The Essential Security Interest (ESI) clause, despite its increasing prevalence in Bilateral Investment Treaties (BITs), has been subject to conflicting interpretations by Tribunals over the last few decades. The provision serves as a powerful tool at the State’s disposal, permitting non-precluded measures taken for the protection of its essential security interests, which would otherwise violate substantive BIT provisions. Although the possibility of misuse warrants careful assessment by Tribunals, the application of tests often seems to reflect an unrealistically high threshold and a predisposition to deny applicability. The ruling in Seda v. Colombia serves as a testament to this inconsistency, with the Tribunal misapplying the relevant tests and misreading the parties’ intent to ultimately confirm the justiciability of the security exception.
Much has been written about the ruling, praising the comprehensive assessment of the components of an ESI clause. In its official press release, Colombia hailed the decision as “protecting national interests, the common good and justice”. Although it upheld the applicability of the ESI clause, the assessment of the Tribunal was far from perfect. This piece seeks to unpack the implications of this “historic” ruling – namely, it assesses the thresholds and tests applied by the Tribunal and its consequences to the ever-expanding discourse surrounding essential security interests.
Factual Background
The dispute stemmed from asset forfeiture proceedings initiated against the Meritage real estate project, developed by the Claimants – a group of American investors, including Angel Seda. During the acquisition of the Meritage property in Medellín, the property was seized by Colombia under the 2014 Colombian Asset Forfeiture Law, due to irregularities in the chain of title and the investors’ past links with drug trafficking. In response, the Claimants instituted ICSID arbitration proceedings under the United States-Colombia Trade Promotion Agreement (TPA). Colombia, in turn, invoked Article 22.2(b) of the TPA, an ESI clause, to exclude the impugned actions from the treaty’s scope.
A Brief Overview of the Ruling
The Claimants’ first contention was that the invocation of the ESI provision as a “new defense” emerging only from the rejoinder, was belated, as the Respondent should have identified the essential security concerns at the time of implementing the contested measures. However, since the provision in question had no temporal qualifier and the Tribunal could consider submissions at any stage of the proceedings, there was no basis for the Tribunal to disregard a defense for which the ‘right to be heard’ was adequately preserved (¶ 615-620).
The Tribunal then assessed Article 22.2(b) of the TPA, which states that nothing in the agreement shall be construed to prevent a Party from adopting measures it considers necessary for the protection of its essential security interests. A footnote to this provision further clarifies that in arbitral proceedings under the investment or dispute settlement chapters of the TPA, a Tribunal shall find the exception applicable when invoked by a Party.
The Tribunal began its assessment by addressing the State’s contention that the provision is self-judging (¶ 431); if deemed so, the State’s perceived necessity of the security measures would be accepted automatically. This would require clear and explicit language, as provided in previous interpretations of the security exceptions (¶ 636), notably the Nicaragua and Oil Platforms judgments. Presently, the phrase “considers necessary” was pivotal in confirming the self-judging nature of the provision, thereby limiting the Tribunal’s scope of review to a light-touch, good-faith assessment.
A significant point of debate was whether the provision was non-justiciable – i.e., whether its invocation was beyond the Tribunal’s review. While the self-judging nature of an ESI clause implies deference to a State’s judgement of the perceived ‘necessity’ of its measures, entailing only a limited, light-touch review. However, a non-justiciable clause goes further, precluding any scrutiny by the Tribunal at all, completely deferring to the sovereign authority of the State. The Tribunal rejected this argument and affirmed its authority to assess the invocation. It found that Article 22.2(b) did not contain language indicating non-justiciability, especially apparent upon comparison with the Singapore-India Comprehensive Economic Cooperation Agreement and the Colombia-India interpretative declaration, both of which reserve non-justiciability in explicit terms (¶ 712-714).
The Tribunal further stated that the footnote to the provision did not support a conclusion of non-justiciability, as the “process of invocation of the ESI Provision entails a finding of applicability by a Tribunal, meaning that the provision does not apply automatically” (¶ 719). If understood as a non-justiciable exception, the ESI provision could emerge as an unchecked tool for States, undermining the investors’ rights. In clarifying its position, the Tribunal emphasized that it did not – and could not – deny the contracting States their sovereign right to define the scope of the exception; rather, it required an unambiguous stipulation to that effect to prevent misuse (¶ 722).
Alternatively, the Claimants invoked Article 10.4, the Most Favored Nation (MFN) clause, to benefit from a more favorable standard of treatment for investors from the Colombia-Switzerland BIT. As this BIT had no security exception, this would amount to a more favorable standard of treatment, which in turn would effectively eliminate the security exception in the present BIT. However, this argument was categorically rejected by the Tribunal, since the MFN clause did not extend to dispute resolution and the Claimants’ attempt to invoke it aimed to shield the dispute resolution provisions of Chapter 10 of the TPA (¶ 796-799).
Finally, the Tribunal applied the facts of the case to the prongs of Article 22.2(b). First, the Asset Forfeiture Proceedings were uncontested as a ‘measure’ under Article 22.2(b). Second, an essential security interest was indeed implicated, owing to the inherent link between organized crime and drug trafficking on the one hand, and public safety, socio-economic stability and security on the other. Third, regarding the nexus, the Tribunal adopted a minimum threshold of ‘plausibility’, merely requiring that the measure is “not implausible” to protect the proffered security interests. The Tribunal found that this standard was satisfied, since there was a plausible link between the proceedings concerning the criminal origin of the Claimants’ assets, the underlying legislation, and the goal of combating drug trafficking. Accordingly, Colombia’s invocation of the security exception was successful, precluding the wrongfulness of the investors’ treatment (¶ 801-802).
Non-justiciable ESI Provisions: An Impossible Threshold?
The tone of the ruling suggests that the Tribunal has set an overly onerous threshold for classifying a security exception as non-justiciable, elevating this standard until it became unattainable in the present case. While concern over misuse of security exceptions insulated from review is valid, the Tribunal explicitly acknowledged that it neither could nor intended to restrict States’ sovereign power to define the scope of the exception (¶ 722). Yet, the Tribunal proceeded to do exactly this.
The interpretation of Footnote 2 illustrates this point. It’s plain text states that upon invocation during arbitral proceedings, the Tribunal “shall” find that the exception applies – a formulation that suggests non-justiciability. However, in countering this reading, the Tribunal adopted a particularly problematic comparative approach, bringing forth provisions from other treaties, interpretative declarations and clarifications that contain more explicit language of non-justiciability. It diluted the implications of the footnote, as one that merely “entails a finding of applicability” (¶ 719). Rather than discerning an evident parties’ intent from the text of the treaty at hand, the Tribunal relied on more concrete expressions of intent elsewhere. Yet, the existence of clearer articulations of non-justiciability in other instruments does not negate the apparent parties’ intent to compel the Tribunal to apply the security exception when invoked – therefore, non-justiciability of the invocation.
This is not to suggest that non-justiciability is normatively desirable, as insulating such measures from the Tribunal’s scrutiny could allow a carte blanche for unfettered measures under the garb of security, all at the investors’ expense. While the Tribunal rightly identified this risk, it seemed to have fundamentally guided the Tribunal’s assessment. It appeared to conduct a deliberately colored assessment of the treaty language to reach a predetermined conclusion favoring justiciability. If the intent was truly to prevent such an escape clause, why did the Tribunal consistently reiterate the right of States to conclude non-justiciable ESI provisions?
The Merits: To Review or Not to Review?
The Tribunal’s inconsistency with its own standard for reviewing the ESI provision is also evident in the effect of the Tribunal’s review. While the Tribunal maintained that the inferences were not based on the merits and merely a light-touch, good-faith review had been carried out (¶ 787) – the analysis suggests otherwise.
On one hand, the Tribunal maintained that the required nexus between the measure and the security interests need only meet a plausibility threshold; since it was not a criminal court, the Tribunal underscored the need to avoid a merits inquiry. Ordinarily, this assessment would be prima facie satisfied by the fact that the measures “could” be not-implausibly linked with the criminal activity and drug trafficking, which was already inferable from the status of the criminal investigations (¶ 790-792).
However, in the very next paragraph, the Tribunal took a step further, proceeding to examine the chain of title and transactional history of the Meritage Lot to establish its criminal origins, thereby entering the merits (¶ 793). Even if this analysis merely confirmed the ultimate result, it exceeded the Tribunal’s own declared limits, thereby setting a poor precedent.
The Curious Case of the MFN Clause
Finally, the Tribunal’s analysis of the MFN clause was a case of accurate conclusions for rather peculiar reasons. The Tribunal rightly denied the applicability of the MFN clause to override the security exception. However, this was based on the conclusion that the MFN provision did not extend to dispute resolution, and that the Claimants had invoked it to safeguard the Tribunal’s power, jurisdiction and available remedies. However, the impact on power, jurisdiction and available remedies presents an unnecessarily convoluted yardstick. If the mere potential to affect available remedies bars the MFN clause’s application, this could preclude its use in fundamental matters such as Fair and Equitable Treatment.
A more appropriate rationale could be derived from Stephan W. Schill’s critique of the CMS v. Argentina ruling: a security exception as an exclusion from the treaty’s scope cannot be overridden by another treaty provision, such as an MFN clause. The ESI clause specifically begins with the phrase “Nothing in this agreement”, signalling its precedence over all treaty provisions, including the MFN clause. Accordingly, the MFN clause cannot be employed to negate the operation of the security exception
Conclusion
The ruling has been widely praised for its unprecedented consolidation of comprehensive tests, enabling a case-by-case assessment. As one of the first publicly known awards addressing the justiciability of security exceptions in BITs, the ruling is likely to have critical implications.
However, the Tribunal’s ruling is far from the definite clarification of the security exception it is often portrayed to be. In an effort to resolve the dilemma surrounding non-justiciable security exceptions, the Tribunal appears to have set an unrealistically high threshold, guided by a preconceived commitment to justiciability. It set boundaries for its own review, only to subsequently overstep them – this pattern is also evident in its treatment of the merits. Accordingly, the Tribunal has consistently arrived at the right conclusions, but for the wrong reasons. While the invocation of the security exception may have sustained, and the justiciability of the provision did not impact the ultimate confirmation of the exception’s applicability, the Tribunal’s method could have resulted in an inaccurate conclusion. Therefore, in a rather limited step forward regarding the circumstances of the case, the Tribunal took a few unfortunate steps back in developing discourse surrounding the assessment of the security exception.
Akshath Indusekhar is a law student at Jindal Global Law School. His areas of interest include Public and Private International Law, International Trade Law and International Investment and Commercial Arbitration.