special treaty norm.Footnote 156
C. The Puzzle of the LG&E Award
Our discussion so far has been relatively abstract by sketching the interpretative possibility and implications of reading the treaty exception as an expression of lex specialis. This method has found reflection in the jurisprudence, albeit with some uncertainty and internal contradictions. The best candidate for a lex specialis approach is LG&E v Argentina, although there are intriguing hints in this direction in the Continental award as well.Footnote 157 Unlike the tribunals that adopt the confluence methodology, the LG&E Tribunal begins its analysis with the treaty exception and also finds expressly in favour of Argentina on the treaty exception (albeit within a defined time period).Footnote 158 In fact, the Tribunal seems to be giving express effect to the specialized public order component of the treaty exception.Footnote 159 The analysis of the treaty provision though is not set in isolation as much of the Tribunal's reasoning draws on customary law.Footnote 160 We are thus in the realm of this second methodology; although both the customary and treaty standards cover a similar subject area, the treaty provision is being applied as an expression of lex specialis.
We might then expect the LG&E Tribunal to draw on the customary standard to assess whether Argentina's chosen means were ‘necessary’ as the only way to maintain public order or to protect essential security interests. However, the Tribunal does not use the ILC articles in this residual fashion but instead adopts its own rather cursory approach to ‘necessity’. It is difficult to understand the precise test applied to assess the necessity of Argentina's chosen means. The Tribunal characterizes Argentina's response as legitimate without explaining why this is the case.Footnote 161
By adopting its own unprincipled approach to the test for ‘necessity’, the LG&E Tribunal may be simply excluding the customary standard in its entirety. In other words, the Tribunal is choosing to follow the UPS Tribunal's hard contracting out approach rather than that of the ICJ in Oil Platforms. The problem though is that the LG&E Tribunal continues its analysis and finds, in a thoroughly unconvincing fashion, that the customary provisions would also excuse Argentina from liability. The strangest part of this later analysis is the summary finding that Argentina's measure (now characterized as an ‘economic recovery package’) was not just legitimate (as concluded under the treaty exception) but ‘the only means to respond to the crisis’.Footnote 162
In sum, the LG&E award is a candidate for a lex specialis approach but it is not a convincing one. The problem lies in the internal contradiction of that award. If the treaty exception is an expression of lex specialis, then an adjudicator should determine the scope of priority to be accorded to the treaty text. This would mean either applying the customary norm in a residual fashion (where not in conflict) or displacing it entirely. The LG&E Tribunal appears reluctant to choose between these approaches, perhaps recognizing the strength of the former (within the internal logic of the lex specialis method) but also its implications (the likely preclusion of most claims to invocation). Its own poorly conceived test of ‘legitimacy’ appears driven more by pragmatism than principle and will hardly aid in guiding successive tribunals in this difficult area.
V. METHODOLOGY III: SEPARATING PRIMARY AND SECONDARY APPLICATIONS
A. The Taxonomy
There is a third methodology that overcomes the deficiencies in the approaches we have considered so far. This would begin by characterizing all of the treaty provisions at issue, both forms of obligation and exception, as primary legal standards that determine whether a State has committed a wrongful act at international law. It is only if breach is determined by the composite application of these rules that an adjudicator would examine the secondary possibility of the customary defence of necessity to preclude the finding of a wrongful act.
This approach finds some welcome reflection in the recent award in Continental v Argentina.Footnote 163 This follows its endorsement by the ICSID Annulment Committee that examined Argentina's claim for annulment of the CMS award:
[Article XI] specifies the conditions under which the Treaty may be applied, whereas [ILC] Article 25 is drafted in a negative way: it excludes the application of the state of necessity on the merits, unless certain stringent conditions are met. Moreover, Article XI is a threshold requirement: if it applies, the substantive obligations under the Treaty do not apply. By contrast, Article 25 is an excuse which is only relevant once it has been decided that there has otherwise been a breach of those substantive obligations.Footnote 164
By separating primary and secondary applications, method III obviously follows the preferred taxonomy of the ILCFootnote 165 and has echoes in the ICJ's approach in the Gabčíkovo-Nagymaros Project (Hungary v Slovakia) case.Footnote 166
A reader at this stage may wonder as to the practical difference between this reading and a hard version of lex specialis (à la UPS v Canada), that would situate the treaty as contracting out of custom. There is little in outcome that will distinguish the two readings. But there may be enormous gains in the message it communicates to adjudicators and other stakeholders in the system. Method III avoids a blunt claim that some might be tempted to make if drawn to the lex specialis account; that the investment treaty project was designed to exclude most if not all customary norms. The reality is far more complex. The treaty project was structured to offer a tailored system of legal regulation of the conduct of States vis-à-vis foreign investors which incorporated select customary norms and also constructed new treaty norms with a different tenor and direction from the protections at customary international law. As I have suggested, we find distinct textual and contextual evidence of tailoring of flexibility in the treaty project. Method III reflects this evidence and locates our analysis not on the general plane but on the particular, being the relation between the individual treaty exception and the customary plea of necessity.Footnote 167 After all, an adjudicator on these questions is ultimately engaged in a process of treaty interpretation where, as Campbell McLachlan has suggested, the treaty text should fall at the centre of the enquiry.Footnote 168
This method also offers distinct doctrinal and strategic advantages, bearing in mind the sociological features of investor–State arbitration examined earlier. It obliges an adjudicator to give effect to both the treaty exception and the customary defence, contrary to the outcome of the blunt confluence reading of method I.Footnote 169 More subtly, it enables an adjudicator to avoid the temptation implicit in the lex specialis reading that remnants of the customary norm will simply continue to control the application of the treaty provision, rendering it all but inutile. Reviewing the jurisprudence to date, it is difficult to escape the sense that many arbitrators remain hard-wired to the customary test. Method III imposes a disciplinary boundary on an adjudicator by characterizing the treaty exception as a first-order primary norm. On the other hand, there may be a criticism that method III will result in redundancy of the customary defence. The argument could run as follows: If analysis of the treaty exception occurs first and assuming its scope of operation is broader than the customary defence, then invocation of the clause will always preclude custom. However, this argument ignores the different spheres of operation of these two sources of international law. Custom applies across all the various areas of international law. Many of these subject areas are not governed by treaties or where they are, do not have a specific preclusion clause such as article XI of the US–Argentine BIT.Footnote 170 Where a legal dispute arises in these other areas, custom remains the mechanism by which the final question of State responsibility is assessed.Footnote 171
Finally and perhaps most importantly, there is a critical normative dimension to this methodology. This is an area of international law that has become increasingly unstable as State parties exercise both ‘voice’Footnote 172 and ‘exit’Footnote 173 vis-à-vis the investment treaty regime.Footnote 174 These actions have been triggered by the enormous take-up of arbitral dispute settlement in the last ten years.Footnote 175 It is not just the invocation of the system that is driving these changes; the manner in which the substantive rules have been interpreted is also part of the matrix. The evolving case law has caused serious assessment among State parties as to whether the ‘classic’ model of investment treaty, constructed in the furnace of post-Second World War contestation between developed and developing States, places too strong a constraint on regulatory autonomy. It is however enormously difficult as a matter of pragmatics and cost not only to amend the thousands of existing bilateral instruments in operation but to do so in a manner that anticipates every possible future contingency.Footnote 176 Investment treaties in this respect might represent a classic embodiment of the problem of incomplete contracting in economic theory.Footnote 177 This raises a key role for the adjudicator to apply existing rules in such a way as to save a contract or treaty from endemic uncertainty (and to ensure continued participation/loyalty of the parties). This could involve a more sophisticated use of the customary rules of treaty interpretation to offer contextual rulings on substantive obligations (such as the fair and equitable guarantee) to better balance the interests of key stakeholders. But for hard cases—the rare but immediate occurrence of financial crisis—express clauses that directly allow for derogation and hence flexibility will be critical in the ability of the adjudicator to adopt this broader role. Method III, by divorcing the treaty exception from the stringent customary plea, offers the adjudicator an important avenue to take up this challenge.
B. Key Interpretative Issues
My preferred method III raises three key interpretative issues: (i) the identification and scope of ‘public order’; (ii) the identification and scope of a State's ‘essential security interests’; and (iii) how to test when a chosen measure is ‘necessary’ to achieve these purposes. My intent here is not to offer simple conclusions to this complex set of issues. I hope instead to present a first-step framework to prompt and frame further analysis and discussion.
1. ‘Public order’ and its adjudication in the WTO
The notion of public order as a defence or qualification to international legal rights exists in other systems of international law. For example, article XIV(a) of the WTO General Agreement on Trade in Services enables WTO member States to take measures ‘necessary to protect public morals or to maintain public order’.Footnote 178 The framers of the GATS provide some general textual instruction to the scope of the public order exception in footnote 5 to article XIV(a):
The public order exception may be invoked only where a genuine and sufficiently serious threat is posed to one of the fundamental interests of society.Footnote 179
This textual direction engages the requisite degree of severity of the threat. The question remains as to how one is to identify the fundamental interests of society that might trigger recourse to the public order exception.
In the Internet Gambling case, the United States had sought to defend its domestic restrictions on Internet gambling under the GATS article XIV(a) exception.Footnote 180 Specifically, the United States argued that the practice of Internet gambling attracted elements of organized crime with resulting threats to both public order and morality.Footnote 181 Adopting a method of dynamic interpretation, the WTO Panel accepted that there is no single meaning of notions of ‘public morals’ and ‘public order’ as these ‘can vary in time and space, depending upon a range of factors including prevailing social, cultural, ethical and religious values’.Footnote 182 On this basis, the Panel was prepared to accord some sensitivity to the applicable WTO member in defining and applying those terms according to their own systems and scales of values.Footnote 183 The sensitivity though was not absolute given the obligation on the Panel to give effect to those terms as a matter of treaty interpretation.Footnote 184
Using the well-worn path of resort to dictionary definitions in the jurisprudence of the WTO, the Panel in Internet Gambling eventually adopted a useful categorization of the distinct concepts of public morals and pubic order. Public morals were taken to denote ‘standards of right or wrong conduct maintained by or on behalf of a community or a nation’.Footnote 185 Public order in turn was seen as a distinct concept directed to the preservation of the fundamental interests of a given society, which would include the maintenance of the rule of law.Footnote 186 The Panel then applied this loose but workable categorization to the facts before it. A policy of restricting Internet gambling in order to prevent underage gambling and the protection of pathological gamblers would relate to public morals, while the fight against organized crime would be matter of public order.Footnote 187
The Panel's approach to the interpretation of ‘public order’ reflects the origins of that juridical concept in many civil law systems. The notion of ordre public in French law encompasses the collection of conditions—legislative, departmental and judicial—that preserve the normal and regular functions of the State.Footnote 188 In short, ordre public is directed to the preservation of rule of law in the State concerned. Viewed in this light, certain of the key disruptive effects of the Argentine financial crisis, including incidences of widespread rioting, looting and executive instability, might fall within this understanding of ‘public order’. This is essentially the interpretative approach taken by the Continental Tribunal. That Tribunal clearly gives independent operation to the ‘public order’ ground of exemption, separate from the exclusion for ‘essential security interests’.Footnote 189 It also accepts ‘public order’ as a synonym for ‘public peace’ (drawing on the Spanish text of the BIT at issue), able to be threatened by actual or even potential insurrections, riots and violent disturbances of the peace.Footnote 190 This interpretative approach does not necessarily excuse Argentina's liability given critical questions surrounding burden and standard of proof, temporality and the appropriate test of means-end scrutiny (all of which are considered below). We have though, as a start point, an interpretative prism to suggest that disruption that threatens the normal functioning of a State can engage a State's ‘public order’ concerns.
2. Human security and the evolving notion of ‘essential security interests’
A more difficult question surrounds the scope of a State's ‘essential security interests’ in the BIT exception. In particular, there is the charged issue of whether this concept is constrained by traditional and static notions of security understood as the ability of the State to counter external threats to its territorial integrity.Footnote 191
The text itself is not limited to ‘national’ security interests, which tends to be the defining approach in certain other derogation clausesFootnote 192 and indeed in other parts of the BIT in question.Footnote 193 Most of the ICSID tribunals offer perfunctory analysis before simply concluding that a State's essential security interests could extend beyond the classic instruments of high politics to encompass financial crises.Footnote 194 Rather than essentially inventing an answer to this problem, an adjudicator should apply the rules on treaty interpretation to test the boundaries of this concept. Article 31(3)(c) of the Vienna Convention on the Law of Treaties obliges an interpreter to take into account ‘any relevant rules of international law applicable in the relations between the parties’.Footnote 195 This article embodies a more general principle of treaty interpretation which Campbell McLachlan has insightfully described as ‘systemic integration within the international legal system’.Footnote 196 Thus understood, article 31(3)(c) enables a tribunal to interpret and apply a treaty instrument in relationship to its normative environment, ‘other’ international law.Footnote 197
The Continental Tribunal makes a step, if a little bluntly, in this direction. It draws on the Preamble of the UN Charter as support for the inclusion of domestic economic concerns within a State's essential security interests.Footnote 198 There are more targeted external sources that can be used to shed light on this question. In particular, the work of the UN Commission on Human Security focuses on the direct impact of financial crisis on the security concerns of States and their citizens.Footnote 199 That Commission argues that traditional notions of State security must be augmented by an express concern for what it terms ‘human security’. The driver is a recognition that the contemporary State is no longer able to act as the sole purveyor of security to its people. The challenges to security are now multi-faceted and encompass events often far beyond State control, including risks of external pollution, terrorist attacks and water shortages.Footnote 200 These changing risks require a new paradigm of ‘human security’ not as a replacement of State security but as a complementary condition.Footnote 201
The Commission specifically isolates financial crises as a downside risk that regularly impairs human security.Footnote 202 The consequences of financial crises—shrinking output, declining incomes and rising unemployment and sharp increases in income poverty—are borne disproportionately by the most vulnerable members of the community.Footnote 203 These adverse impacts on human security require steps not only to prevent but also to mitigate the effects of crises. Mitigation requires the adequate provision of social protection in the form of health care, food, shelter, water and income support.Footnote 204 The problem though is that precisely encountered by Argentina; the ability of a State to finance social protection is at its weakest when crises compound.Footnote 205
This analytical framework suggests that outbreaks of financial crises might engage security concerns (defined to encompass human security objectives), thereby triggering the operation of the BIT exception. There is an obvious objection to the analysis sketched so far. I began this part of the article by searching for ‘relevant rules of international law’ to guide an interpretation of the BIT exception. A report of a consultative body to the UN could hardly be said to constitute a formal source of international law. There is though a discernible link between the UN Commission's study and the corpus of human rights conventions, especially the International Covenant on Economic, Social and Cultural Rights (ICESCR).Footnote 206 For the Commission, there is a complementarity such that: ‘Human security helps identify the rights at stake in a particular situation. And human rights helps answer the question: How should human security be promoted?’Footnote 207 The State parties of the ICESCR (including Argentina) are obliged to ensure their citizens are provided with the right to an adequate standard of living (including food, clothing and housing)Footnote 208 as well as the right to the highest attainable standard of physical and mental health.Footnote 209 While State parties may progressively realize these rights,Footnote 210 they are under an immediate obligation to ensure the satisfaction of ‘minimum essential levels’.Footnote 211 Where the very ability of a State to meet these continuing and elementary needs of citizens is endangered through financial crises, human rights law offers a critical anchor to justify a claim to engagement of ‘essential security interests’.Footnote 212 We find, once again, intriguing hints in this direction in Continental where the Tribunal expressly draws on a concept used in human rights law in suggesting it would accord a State ‘a significant margin of appreciation’ when it came to applying emergency measures.Footnote 213
This analysis may be criticized as suggesting a broad and overly permissive approach to the invocation of the treaty exception, with the potential for opportunistic abuse by State parties to investment treaties. It is important, however, to recall the exact stage of the analysis. So far, I have only sought to offer two distinct prisms through which the downside effects of financial crisis might engage a State's ‘public order’ and ‘essential security interests’ under the treaty exception. This does not complete the analysis nor necessarily excuse Argentina's liability.
There are at least three further conditions to a successful treaty defence when applied to a given fact dynamic. First, the burden of proof should rest on the invoking State (as the better informed party) to adduce objective evidence of the requisite elements of disorder and/or inability to meet key human rights obligations. That general burden though is distinct from the requisite standard of proof.Footnote 214 The latter concerns the quantity and nature of evidence necessary for a party to persuade the adjudicator of the correctness of their position and as such, to discharge its overall burden. These are not simply technical questions but instead impact on the very incentives of parties to litigate and defend particular claims. At least on the question of ‘essential security interests’, it would seem reasonable to require a State to justify its invocation of human rights law by offering expert validation in the form of an appropriate determination of a UN human rights body. Second, there is an important temporal limiter. The invocation of the treaty defence should only last as long as these objective conditions allow as to do otherwise would build in a range of perverse incentives. On this critical point, the Continental award represents a lodestar in the jurisprudence; that Tribunal finds against Argentina on only one measure (a debt swap) and does so because of its imposition in 2004 ‘when Argentina's financial conditions were evolving towards normality’.Footnote 215 Finally and most importantly, there is the overarching test for determining whether the State's actual measures were ‘necessary’ to the achievement of those objectives. It is in this part of the inquiry that an adjudicator has a critical role to play in checking for abuse of the invocation of the exception.
3. Necessity: less restrictive means analysis and/or proportionality review
The necessity component of the treaty exception asks a question of the closeness or fit between the chosen means and the asserted regulatory purpose of the State in question. There are though various methods of engaging in means-end inquiry. It is possible to discount one candidate at the outset; the customary test that the chosen measure be the ‘only way’ for the State to achieve its asserted goal. Given that the treaty exception under our preferred methodology III acts as a primary norm, there is no obligation on an adjudicator to adopt this exceedingly strict secondary test.
In a recent paper, William Burke-White and Andreas von Staden offer another possible method of means-end inquiry. Drawing on the margin of appreciation doctrine in the jurisprudence of the European Court of Human Rights, the authors suggest:
[The European Court of Human Rights] will, in a first step, review whether the impugned measure can be characterized as pursuing such a legitimate objective. In doing so, the court assesses the measure in an abstract way and will accept the domestic legislature's ‘judgment of what is “in the public interest” unless that judgment be manifestly without reasonable foundation.’ The result, then, is a ‘wide margin of appreciation’. In a second step, however, the Court will inquire whether the operation measures chosen to achieve the legitimate objective in question meet a ‘fair balance’ test: that test is failed if the individual claimant, in light of the objective pursued has to bear an ‘excessive’ or ‘disproportionate’ burden …. The transfer of such a two-stage review process to BIT NPM [Non-Precluded Measure] clauses would be straightforward: In the first step, an arbitral tribunal would determine whether a measure's purpose falls reasonably within one of the permissible objectives, with the invoking state being granted an appropriate margin. If the purpose is legitimate, then the tribunal would in a second step weight the public interest pursued against the burden imposed on the foreign investor and determine whether the latter were proportional.Footnote 216
My analysis shares some commonality with parts of the Burke-White and von Staden thesis. I have argued that an adjudicator should exhibit institutional sensitivity on the question of the engagement of permissible objectives in the treaty exception. This might be in relation to the respondent State (say on disruption to ‘public order’) but it could also engage the competence of an expert human rights body (especially on a human security understanding of ‘essential security interests’). However, I depart from those authors in their endorsement of the ‘second step’. They are suggesting a particular form of proportionality review in testing whether a chosen measure is ‘necessary’ for the asserted regulatory objective.Footnote 217
Proportionality assessment in a range of constitutional and treaty systems is commonly understood to involve at least three steps.Footnote 218 The first—‘rationality review’—merely requires the adjudicator to examine whether the means adopted by the State are rationally related (usually by making a contribution) to the objective in question. The second test—‘less restrictive means’(LRM) analysis—is more stringent. The adjudicator here examines whether there is any alternative measure that achieves the same level of benefit but is less restrictive of the right in question. The last stage—‘balancing’ or ‘proportionality’—is significantly different as it requires an adjudicator to determine whether the costs imposed by the measure are excessive or disproportionate to the benefits of the policy objective. The inquiry now focuses on the regulatory objective itself. This can be abandoned depending on the adjudicator's own assessment of the regulatory objective against attendant costs.Footnote 219 It is critical that we recognize that this presented framework offers a nested sequence of tests. Each test invalidates everything invalidated by its predecessor and something more.Footnote 220 Where this analytical framework has been adopted in individual legal systems, it is not uncommon for courts to dispose of most cases under the LRM test and rarely move to the final balancing stage.Footnote 221
This full and complex analytical framework is not what the proponents of ‘proportionality review’ in the investment treaty setting have in mind. They are advocating for the elevation only of the third (balancing) stage. It is important to appreciate the institutional choices inherent in this normative claim. The judicial organ becomes responsible for assessment and weighting of relative values rather than national legislatures. Yet this form of weighting often involves complex value-laden and empirical judgments. It is highly doubtful that courts, in general, are better assessors of values and empirical questions than elected representatives.Footnote 222 Lawyers are far better positioned to adjudge deficiencies in process rather than such complex decisions.Footnote 223
It is true that proportionality review has found reflection in the jurisprudence of various European courts. Once again though, it is usually preceded by the earlier analytical steps outlined earlier and as such, these would seem to be rational candidates for consideration in the investment treaty setting. Moreover, it is important to bear in mind the strategic imperatives that have driven the affirmation of greater judicial authority in the European context and as such, the limited justifications for a transplant to the investment treaty context. For example, Miguel Maduro has argued that the elevation of this doctrinal framework to article 30 of the 1957 Treaty Establishing the European Economic Community has been carefully chosen as part of a broader normative project. Maduro has presented the increased judicial activism of the European Court of Justice in applying this test as reflecting a normative goal of positive integration or ‘market building’ in the European context.Footnote 224 The Court's increased judicial activism is in turn a response to the failure of the legislative arms of the European political process to implement this broader goal.Footnote 225
The choice then of this test in part of the European acquis is explicable due to very specific factors; the existence of a normative goal (positive integration) and an institutional malfunction (the failure of the European legislative process to implement that goal). Needless to say, neither of these directives is even remotely present in the loose network of international investment treaties. We should, as a result, exercise considerable caution in cross-regime citation and reliance.Footnote 226 Moreover, it is important to bear in mind the broader implications of the form of dispute settlement that characterizes this area of international law. There is no appellate mechanism that—properly constructed—might operate to discipline unprincipled judicial activism under the third stage of a proportionality framework. This is a critical constraining factor given, as we have seen, the particular sociology of investor–State arbitration (and especially its intuitive claim to an exclusive telos of investment protection).
With this in mind, there is considerable merit in proceeding cautiously. One logical possibility is that we start with a variant of LRM analysis, which is in fact the test suggested most naturally by the word ‘necessary’ in the treaty exception. LRM analysis offers a critical advantage over balancing/proportionality, where the latter stands as a first and final inquiry by the adjudicator. In an LRM test, the adjudicator does not pass judgment on the relative importance of the goal chosen by the invoking State. Her assessment focuses only on: (i) whether there is an alternative measure available to the State that achieves the same level of benefit as the chosen measure, where (ii) the alternative measure results in less restrictive effects on foreign investment. Where such an alternative exists, then the State's chosen measure will not be ‘necessary’ and the State may not rely on the treaty defence. It is important to note that an alternative measure will only be assessed under this test if it achieves the same level of benefit as the chosen measure. There is then no possibility that the adjudicator can substitute her own judgment for that of the State in the appropriate level of benefit to flow from the chosen measure. Of all the various cases constituted to date, it is only the Continental Tribunal that evinces a sophisticated understanding of the more modest role implicit in an LRM test.Footnote 227
There are, however, legitimate criticisms of the LRM test that still require serious consideration. The most fundamental goes to the issue of how an adjudicator is to treat the cost of the alternative measure vis-à-vis the chosen measure. The CMS Tribunal for example, identified a variety of alternatives to Argentina's chosen measure, including ‘the granting of direct subsidies to the affected population’.Footnote 228 In a simple LRM test, the adjudicator would look to whether the alternative measure (subsidization of consumers of utility services) achieves the same level of benefit at less restriction to foreign investment than the chosen measure (abrogation of investor rights under utility contracts). The logical outcome of this analysis is that the State would always be required to bear the costs of the alternate, less restrictive option (of subsidization of utility services). On this hypothetical at least, the significant costs of subsidization would make this a highly tenuous alternative given likely real world, budgetary constraints on the part of an invoking State.
There is another possible version of LRM analysis that offers a more sophisticated treatment of the differential costs of opposing alternatives. Such an approach would not only identify a less restrictive alternative but assess whether it is a reasonable alternative given the different costs involved. To be precise, the State's chosen measure will be necessary if every alternative measure achieves the same level of benefit at less restriction but requires unreasonable administration and enforcement costs. Under this reasonable LRM test, a tribunal could then evaluate the cost and time delays that would be incurred by a State in adopting the alternate course of action. The notion of a reasonable LRM test is not without precedent and has found reflection in the case-law of the WTO.Footnote 229
A reader at this stage may wonder whether a reasonable LRM test collapses into the free-standing test of proportionality review criticized earlier.Footnote 230 It is true that a reasonable LRM test involves some form of balancing. The type of balancing is though very different to full-blown balancing or proportionality. In the latter, the adjudicator directly balances the benefit of the goal as against the restriction to foreign investment. In contrast, the reasonable LRM test does not at any point authorize an adjudicator to judge the relative importance and benefit of the goal. She is only entitled to balance the different enforcement cost with the restriction on foreign investment.
There may be some who regard the reasonable LRM test as too deferential to a regulating host State. This seems a misplaced concern as we have only assessed the test against one somewhat extreme example; whether subsidization of consumers might constitute a reasonable less restrictive alternative to Argentina's chosen measure of abrogation of the contractual rights of foreign investors operating utility concerns. The idea that the full costs of responding to the escalating costs should be borne entirely by the State and by extension, its citizens seems an unreasonable outcome especially given the absence of any evidence of discrimination directed at foreign economic actors. On the other hand, there are other possibilities that might constitute reasonable less restrictive alternatives to simple abrogation of investor rights.
One possibility was raised in passing in the LG&E award. The LG&E Tribunal noted that Argentina's measure—suspension of contractual rights of foreign investors—was applied ‘across the board’ rather than through individual assessment of particular utility contracts.Footnote 231 Individual assessment would presumably entail the assessment of whether rights under a particular utility contract should be abrogated given their likely contribution to the continuation and scope of the crisis. A contractual right that would entitle a foreign investor to increase the tariff rate for electricity or gas supply might be treated differently to a contract involving the provision of telecommunications services. There could be a sensible case for abrogation of the latter but retention of the former. In this case, individual assessment is clearly a less restrictive alternative to the ‘across the board’ measure; it offers review of each case on its own merits and the possibility that some public utility contracts may escape abrogation of rights. Moreover, the enforcement cost of such an alternative is clearly far less than the extreme option of direct subsidization. This is not to say that individual assessment is necessarily a reasonable LRM. There may still be a convincing argument that the time commitment in setting up a system such as this would prevent Argentina from offering an immediate response to the escalation of the financial crisis. That aside, this opposing hypothetical offers a useful indication of how the reasonable LRM test might be sensibly applied by future adjudicators to carefully balance the interests of both foreign investors and an invoking State.
VI. CONCLUSION
The conflicting jurisprudence to emerge from the Argentine cases on the scope of the treaty defence has already prompted a shift in treaty practice. Newer investment instruments, especially those concluded by the US, have been amended to ensure the invocation of the treaty exception becomes a matter of competence for signatory States alone.Footnote 232 The shift towards auto-interpretation in this area is however only a prospective one and limited to certain States. There remain dozens of pending cases where adjudicators will be required to interpret and apply the operative components of the treaty defence
In their deliberations, future tribunals should carefully test the relationship between the treaty defence and the customary plea of necessity. The dominant methodology in the jurisprudence to date of conflating these two legal standards fails to engage this fundamental question. It has simply been assumed that the treaty exception and the customary defence are one and the same. This is mistaken both on a close interpretation of the two legal standards and on a broader historical analysis of the emergence of investment treaty norms. There are real-world implications which flow from this unsatisfactory set of cases. Argentina is now liable for hundreds of millions in dollar damages despite objective evidence of legal error by the method I tribunals. If carried through, Argentine threats of non-compliance with ICSID awards will present one of the greatest challenges to the viability of the system of investment treaty arbitration.
We face a pressing need for future adjudicators to construct a robust and defensible approach to the relation between the treaty exception and customary plea. The choice, to my mind, really comes down to an election between methods II and III. The LG&E award reveals a flirtation with the precepts of reading the treaty defence as an expression of lex specialis. It fails though to engage the fundamental question surrounding that methodology II, which is the scope of priority to be accorded to the treaty defence. There is a plausible claim that custom continues to operate in a residual fashion unless in direct conflict with the specialized treaty defence. This approach would justify an adjudicator applying the strict customary test of means-end inquiry to the treaty exception. The practical outcome of this would be a type of method I redux, the preclusion of the defence in all but the most extreme situations. Methodology III is, in my view, the more convincing and coherent reading of the relationship between the two legal norms. It obliges an adjudicator to interpret the treaty defence on its own terms without simply transplanting from customary law. Here also, the quality and rigour of reasoning is as important as outcome in the delineation of concepts such as ‘public order’, ‘essential security interests’ and whether a chosen law is ‘necessary’ to achieve those goals. This is perhaps the message that should speak loudest to future adjudicators of these questions.
