Self-judgment in international law: Between judicialization and pushback

Abstract

The legally binding unilateral application of norms holds potential for abuse. Nonetheless, self-judgment is alive and kicking. Self-judgment language commonly features in treaties and states frequently invoke their authority to ‘self-judge’ sensitive issues, such as matters related to national security, before international judicial bodies. In many of these cases, the controversy whether a norm has a self-judgment quality or not has been decisive for the outcome of the dispute. Yet, the meaning and consequences of self-judgment remain contested.

This article develops self-judgment as the authoritative application of international legal norms by states. It posits that steps towards the judicialization of self-judgment by judicial bodies have given rise to state efforts to preserve unfettered discretion. Notably, states have responded to attempts by judicial bodies to gain authority over the application of self-judgment by drafting provisions more explicitly. This dynamic continues to make self-judgment a site of judicialization and pushback. The only way to understand the meaning, limitations and development of self-judgment is by studying this process. Doing so conceptually refines self-judgment and allows for more meaningful references to the notion in practice.

1. Introduction

In 2013, Kenneth Vandevelde, a former investment treaty negotiator from the United States, observed that the ‘slow proliferation of treaties with self-judging language … is particularly disturbing in the absence of a consensus about what it means for an exception to be self-judging’. ¹ He astutely observed two phenomena that seem to be in an uneasy relationship.

First, self-judgment seems to be ever more present in treaty practice, legal rhetoric, and judicial proceedings. States frequently claim before international judicial bodies that certain norms are of a ‘self-judgment’ quality. This, so goes the common argument, entails the unreviewable right of the invoking state to decide whether the requirements of the norm have been fulfilled. ² In recent years, for example, several states invoked national security exceptions in the World Trade Organization (WTO) framework and before the ICJ based on exactly this argument. ³ In treaty practice, self-judgment also seems to be en vogue: explicit self-judgment provisions are included in a growing share of investment agreements. ⁴ The classic example of explicit self-judgment is a treaty provision that applies ‘if the state considers’ the requirements of the provision to be fulfilled. ⁵ For instance, Article 13(2) of the 2016 Japan-Iran bilateral investment treaty (BIT) states that:

Nothing in this Agreement shall be construed to prevent a Contracting Party from adopting or enforcing measures:

(a) which it considers necessary for the protection of its essential security interests … ⁶

Second, there is a striking imbalance between the practical relevance of self-judgment and the attention it has received as a legal phenomenon. One would assume that this practical relevance would have prompted much scholarly attention. Indeed, the label of ‘self-judgment’ features regularly in scholarly discussions. ⁷ Yet, academic efforts have almost exclusively explored what specific self-judgment provisions mean in the context of their respective fields of international law, notably under Article 36(2) of the ICJ Statute, WTO law, and international investment law. ⁸ Frequently, scholarship takes the existence of self-judgment as a legal phenomenon as a given but stops short of spelling out what its features and implications are. As a result, states, scholars, and judicial bodies often employ the term ‘self-judgment’ loosely, lumping together a range of state actions related to unilateralism, the rejection of delegation of decision-making power, and the capacity to decide autonomously in particularly sensitive areas.

Allowing states to decide authoritatively on the application of norms that are commonly related to their national security interests holds significant risk of abuse. If states are free to apply or disapply international law as and when they please, treaties and the judicial bodies created to ensure compliance with these treaties risk being undermined. At the same time, international law has long acknowledged space for self-judgment within the international legal order. ⁹ States seem to value self-judgment as a legal technique to grant flexibility. Notably, it is said to function as a ‘safety valve’, allowing states to opt out of their international legal commitments when political or economic pressures become too high. ¹⁰

This article argues that these conflicting interests have made self-judgment a site of continuous judicialization and pushback. A tension exists between those who favour an increased reliance on law and international judicial bodies to regulate self-judgment norms, and those who seek to maximize unimpeded state decision-making. The term judicialization describes, with different emphases, an increasing reliance on judicial bodies and judicial means of decision-making to address political controversies. ¹¹ The article chooses the lens of judicialization to analyse self-judgment in the context of judicial dispute settlement ¹² because this is where the effects of self-judgment become most visible. ¹³

The judicialization of self-judgment in international law can be understood to have proceeded in two stages. First, the ICJ articulated a presumption against implicit self-judgment. This prompted some pushback in treaty-making practice, as states shifted towards a more frequent use of explicit self-judgment in investment and trade law. In a second step, international judicial bodies subjected explicit self-judgment provisions to a good faith review. While a part of the state community seems to accept this shift, some states are pushing back, by drafting treaty provisions even more expressly to limit the competence of international judicial bodies to review the invocation (‘reinforced self-judgment’). The current contestation of the effects of reinforced self-judgment demonstrates that the struggle over the ‘last word’ continues. Only by appreciating this struggle can self-judgment be refined conceptually and employed more meaningfully in judicial practice. The article seeks to lay the groundwork to do so. A better understanding of self-judgment also helps to distinguish between legally relevant claims of self-judgment and mere rhetoric. In that, the article also aims to make the legal implications of self-judgment provisions more predictable.

This article will begin by providing some background to the concept of self-judgment and judicialization (Section 2) before presenting the two stages of the judicialization of self-judgment (Sections 3 and 4) and the respective pushback against these developments.

2. The concepts of self-judgment and judicialization

When approaching the concept of self-judgment, it is useful to recall H.L.A. Hart’s ‘core of settled meaning’. ¹⁴ Hart accepted that legal norms have an ‘open texture’ that allows for different reasonable interpretations. ¹⁵ At the same time, legal norms have a core of settled meaning that serves as the premise of different interpretations. Other legal philosophers have come to similar conclusions. For example, Philipp Heck, back in 1914, distinguished between the ‘conceptual core’ and the broader ‘conceptual space’ allowed for by a particular term. ¹⁶ These insights concerning the interpretation of (legal) terms can be applied to self-judgment norms too. Although the concept of self-judgment is contested, one can identify a core within which emanations of self-judgment can be pinpointed. ¹⁷

This core is the authoritative application of legal norms. Authority is a complex phenomenon with a long and varied history transcending individual disciplines. ¹⁸ At a general level, authority has been understood as a relationship that grants the power to impose one’s will on others regardless of the correctness of the underlying judgment. ¹⁹ In the realm of international law, one understanding of authority describes the capacity of actors to assign meaning to norms that bind other actors. ²⁰ A state that is entitled to apply a self-judgment norm is given the authority to determine whether the legal and factual requirements of the norm are fulfilled and once it makes that determination other subjects of international law are – at least to some extent – bound by this determination. This can be conceptualized as discretion or deference. ²¹ However, it is important to note that the link between self-judgment and legal authority does not necessarily imply that invocations of self-judgment norms are within the unfettered discretion of the invoking state. While this may at first glance appear to contradict the notion of ‘self-judgment’, a key claim of this article is that the meaning of self-judgment has developed over time in stages that differ in the degree to which self-judgment allows for the exclusive decision-making of states.

When conceptualizing self-judgment, there are two grounds on which states base their authority in the application of the norm. ‘Implicit self-judgment’ is based on the nature of the norm and needs to be determined through interpretation, taking into account the intention of the norm-creators. ²² States usually argue in favour of implicit self-judgment in matters of great subjective importance (most often security matters), seeking to ground their authority to self-judge precisely in this importance. Argentina, for example, argued in disputes arising out of the 2001–2002 financial crisis that the essential security interests exception in its BIT with the United States constituted an implicit self-judgment norm. ²³

In contrast, ‘explicit self-judgment’ describes the quality of provisions that expressly grant states a degree of authority to determine whether or not the requirements of the provision are fulfilled and whether the provision can be applied. Explicit self-judgment often features in provisions that condition their application on the state ‘considering’ or ‘determining’ that the clause applies. ²⁴ The paradigmatic example of an explicit self-judgment provision is Article XXI(b) of the 1994 General Agreement on Tariffs and Trade (GATT), ²⁵ the beginning of which reads:

[n]othing in this Agreement shall be construed to prevent any party from taking any action which it considers necessary for the protection of its essential security interest … (emphasis added)

Exploring the concept of self-judgment requires more than engaging with any one contingent conception of self-judgment at any particular point in time. One must also seek to understand how self-judgment developed over time. Capturing this development in turn requires an analytical frame. A valuable concept to analyse the relationship between the changing conceptions of self-judgment is that of judicialization.

Judicialization has been a key parameter to discuss, measure, and analyse the role of law in international relations since the late 1990s. The term judicialization describes, with different emphases, an increasing reliance on judicial bodies and judicial means of decision-making to address political controversies. For Andreas Føllesdal and Geir Ulfstein, judicialization captures the multiplication of international courts as well as their users and their functions. ²⁶ Ran Hirschl describes judicialization as ‘the expansion of the province of courts and judges in determining public policy outcomes’. ²⁷ Karen Alter and others focus on the ‘delegation to an adjudicatory institution’ ²⁸ when discussing judicialization. It is the delegation of authority that appears particularly pertinent for the purposes of investigating self-judgment. This delegation of authority from states to international judicial bodies has also been described as one aspect of the broader process of the legalization of international relations. ²⁹

At the same time, judicialization has been no one-way street. It has been characteristic of the development of self-judgment that there have been continuous waves of state pushback against judicialization. Pushback describes acts of resistance against a court’s authority ‘exercised according to the “rules of the game” and within the institutional system of the [international judicial body]’. ³⁰ It is a process aimed at shifting authority away from international judicial bodies back to states. ³¹ Scholarship generally employs pushback to denote acts of resistance against individual institutions. However, recent work understands resistance against specific international judicial bodies as attempts to roll back the judicialization of international relations more broadly. ³² In this way, pushback complements the lens of judicialization and allows us to accurately grasp the development of self-judgment as a contested concept.

If judicialization is understood as the delegation of authority to international judicial bodies, and self-judgment as a legal technique to entrench authority of states in the application of legal norms, judicialization is a particularly promising instrument to make sense of the development of self-judgment over time. Changes in the understanding of self-judgment can be interpreted as either furthering or pushing back against judicialization. If, for example, the prevalent understanding of self-judgment under international law used to be defined by unfettered discretion and law-appliers are progressively shifting to accept degrees of reviewability, this is a development towards judicialization because authority is shifting from states to international judicial bodies (delegation). Having said that, one must be aware that judicialization is generally employed to capture macro trends in the international landscape that will rarely be brought about by any individual judicial decision. Nonetheless, as this article shows, judicial decisions have and continue to play a crucial role in the development of the concept of self-judgment.

Against this backdrop, self-judgment and judicialization should not be viewed as a dichotomy. Instead, together with the notion of pushback, judicialization helps to make sense of the development of self-judgment, highlighting its changing, gradational nature. In other words, the dynamics of self-judgment can be visualized as a pendulum swinging between judicialization and pushback.

3. The presumption against implicit self-judgment

The idea of states possessing particular authority to apply specific legal norms based on their nature can look back at a long intellectual legacy in general international law. In some emanations of the early just war doctrine, especially in the scholarship of Alberico Gentili, the sovereign’s decision and corresponding belief sufficed to make a war objectively just. ³³ More recently, at the turn of the twentieth century, the notion of implicit self-judgment was invoked in debates on arbitration treaties. The relevant provisions carved out an exception to the obligation to resort to arbitration if the dispute concerned grounds such as ‘vital interests’, sovereignty, or states’ domestic jurisdiction. ³⁴ While some of these exceptions used explicit self-judgment, many did not mention who possessed the final authority on the invocation of the clause. Nonetheless, it was widely accepted that it was the parties to the treaties themselves who unilaterally decided on their application. ³⁵

From its historical roots until today, states usually advance claims of implicit self-judgment when they believe that the subject matter is so important for their interests that no other actor can or should have a say in the application of said legal norm. Bearing this in mind, it comes as no surprise that claims of implicit self-judgment surfaced most prominently in relation to particularly sensitive issues, such as national security. The ICJ and international investment tribunals were decisive in articulating a presumption against self-judgment.

3.1 Coining the presumption: Nicaragua v. United States and its repercussions

The ICJ provided some of the most important contributions to the understanding of self-judgment. Of these, the landmark decision Nicaragua v. United States stands out. ³⁶ Nicaragua based some of its claims on the 1956 Treaty of Friendship, Commerce and Navigation (FCN Treaty). ³⁷ Article XXI(1)(d) of the FCN Treaty laid down that the treaty ‘shall not preclude the application of measures … necessary to protect [a party’s] essential security interests’. The United States invoked the provision. In his pleadings for Nicaragua, Abram Chayes sketched a continuous evolution from early vital interest exceptions to the argument of the United States that the Court should refrain from adjudicating the dispute which involved matters of armed conflict. ³⁸

It was in light of this legacy that the Court pronounced on the arguable implicit self-judgment nature of the FCN Treaty’s security exception. The section in which the Court addressed the United States’ invocation of Article XXI(1)(d) of the FCN Treaty provides a central puzzle piece to the development of self-judgment:

Article XXI defines the instances in which the Treaty itself provides for exceptions to the generality of its other provisions, but it by no means removes the interpretation and application of that article from the jurisdiction of the Court as contemplated in Article XXIV. That the Court has jurisdiction to determine whether measures taken by one of the Parties fall within such an exception, is also clear a contrario from the fact that the text of Article XXI of the Treaty does not employ the wording which was already to be found in Article XXI of the General Agreement on Tariffs and Trade … The 1956 Treaty, on the contrary, speaks simply of ‘necessary measures’ not of those considered by a party to be such. ³⁹

Later in the judgment, the Court emphasized again that:

…by the terms of the Treaty itself, whether a measure is necessary to protect the essential security interests of a party is not, as the Court has emphasized (paragraph 222 above), purely a question for the subjective judgment of the party; the text does not refer to what the party “considers necessary” for that purpose. ⁴⁰

With this decision, the ICJ articulated a presumption against self-judgment. ⁴¹ The decisive factor to determine the non-self-judgment character of Article XXI of the FCN Treaty for the ICJ was the wording of the clause. The Court expressly backed up its holding that the application of Article XXI(1)(d) of the FCN Treaty cannot be ‘purely subjective’ because the provision lacks the self-judgment language of Article XXI of the GATT (‘considers necessary’). ⁴² For modern international lawyers, this may seem intuitive but in the light of the legacy of ‘vital interests’ the Court’s holding was a crucial clarification. By requiring express self-judgment language, the ICJ considered it not enough that an issue was particularly sensitive by its nature, e.g., by concerning issues of national security. ⁴³ This can be contrasted with early approaches to self-judgment. In 1929, Robert Wilson had still assumed a presumption in favour of self-judgment stating that ‘the presumption is that each state may decide for itself whether a particular dispute falls inside or outside of the reserved classes of questions’. ⁴⁴ In contrast, the Court moved towards a conception that considered self-judgment to be the exception even in areas of major importance to the national interest. The ICJ reiterated this view in Oil Platforms regarding a similar clause in the Iran-US Treaty of Amity. ⁴⁵

The ICJ’s holdings were later entrenched by investment arbitration tribunals in a number of arbitrations arising in the aftermath of Argentina’s economic crisis between 1999 and 2002. ⁴⁶ In response to measures taken by Argentina to rein in the effects of the crisis, dozens of foreign investors filed claims against the state before investment tribunals under BITs. ⁴⁷ Argentina notably relied on the essential security interests clause, Article XI, of the 1991 Argentina-US BIT. ⁴⁸ Drafted in similar language to the provisions discussed by the ICJ in Nicaragua, Article XI of the Argentina-US BIT states that:

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.

Argentina argued that its crisis measures served the protection of its security interests under Article XI of the treaty. Argentina’s position was that the clause was ‘self-judging’, though over time it seems to have taken different approaches as to the legal consequences of this label. ⁴⁹ It appears to have endorsed the most comprehensive view of the meaning of self-judgment in the proceedings that gave rise to the first publicly available award relating to Argentina’s crisis measures, CMS v. Argentina, which was handed down in May 2005. The tribunal described Argentina’s position as follows:

[T]he Respondent … believes that it is free to determine when and to what extent necessity, emergency or the threat to its security interests need the adoption of extraordinary measures. ⁵⁰

In CMS, the tribunal rejected interpreting Article XI of the Argentina-US BIT as a self-judgment norm. The tribunal relied on the written form presumption of self-judgment established in Nicaragua v. United States, stating that it ‘is convinced that when States intend to create for themselves a right to determine unilaterally the legitimacy of extraordinary measures importing non-compliance with obligations assumed in a treaty, they do so expressly’. ⁵¹ CMS v. Argentina, despite its subsequent annulment, set the tone for an avalanche of awards. Without exception, the tribunals dismissed Argentina’s self-judgment arguments based on similar points as the CMS award. ⁵² In Enron v. Argentina, for example, the tribunal held that:

Truly exceptional and extraordinary clauses such as a self-judging provision normally must be expressly drafted to reflect that intent, as otherwise there can well be a presumption about it not having that meaning in view of its exceptional nature. ⁵³

The jurisprudence of the ICJ and investment tribunals established that states could no longer expect to authoritatively apply norms of international law, even if a matter was particularly sensitive. This necessarily entails a shift in authority from states to international institutions when they are competent to review self-judgment claims. Accordingly, the development of the presumption against self-judgment constituted a first step towards the judicialization of self-judgment.

3.2 Pushback: The shift towards explicit self-judgment

The articulation and acceptance of the presumption against self-judgment had at least two consequences. First, it incentivized states to look for functional equivalents to implicit self-judgment. For example, states began to argue more assertively in favour of a ‘margin of appreciation’ or ‘deference’ when matters of national security were at stake. The question as to where the limits of the appropriate degree of international control of state action in sensitive areas lie, has formed a focal point of debate on international judicial bodies in recent years. ⁵⁴

Second, the presumption increased the burden on states to establish the self-judgment character of a norm by requiring explicit self-judgment if they wanted to limit the competences of judicial bodies in particular areas. This is precisely what states did by using explicit self-judgment provisions in their treaty drafting more frequently. As international judicial bodies limited the scope for implicit self-judgment, states sought to carve out discretionary space by adapting their treaty practice. Over the 1990s and 2000s, as the decline of implicit self-judgment sank in, explicit self-judgment provisions began to proliferate rapidly, particularly in international investment and trade agreements. In response to the Nicaragua decision, ⁵⁵ the United States began to include explicit self-judgment provisions as a standard feature in its investment agreements. ⁵⁶ It was however, not alone in doing so. Canada, Turkey, and Japan were among several states that also began to include self-judgment language in their trade and investment treaties. ⁵⁷ Similarly, the EC included self-judgment provisions in association agreements, and self-judgment language found its way into the Energy Charter Treaty. ⁵⁸ Given the influential nature of US treaty practice, it is likely, though still speculative, that the explicit self-judgment provisions were adopted from US treaties into other agreements. In sum, the establishment of the presumption against implicit self-judgment was followed by an increasing move towards explicit self-judgment, and this move is best read as pushback against the judicialization of self-judgment.

4. Judicialization through good faith review

The struggle over the authority to apply self-judgment norms continues in a second phase of judicialization. This section traces how good faith as a limitation to self-judgment emerged as an argumentative practice in the context of self-judgment provisions in early arbitration treaties, self-judgment reservations to Optional Clause declarations, and national security exceptions in international trade law. Good faith review did not provide the basis for judicial decisions on self-judgment until the 2000s, but early arguments in favour of a good faith limitation were crucial in paving the way to this second phase of the judicialization of self-judgment.

The notion of good faith review is based on the premise that self-judgment as an unreviewable exit route from treaty obligations holds potential for abuse. At a general level, good faith can be substantiated with the notion of reasonableness, implying ‘any nation [has to] take into account the reasonable expectations of all other members of the international community’. ⁵⁹ One central function of good faith, therefore, is to limit the reliance on norms even if the act of a state is covered by the wording or substantive scope of a particular provision. Thus, it emphasizes the object and purpose of international legal norms.

There are two ways in which good faith can be conceptualized as a legal principle that limits states in their invocation of self-judgment norms. The first way focuses on the intersection of good faith in treaty law and self-judgment. The notion that treaty provisions must be ‘interpreted’ and ‘performed’ in good faith is one of the bedrock principles of the 1969 Vienna Convention on the Law of Treaties, enshrined in the preamble, Article 26, and Article 31. ⁶⁰ Against this background, good faith translates into an obligation to interpret and apply treaties reasonably in a way that does not defeat their purpose. The ICJ also pointed out how good faith serves as the foundation of treaty law and unilateral acts alike. ⁶¹

The second way to understand good faith as limitation of self-judgment is based on the obligation to exercise discretion in good faith. Often established based on a comparative legal exercise with reference to détournement de pouvoir, Ermessensmissbrauch or English administrative law, the main thrust of this emanation of good faith is the prohibition of the exercise of discretion for purposes other than those for which the power was conferred to the state. ⁶²

The idea to review whether a norm has been invoked in compliance with a good faith obligation is the notion of ‘good faith review’. It is a standard of review denoting that a reviewing body shall not assess the invocation of a self-judgment norm as to its ‘correctness’ but only determine whether or not the invocation occurred in ‘good faith’. ⁶³ In line with the understanding of good faith in international law more generally, this is usually done by analysing whether states acted rationally and in compliance with the purpose of a particular norm. ⁶⁴ Good faith review is deferential because it allows states considerable leeway in their application of self-judgment norms. ⁶⁵ As such, it has been described as a ‘minimal review’, ⁶⁶ an ‘extremely lenient standard’, ⁶⁷ and the ‘minimum permissible level of scrutiny in international adjudication’. ⁶⁸

Good faith review and unfettered discretion seem similar at first sight. While the invocation of the self-judgment norm is not completely unreviewable under a good faith standard, it still allows for considerable state discretion. ⁶⁹ However, upon closer examination, the difference is categorical. Once it is acknowledged that a self-judgment norm is not by its nature beyond the review of the judicial body, adjudicators find themselves in murky waters. If judicial bodies discard the non-justiciable understanding and instead base the review on notions that require further concretization, such as good faith, they gain considerable influence over the application of self-judgment norms. Viewed from this perspective, the application of a good faith review to self-judgment norms is not a decisive shift because of the intensity of the review but rather because it constitutes a review in the first place. As observed by Donald Greig, the purpose of good faith review can be to bring a formerly non-justiciable norm ‘within the scope of judicial review’. ⁷⁰ Applying a good faith review, thus, can be considered a technique to judicialize self-judgment.

This section analyses the judicialization of self-judgment through good faith review. It first offers a brief overview of the early practice of explicit self-judgment (Section 4.1). It then explores how good faith emerged as an argumentative practice in scholarship and judicial proceedings (Section 4.2) before it was mainstreamed through a series of influential decisions (Section 4.3). Finally, the section turns to the practice of ‘reinforced self-judgment’ which can be seen as pushback against the judicialization of explicit self-judgment (Section 4.4).

4.1 Early explicit self-judgment

Explicit self-judgment provisions first emerged in the late nineteenth century when third-party international dispute settlement slowly began to increase. The purpose of the first generation of explicit self-judgment provisions was to limit the competence of arbitral tribunals, which derived their jurisdiction from compromissory clauses or general arbitration treaties in the late nineteenth century. The first example of an explicit self-judgment provision can be found in the 1890 arbitration treaty between Guatemala and El Salvador. Article VIII(3) excepted disputes from compulsory arbitration ‘which, in the opinion of one only of the nations interested in the dispute, would compromise its autonomy and its independence’. ⁷¹ The provision was most likely inspired by negotiations at the First Conference of American Republics held in Washington 1889–1890. At the Conference, the draft of a general arbitration agreement provided that all disputes could be submitted to arbitration except when ‘in the judgment of any one of the nations involved in the controversy, [the dispute] may imperil its independence’. ⁷² While the draft was never turned into a treaty, the idea seems to have stuck.

Bilateral arbitration treaties became en vogue after the turn of the twentieth century, especially in the aftermath of the First Hague Peace Conference of 1899, when dozens of treaties were concluded within a few years. ⁷³ While the wording of the exceptions differed in detail, they commonly excluded from the scope of arbitration disputes that affected the ‘honour’, ‘independence’ or ‘vital interests’ of a state party. ⁷⁴ A provision in the Anglo-French arbitration treaty of 1903 proved to be a blueprint in this regard. ⁷⁵ Some of the treaties concluded between 1899 and 1914 not only set out these broad exceptions but also combined them with explicit self-judgment language, according to which recourse to arbitration was excluded if one party ‘considered’ the question to, for example, concern its vital interests. ⁷⁶ As a result, explicit self-judgment provisions had found their way into the mainstream of international dispute settlement. ⁷⁷

When states began to introduce explicit self-judgment exceptions into arbitration treaties it did not take long until scholars first attempted to link the self-judgment to the obligation of good faith. Heinrich Lammasch and Otfried Nippold for example, claimed self-judgment provisions must not be invoked arbitrarily because treaties have to be applied bona fide by the invoking state. ⁷⁸ But the notion of a ‘good faith’ limitation to interpreting early explicitly self-judgment provisions was by no means universally accepted. Indeed, most contributions dealing with self-judgment provisions in early arbitration treaties did not address the issue of good faith. ⁷⁹

In the inter-war period, self-judgment reservations in general arbitration treaties first reduced in numbers and then slowly withered away – as did general arbitration treaties more broadly. Self-judgment soon returned in other forms.

4.2 An emerging argument: Good faith as a limitation to self-judgment

After 1945, self-judgment provisions quickly resurfaced in international legal instruments. Even though there were several occasions when states or judicial bodies could have forced binding interpretations of self-judgment provisions, the period from 1945 to the 2000s turned out to be largely one of avoidance. States were generally reluctant to invoke self-judgment provisions themselves and cautious to challenge their invocation legally if other states did so, while international judicial bodies avoided ruling on invocations of explicit self-judgment provisions. ⁸⁰ At the same time, good faith as a limitation of self-judgment increasingly emerged as an argumentative practice. This can be aptly illustrated using one of the main examples of explicit self-judgment during that period: reservations to declarations of acceptance under Article 36(2) of the ICJ Statute (‘self-judgment reservations’).

The earliest and most famous self-judgment reservation was made by the United States in 1946, excepting ‘disputes with regard to matters which are essentially within the domestic jurisdiction of the United States of America as determined by the United States of America’. ⁸¹ Ten other states submitted declarations with similarly worded self-judgment reservations, five of which are still in force today. ⁸²

The ICJ faced invocations of self-judgment reservations in several cases. Notably, in Norwegian Loans, ⁸³ Interhandel, ⁸⁴ and Aerial Incident ⁸⁵ the parties to the dispute and several judges in their individual opinions engaged with the possibility of a good faith review. In Norwegian Loans, Norway invoked the French self-judgment reservation on the basis of reciprocity, arguing that the case exclusively concerned its domestic jurisdiction. In its submissions, the Norwegian government, however, seemed to consider the invocation of the reservation subject to the general good faith requirement of international law. Norway’s agent stated:

Certainly, such a reservation must be interpreted in good faith, and a government that is hiding behind it to deny the Court’s jurisdiction in a case that is manifestly not “essentially within the domestic jurisdiction” would commit an abuse of rights that would not prevent the Court from acting. ⁸⁶

The importance of Norway’s argument lies in its premise. Norway did not specify what steps it believed the Court could take if it found that an abuse of rights had been committed. Nonetheless, its submission necessarily implied not only a legal (as opposed to solely moral) good faith limitation on the state invoking a self-judgment provision but also a competence of the Court to assess whether this condition had been fulfilled. Without such an assessment, it would not be possible for the Court to ‘act’ in any way. In hindsight, Norway’s position thus pioneered the justiciable good faith limitation on self-judgment provisions. ⁸⁷

The Court, however, did not engage with the question of good faith in its decision. Rather, it accepted Norway’s invocation of the self-judgment provision without assessing the validity of the French reservation, because ‘[t]he validity of the reservation ha[d] not been questioned by the Parties’. ⁸⁸

A similar set of arguments was rehearsed in Interhandel, a case in which the United States relied on its self-judgment reservation. In the oral submissions, the Swiss agent claimed that one had to consider the discretion to invoke a self-judgment reservation as limited by considerations of good faith and the prohibition of the abuse of discretionary powers. ⁸⁹ If a self-judgment reservation was invoked arbitrarily, such an invocation had to be considered a ‘nullity’. ⁹⁰ The Court again avoided taking a position on the validity and possible legal consequence of the clause. It declared the Swiss claim inadmissible on the grounds that local remedies had not been exhausted without engaging with the United States’ invocation of the self-judgment provision.

In Aerial Incident, the United States first seemed to backtrack from its position in Interhandel and asked the ICJ to conduct a good faith review of Bulgaria’s invocation of the reservation. While the United States did not question Bulgaria’s right to rely on reservations of the United States based on the principle of reciprocity, it argued that the self-judgment reservation ‘does not permit the United States or any other State to make an arbitrary determination, in bad faith’. ⁹¹ According to the United States, Bulgaria could not determine the domestic character of the aerial incident because this ‘would fly in the face of actuality and would ignore the international character’ of the claim. ⁹² The United States considered Bulgaria’s invocation ‘apparently premised on the proposition that there are no limits upon the right and ability of a State to determine, under the reservation in question, that a matter lies essentially within domestic jurisdiction’. ⁹³ In its submissions, the United States further argued that the self-judgment reservation does not empower any government ‘to make an arbitrary determination that a particular matter is domestic’. ⁹⁴

The Court, however, did not have to rule on the contentious matter. In 1960, the United States asked the Court to discontinue the proceedings. In the letter setting out the reasons for this decision, it fundamentally changed its position regarding the competence of the ICJ to rule on Bulgaria’s invocation of the self-judgment reservation. It stated:

The necessary premise of the [good faith] argument was that the Court must have jurisdiction for the limited purpose of deciding whether a determination under [the self-judgment reservation] is arbitrary and without foundation. On the basis of further study and consideration of the history and background of [the self-judgment] reservation and the position heretofore taken before the Court, it has been concluded that the premise of the argument is not valid and that the argument must therefore be withdrawn … [A] determination is not subject to review or approval by any tribunal, and it operates to remove definitively from the jurisdiction of the Court the matter which [the United States] determines. A determination under [this] reservation that a matter is essentially domestic constitutes an absolute bar to jurisdiction irrespective of the propriety or arbitrariness of the determination. ⁹⁵

The litigation over reservations to Optional Clause declarations gave rise to the first sophisticated engagements with some of the challenges posed by explicit self-judgment provisions. Norway, Switzerland, and – despite its subsequent change of heart – the United States were the first states to seriously invoke good faith as a limitation on the invocation of those provisions.

4.3 Good faith review mainstreamed

From the 2000s onwards, the idea to limit self-judgment based on good faith gained mainstream acceptance. In a series of decisions international judicial bodies implemented good faith review to control the invocation of self-judgment provisions, thereby advancing the judicialization of self-judgment.

4.3.1 Between the lines: International investment law

It was in international investment law that judicial decisions first discussed the notion of a good faith review. While Argentina considered self-judgment to oust the review of arbitral tribunals completely in CMS, it subsequently modified its position to a ‘good faith understanding’ of self-judgment. The argument to limit the security exceptions to invocations which occurred in ‘good faith’ was based on the United States’ BIT negotiations with other states. In the negotiation records, it was frequently stated that the respective parties would ‘expect’ the other side to apply the self-judgment exception only in good faith. This was taken by Argentina to imply the possibility of a ‘good faith review’ by arbitral tribunals as well. ⁹⁶

The tribunals in the Argentinian cases rejected the implicit self-judgment nature of Article XI of the BIT and accordingly did not have to decide whether they would have been able to conduct a good faith review of the provisions had they found them to be of a self-judgment character. However, some tribunals commented on the issue in obiter dicta. Most importantly, the tribunal in LG&E v. Argentina stated that even if it had accepted the self-judgment character of the clause, ‘Argentina’s determination would be subject to a good faith review anyway’. ⁹⁷

Some tribunals also seemed to grasp self-judgment as a legal concept transcending specific areas of international law and accordingly contextualized their interpretive task. For example, the tribunal in Sempra v. Argentina, when denying the self-judgment character of the provision in the Argentina-US BIT, espoused a progressive view on explicit self-judgment as well. In reference to the national security exception in the GATT, it stated that:

not even in the context of GATT Article XXI is the issue considered to be settled in favour of a self-judging interpretation, and the very fact that such article has not been excluded from dispute settlement is indicative of its non-self-judging nature. ⁹⁸

In sum, investment tribunals were the first judicial bodies to accept the good faith review of self-judgment norms and the concomitant shift in authority from states to adjudicators that is captured here by the notion of judicialization. They did, however, only do so ‘between the lines’ in obiter dicta. Nonetheless these holdings informed mainstream scholarship. Soon after the holding in LG&E v. Argentina, the judicialization of self-judgment reached the ICJ.

4.3.2 Judicialization at the International Court of Justice

In 2008, more than half a century after Norway first proposed a good faith review of a self-judgment norm before the ICJ, the Court endorsed the notion in Djibouti v. France. ⁹⁹ In this decision, the ICJ ruled that France’s invocation of an explicit self-judgment provision in a bilateral treaty on mutual criminal assistance with Djibouti was subject to a ‘good faith review’ by the Court. For the concept of self-judgment in international law, the judgment represents a turning point.

In 1986, France and Djibouti concluded a Convention on Mutual Assistance in Criminal Matters which established duties to co-operate but included a national security exception in Article 2(c). ¹⁰⁰ The self-judgment exception allowed the parties to refuse giving judicial assistance ‘[i]f the requested State considers that execution of the request is likely to impair its sovereignty, security, ordre public or other essential interests’. ¹⁰¹ In the 1990s, France seconded Bernard Borrel, a French national, to the Ministry of Justice of Djibouti. In 1995, the corpse of Mr Borrel was discovered, triggering investigations in Djibouti and France. The Djiboutian authorities concluded that Mr Borrel’s death was by suicide and terminated the investigation in 2003. However, the French investigation revealed connections to high-ranking officials in Djibouti. In 2004, after re-opening the investigation itself, Djibouti requested France to execute a letter rogatory and provide files relating to the enquiry into Mr Borrel’s death. The investigating French judge refused to do so, relying inter alia on Article 2(c) of the Convention. ¹⁰² In 2006, Djibouti filed a case against France at the ICJ, seeking a ruling that France was obligated to provide assistance.

France adopted a classic, broad understanding of the national security exception. It argued that the invocation of Article 2 was ‘left to the exclusive discretion of the State which intends to implement it’ and that it alone determines whether a request for mutual legal assistance is likely to prejudice its sovereignty, security, public order, or its other essential interests. ¹⁰³

Djibouti suggested that the Court should impose a justiciable good faith obligation on France regarding the application of the self-judgment provision and tried to ground its point in the general obligation to exercise discretionary powers in good faith. ¹⁰⁴ The consequence of following the French interpretation would be that the ICJ would be ‘automatically removed from the picture’, as it would simply be required to give effect to the French assessment. ¹⁰⁵ These factors, taken together, would render the Convention meaningless (‘simplement réduites à néants’) because each party would have the unqualified right to refuse to comply with its obligations without any possibility of being held legally accountable. ¹⁰⁶

The Court sided with Djibouti agreeing that it had the competence to conduct a good faith review of the self-judgment provision, Article 2(c) of the Convention. It held that Article 2(c) of the Convention provides a state with ‘very considerable discretion’ but ‘this exercise of discretion is still subject to the obligation of good faith codified in Article 26 of the Vienna Convention on the Law of Treaties’. ¹⁰⁷ In other words, the Court accepted that the self-judgment provision did not imply unfettered discretion for France but was tempered by a justiciable good faith obligation. Based on these holdings, the ICJ found that France had met the good faith requirement and accepted France’s invocation of the self-judgment provision in Article 2(c).

It was the first time that the ICJ acknowledged a limitation regarding the invocation of an explicit self-judgment provision, and it marked an important step in the development of self-judgment provisions from unreviewable exit routes to justiciable norms.

4.3.3 National security in international trade law

The second key area of judicialization concerns the national security exceptions under the GATT and later within the WTO. Today, there are national security exceptions under Article XIVbis of the GATS, Article 73 of the TRIPS and, most prominently, Article XXI of the GATT 1994. For long, Article XXI of the GATT had been considered the epitome of self-judgment provisions and has been referred to in this capacity in decisions of international judicial bodies outside of the GATT system. ¹⁰⁸ The provision reads:

Article XXI of the GATT 1994

Nothing in this Agreement shall be construed

(a) to require any contracting party to furnish any information the disclosure of which it considers contrary to its essential security interests; or

(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.

Its language is identical to Article XXI in the GATT 1947, the predecessor of the GATT 1994. ¹⁰⁹ The self-judgment language in paragraph (a) and (b) (‘it considers’) has been the source of protracted interpretative controversy and the question of limitations in the invocation of the provision arose already during the drafting process. ¹¹⁰ The Norwegian Chairman of the preparatory committee in Geneva, Eric Colban, seemed to consider the security exceptions beyond the purview of conventional legal safeguards. He opined that ‘the atmosphere inside the IO will be the only efficient guarantee against abuses of the kind to which the Netherlands Delegate [who had pointed out the vagueness of the provision] has drawn our attention’. ¹¹¹

Even though the national security exception was referred to several times in the decades following the inception of the GATT, the members always avoided a decision on the interpretation of the provision. ¹¹² Even when a dispute escalated to the extent that a panel was set up, the members avoided a ruling on the self-judgment provision. Most prominently, in a trade dispute between the United States and Nicaragua in the 1980s, the terms of reference of the panel excluded ‘the validity of, or motivation for, invocation of Article XXI(b)(iii)’ and the assessment of what measures had been ‘necessary’. ¹¹³ In other instances, for example in the EU-US dispute about the Helms-Burton Act, members relied on diplomatic means of dispute settlement rather than pursuing a (quasi-)judicial path. ¹¹⁴

While the majority of members maintained that the invocation of the national security exception was not subject to justiciable limitations, ¹¹⁵ scholars began arguing in favour of the good faith limitation as it began to gain traction in other areas of international law. ¹¹⁶ Influenced by this discourse and the turning tides before investment tribunals and the ICJ, good faith review reached the WTO in Russia – Traffic in Transit. ¹¹⁷ This dispute arose in the context of the deteriorating political relations between Ukraine and Russia in 2013–2014 and Russia’s subsequent annexation of Crimea and support of separatists in Ukraine. Russia and Ukraine implemented economic sanctions targeting each other. It was a bundle of such measures, including special transit requirements, which Russia had imposed on goods originating in Ukraine, that Ukraine challenged before a WTO panel in 2016.

Russia invoked the national security exception in Article XXI(b)(iii) of the GATT to defend its measures. Its central claim was that the invocation of Article XXI divested the panel of its jurisdiction. ¹¹⁸ This was based on the notion that the provision was ‘totally self-judging’ ¹¹⁹ and, therefore, conferred ‘sole discretion’ to the member invoking the exception as to the ‘necessity, form, design and structure’ of the measures taken. ¹²⁰ The United States, as a third party to the dispute, emphatically agreed, although it argued that the invocation did not concern the panel’s jurisdiction but was ‘non-justiciable’ as there were ‘no legal criteria by which the issue of a Member’s consideration of its essential security interests can be judged’. ¹²¹

Ukraine contested these arguments. It claimed that the self-judgment language did not mean that a member enjoyed ‘total discretion’ and argued that a panel’s ‘objective assessment must include an examination of whether a Member invoking Article XXI has done so in good faith’. ¹²² For Ukraine, good faith implied that the provision could not be invoked ‘to pursue protectionist objectives or to apply a disguised restriction on trade’. ¹²³ In a telling shift from earlier contentious debates regarding the national security exception under the GATT, other third parties ¹²⁴ that submitted their views argued that the panel possessed jurisdiction to rule on the invocation of Article XXI and that the dispute was also justiciable. ¹²⁵

The panel endorsed a judicialized understanding of Article XXI of the GATT by emphasizing its competence to conduct a review and developing a nuanced yardstick to do so. It found that the circumstances listed under sub-paragraphs (i)-(iii) (‘relating to fissionable materials’; ‘relating to the traffic in arms’; ‘taken in time of war’) were subject to an objective determination. ¹²⁶ In addition, the panel concluded that the self-judgment language only qualified the ‘necessity’ of a measure and what constitutes ‘essential security interests’. The self-judgment wording neither divested the panel of its jurisdiction nor made the matter non-justiciable. Additionally, the designation of essential security interests and its connection with the measures at issue taken under Article XXI of the GATT were subject to a good faith review. ¹²⁷ For the panel, the good faith review had two consequences. First, members were required to demonstrate the veracity of their claims, i.e., to provide reasons for the invocation of the provision. ¹²⁸ The second and, according to the panel, the ‘most importan[t]’ aspect of the good faith obligation, was that the measure taken had to meet a ‘minimum requirement of plausibility’ in relation to the proffered essential security interests. ¹²⁹

When applying the standard to the case, the panel accepted that an emergency in international relations existed objectively and that Russia had taken its measures under Article XXI of the GATT to protect its essential security interests in compliance with the good faith requirements. ¹³⁰ The crucial contribution of Russia – Traffic in Transit towards the judicialization of explicit self-judgment lies in the development of the good faith test, which significantly limits the leeway for members when invoking Article XXI. The proceedings also constituted a turning point in the positions expressed by other parties in their approach to self-judgment within the framework of the WTO. What had once been the paradigmatic expression of unfettered state power within a treaty framework, had now been reined in. Explicit self-judgment did not shield states from scrutiny in the international legal system. ¹³¹

When the matter came before the Dispute Settlement Body, both parties to the dispute accepted the ruling. Russia – with which the panel had sided in the final analysis – also welcomed the ‘historic contribution’ and the ‘guidance and clarity regarding the consequences of invoking’ Article XXI which the report had provided. ¹³² The United States considered the report ‘unpersuasive’ and ‘seriously flawed’. ¹³³

All proceedings which dealt with explicit self-judgment norms, including Russia – Traffic in Transit, had accepted the invocation of the self-judgment provision in the final analysis. This changed with the second WTO panel report on a national security dispute: Saudi Arabia – IPR. ¹³⁴ The self-judgment aspect of the case concerned Article 73 of the TRIPS, identical in language to Article XXI of the GATT. In the underlying dispute, Qatar claimed that Saudi Arabia had taken a number of measures to prevent a Qatari state-owned company, beIN, from obtaining legal representation in domestic proceedings via which it tried to prevent copyright infringements by a piracy broadcaster. Qatar also claimed that Saudi Arabia had refused to launch criminal investigations against the same copyright infringer. Saudi Arabia invoked Article 73 of the TRIPS, arguing that its acts and omissions were lawful because the severance of its diplomatic relations with Qatar constituted an emergency in international relations. ¹³⁵

Once again, many third parties intervened, largely emphasizing their support for the yardstick developed in Russia – Traffic in Transit to be transferred to Article 73 of the TRIPS. The United States maintained its position from the previous case that the panel should not hand down a decision. ¹³⁶ Bahrain sided with the United States and considered the invocation of Article 73 of the TRIPS of a self-judgment character and, therefore, beyond the competence of the panel. ¹³⁷

Saudi Arabia challenged the justiciability of its invocation of Article 73, claiming that the dispute was not a trade dispute but rather a ‘political, geopolitical and essential security dispute’. ¹³⁸ This was – unsurprisingly – rejected by the panel. ¹³⁹

In applying the good faith test, the panel accepted that Saudi Arabia’s severance of its diplomatic relations with Qatar met the minimum threshold to be objectively considered an emergency in international relations. ¹⁴⁰ In light of the international crisis between a number of Gulf states and Qatar, it also found that Saudi Arabia met the plausibility threshold to deny beIN legal representation because the measure constituted ‘an aspect of Saudi Arabia’s umbrella policy of ending or preventing any form of interaction with Qatari nationals’. ¹⁴¹

The panel did demonstrate, however, that the good faith limitation of self-judgment provisions can have teeth in practice when assessing Saudi Arabia’s refusal to apply criminal procedures to hold the piracy broadcaster accountable. In this regard, the panel found the ‘minimum requirement of plausibility in relation to the proffered essential security interests’ was not met. ¹⁴² The panel was ‘unable to discern any basis for concluding that the application of criminal procedures or penalties to [the piracy broadcaster] would require any entity in Saudi Arabia to engage in any form of interaction with beIN or any other Qatari national’. ¹⁴³ In other words, Saudi Arabia’s measure taken to further the proffered essential security interests was so unrelated to the emergency in international relations that it could not be considered a good faith measure under Article 73 of the TRIPS. It was the first time an international judicial body decided that the invocation of an explicit self-judgment provision did not entail the lawful application of the provision which the state had envisaged.

In distinction to Ukraine, Saudi Arabia appealed the panel decision. As the Appellate Body had become defunct by the end of 2019, the decision never acquired legally binding status and remains in the ether of international adjudication. Its holdings can nonetheless be considered an important application of the standards laid down in Russia – Traffic in Transit. The approach has since been broadly confirmed in several recent panel reports dealing with complaints against measures of the United States, which were found in violation of the WTO Agreements despite its invocation of the national security exception. ¹⁴⁴

4.4 Pushback: Reinforced self-judgment

How did states react to the judicialization of explicit self-judgment provisions? Two kinds of reactions can be discerned. One group of states seems to have accepted and internalized the judicialization of self-judgment exceptions. In the context of the WTO, this is first and foremost reflected in the arguments of third parties in the proceedings who argued in favour of the judicialization of the provision. ¹⁴⁵ When it comes to investment treaty practice, some states have recently included limitations to explicit self-judgment provisions, such as the obligation not to apply the provision in an arbitrary manner, similar to the good faith threshold. ¹⁴⁶

A second group of states have doubled down on their efforts to self-judge sensitive matters. After the presumption against self-judgment was established and the review of self-judgment provisions based on good faith began to gain traction, some states started amending self-judgment provisions to buttress the authority of the party invoking them vis-à-vis international judicial bodies. ¹⁴⁷ Notably, some states added ‘clarificatory’ footnotes to self-judgment provisions stating that a tribunal ‘shall find that the exception applies’ if one of the parties invokes it. ¹⁴⁸ As this practice further reinforces the self-judgment character of a provision it can be referred to as ‘reinforced self-judgment’.

Reinforced self-judgment provisions originated in the mid 2000s treaty practice of the United States and India. While the approaches of the two states differ in their wording, they converge in their objective: both are aimed at bolstering the authority of the state vis-à-vis international judicial bodies. The first agreement in which the practice of reinforced self-judgment materialized was the 2006 Peru-US Trade Promotion Agreement (TPA). ¹⁴⁹ Article 22(b) of the agreement reads:

Nothing in this Agreement shall be construed: to preclude a Party from applying measures that it considers necessary for … the protection of its own essential security interests. ¹⁵⁰

This can be considered a ‘classic’ self-judgment national security provision. In addition, however, the parties added a footnote to the provision. The footnote specifies that:

[f]or greater certainty, if a Party invokes Article 22.2 in an arbitral proceeding … the tribunal or panel hearing the matter shall find that the exception applies. ¹⁵¹

Similar provisions were included in the trade and investment agreements between the United States and Panama, ¹⁵² Colombia, ¹⁵³ and Korea. ¹⁵⁴ Notably, all of these agreements were concluded in 2006 and 2007 – shortly after the implicit self-judgment character of essential security interests provisions had been rejected in CMS v. Argentina and the LG&E v. Argentina tribunal had cast doubt on the traditional conception of self-judgment when endorsing a good faith review. ¹⁵⁵ States that had agreed to the first reinforced provisions in treaties with the United States in turn incorporated reinforced self-judgment in their own treaty practice. ¹⁵⁶ In this way, the US-variant of reinforced self-judgment through ‘clarifying’ footnotes has also migrated to Asia in recent years, for example in the 2014 Australia-Korea Free Trade Agreement (FTA). ¹⁵⁷

The second key example of reinforced self-judgment can be found in the treaty practice of India. Starting with its 2005 FTA with Singapore, ¹⁵⁸ India has added annexes or ‘exchanges of letters’ to selected trade and investment agreements. These additions fulfil a very similar purpose to the American ‘clarifications’. In combination with self-judgment national security provisions in the agreements themselves, the additions reinforce the authority of the treaty parties to make determinations beyond the review of the tribunal generally competent under the treaty. For example, India issued a joint interpretative note to its 2009 BIT with Bangladesh in 2017, stating that an invocation of Article 12 of the BIT ‘shall be non-justiciable in that it shall not be open to any arbitral tribunal to review the merits of any such decision’. ¹⁵⁹

The latest generation of Indian investment agreements, based on the 2016 Indian Model BIT, also suggests that the drafters considered the self-judgment security exception to be non-justiciable due to its reinforcement. A good example of this practice of reinforced self-judgment is Article 33 of the 2018 Belarus-India BIT. ¹⁶⁰ It contains a rather mundane self-judgment national security exception:

Article 33 Security Exceptions

33.1 Nothing in this Treaty shall be construed:

(i) to require a Party to furnish any information, the disclosure of which it considers contrary to its essential security interests; or

(ii) to prevent a Party from taking any action which it considers necessary for the protection of its essential security interests including but not limited to …

33.4 This Article shall be interpreted in accordance with the understanding of the Parties on security exceptions as set out in the Annex which shall form an integral part of this Treaty.

The relevant annex provides:

Where the Party asserts as a defence that conduct alleged to be a breach of its obligations under this Treaty is for the protection of its essential security interests protected by Article 33, any decision of such Party taken on such security considerations and its decision to invoke Article 33 at any time, whether before or after the commencement of arbitral proceedings shall be non-justiciable. It shall not be open to any arbitral tribunal constituted under Chapter IV or Chapter V of this Treaty to review any such decision, even where the arbitral proceedings concern an assessment of any claim for damages and/or compensation, or an adjudication of any other issues referred to such arbitral tribunal. ¹⁶¹

India has incorporated this approach in all of the new-generation BITs that it has concluded since 2018. ¹⁶² In an effort to implement the same approach in its BITs already in force prior to the development of its 2016 Model BIT, India has also begun to conclude ‘Joint Interpretative Notes’ on existing BITs, such as the Bangladesh-India BIT mentioned above, that affirm the same interpretation. ¹⁶³

As of now, reinforced self-judgment provisions seem to feature only in trade and investment treaties. This is likely the case because states have been particularly concerned about the expanding power of international judicial bodies in these fields from the 1990s onwards with the creation of the WTO and the proliferation of investment arbitration. ¹⁶⁴ With the Argentinian saga and the rise of national security litigation in the WTO, investment and trade have also emerged as areas where national security provisions – the most common self-judgment provisions – have most frequently been the subject of judicial decisions. But if states consider that international judicial bodies curtail their decision-making capacity too stringently in other areas of international law, reinforced self-judgment could spread further in the future. In this regard, one may recall attempts by states to regain control over the interpretation and application of human rights treaties. ¹⁶⁵

Arguably, reinforced self-judgment merely clarifies what the drafters intended. Yet, the attempt to construe self-judgment norms as non-justiciable is at odds with the good faith approach which has been increasingly accepted in recent years. ¹⁶⁶ Reinforced self-judgment as espoused in the interpretation of the United States and India serves as an instrument for states to re-capture authority from judicial bodies. As such, reinforced self-judgment can be understood as an attempt to push back against the judicialization of self-judgment provisions that occurred over the past years.

5. Conclusion

Three principal conclusions emerge from this article. First, self-judgment has been a site of continuous contestation. That the meaning of self-judgment is contested should not cause surprise. Norm contestation is a broad concept, describing practices aimed at changing the validity of a norm or how a norm is applied through objection. ¹⁶⁷ International and domestic norms are frequently contested, to the extent that Pierre Bourdieu considered contestation to be the essence of the practice of law itself. ¹⁶⁸ The observation is nonetheless meaningful because it captures the dynamic defining the legal concept of self-judgment.

Throughout this process, the judicialization of self-judgment has experienced pushback. As implicit self-judgment was judicialized, explicit self-judgment spread, and with the acceptance of good faith review, reinforced self-judgment emerged to push back the authority of international judicial bodies even more assertively.

A miniature history of self-judgment explained through this dynamic could be read thus: states originally conceived self-judgment as an instrument for states to exclusively exercise their authority. The presumption against implicit self-judgment could already be read as a – comparatively subtle – step towards judicialization. The burden on states to establish the self-judgment character of a norm was increased by requiring explicit self-judgment if they wanted to assert their authority vis-à-vis international judicial bodies and other states in particular areas. This is precisely what states did by increasingly incorporating explicit self-judgment provisions when drafting treaties. In response, international judicial bodies progressively judicialized explicit self-judgment by subjecting it to a good faith review. Consequently, one can again observe pushback through legal argumentation (e.g., in the WTO disputes) and, most crucially, through the practice of reinforced self-judgment. It emerges that the past decades have seen constant attempts to shape the understanding self-judgment.

Second, explicit self-judgment provisions no longer provide states with unfettered discretion in their application. This shift constituted the most important recent step towards judicialization. Self-judgment language is no longer enough to shield states from other states or international judicial bodies questioning their application of norms of international law. Although the WTO panels made little reference to the ICJ’s decision, the chronological development and the approach taken by the panels suggest that Djibouti v. France paved the way for this shift in the WTO context. This shift is also a move away from the ICJ’s dictum in Nicaragua v. United States, which at least implied the possibility of unfettered self-judgment and shaped the understanding of self-judgment for a long time. As a result, the good faith test developed in Russia – Traffic in Transit ‘severely limits the self-judging element’ ¹⁶⁹ of self-judgment provisions.

Third, the prevailing standard to review self-judgment provisions, which has crystallized as a tool of judicialization across different jurisdictions, is a ‘good faith review’: The careful endorsements in investment law, by the ICJ and by WTO panels that accepted the competence to review invocations of explicit self-judgment provisions, have all employed good faith review as a technique that promises to respect the particular wording of the norms while at the same time maintaining the normativity of the underlying provisions. Employing good faith has often been criticized as vague. ¹⁷⁰ However, the manner in which it has actually been applied by international judicial bodies – namely in the form of a purposive interpretation of the treaty provision and a light-touch review – falls squarely within the established methods of treaty interpretation.

If one is trying to discern a more general picture, reinforced self-judgment as well as the shift towards explicit self-judgment can be considered forms of pushback against the judicialization of self-judgment. While the majority of states seem to have endorsed or at least acquiesced to the judicialization of self-judgment in general, some states have pushed back against the concomitant shift of authority towards international judicial bodies. From a functional perspective, these states likely attribute greater weight to the flexibility that a traditional conception of self-judgment has provided them. The future of self-judgment hinges on which side the scales will tip between judicialization and pushback.