3.1 The Foundations of Customary International Law as a Source of International Law

The very existence of customary international law as a source of international law is largely a result of a logical deduction. In the early jurisprudence of the PCIJ and other bodies, no effort was made to establish consistent state practice and opinio juris for the existence of custom as a source of law,Footnote ²⁸ and instead the PCIJ and other authorities deduced the existence of custom.Footnote ²⁹ Similarly, when the International Law Association sought to complete its study on customary international law, it acknowledged that it would draw on the practice of states to identify the rules of customary international law, though that was not simply an exercise in using custom to support the existence and rules of custom itself.Footnote ³⁰ In the ILC study on customary international law, the special rapporteur relied almost exclusively on the case law of the ICJ, among other courts and tribunals, and cited to little state practice or opinio juris as expressed by states on point.Footnote ³¹ Of course, those cases that were cited might or might not have cited to state practice or opinio juris in turn. These approaches do not make the conclusions wrong; after all, reliance on prior judicial decisions is a valid subsidiary source for the content of any of the formal sources.Footnote ³² But such reliance on prior cases demonstrates that the identification of customary international law as a source of international law is not entirely established by state practice and opinio juris alone and is likely a logical deduction from the behaviour of states.

In turn, the very constitution of customary international law by state practice and opinio juris alone is also not clearly based on a logical inductive examination of state practice and opinio juris.Footnote ³³ Instead, the requirement of state practice and opinio juris is primarily derived from the PCIJ and ICJ Statutes, upheld by legal tradition and reaffirmed by, inter alia, the ICJ.Footnote ³⁴ From the language of the Statutes, it is not completely clear that customary international law is formed by state practice and opinio juris as elements.Footnote ³⁵ There is certainly space to argue that these aspects are factors or perhaps even evidentiary measures for custom.Footnote ³⁶ In fact, the process of identifying rules of customary international law often treats state practice and opinio juris more like factors or evidence than truly as elements.Footnote ³⁷ Nonetheless, they are commonly understood to be elements.Footnote ³⁸

3.2 The Deduction of Customary International Law without Resort to the Elements

Both the ICJ and PCIJ have suggested that some rules of customary international law might not be established by the elements of state practice and opinio juris.Footnote ³⁹ As noted above, the ICJ chamber in the Gulf of Maine case took note of some norms in international law that were not proved through induction but by their role in ‘ensuring … coexistence and vital cooperation’.Footnote ⁴⁰ The ICJ seems to be saying that there are some rules that are established through an inductive inference from state practice and opinio juris, but that there are other rules that are not.Footnote ⁴¹ This approach is not isolated. In the Corfu Channel case, the ICJ relied on ‘elementary considerations of humanity’ to identify rules of customary international law.Footnote ⁴² We could interpret this passage in an alternate way: that some rules of customary international law, while constituted by state practice and opinio juris, can be proved to exist by simply being logical for coexistence, vital cooperation or principles of humanity. And yet, the ICJ has also held quite clearly that customary international law norms cannot simply be deduced from ‘humanitarian considerations’ or ‘moral principles’.Footnote ⁴³

The ICJ takes several approaches to deducing the existence of customary international law. The first approach – what Talmon terms ‘normative deduction’ – is to infer the existence of a rule from other rules. For example, the ICJ has relied on the contents of treaties to prove the existence of customary international law.Footnote ⁴⁴ While it identified the right of self-determination in the Namibia advisory opinion, it used the UN Charter to interpret the rule as applying for all peoples.Footnote ⁴⁵ In a similar fashion, it has looked to UNGA resolutions to prove customary international rules.Footnote ⁴⁶ However, in both instances one could argue that the ICJ was not identifying a new rule of customary international law, but was instead using the UN Charter and UNGA resolutions to clarify the content of a rule through systemic interpretation. In the Territorial and Maritime Dispute, Nicaragua and Colombia agreed that customary international law would apply to the case because Colombia was not a party to the UN Convention on the Law of the Sea (UNCLOS), and agreed that the provisions in UNCLOS regarding the baselines and entitlement to maritime zones, exclusive economic zone (EEZ), and continental shelf reflected customary international law,Footnote ⁴⁷ though they disagreed over the content of the rules concerning the continental shelf beyond 200 nautical miles.Footnote ⁴⁸ It is possible to argue that the ICJ was merely affirming that Nicaragua had an obligation toward the other states parties to UNCLOS to comply with Article 76 when it referred to the ‘interrelated’ nature of ocean space.Footnote ⁴⁹ However, it did explicitly cite to obligations inter partes between states not parties to the dispute. In fact, it argued that UNCLOS created a new ‘legal order’, so Nicaragua was under this obligation in a dispute with a state not a party to UNCLOS.Footnote ⁵⁰ Thus, the failure of Nicaragua to comply with the procedural requirements of UNCLOS meant it had not established its continental margin vis-à-vis Colombia, a state not a party to UNCLOS.Footnote ⁵¹ Judge Robinson criticized this conclusion, arguing that the procedural condition was not part of customary international law and could not be applied.Footnote ⁵² What is not entirely clear in these cases is whether the ICJ was using the existence and content of treaties to prove the existence of a custom or clarifying the content of the rule.Footnote ⁵³

The ICJ also deduces rules of customary international law from other customary international rules. In the Burkina Faso/Mali Frontier Dispute, it held that uti possidetis was customary international law because ‘it is logically connected with the phenomenon of the obtaining of independence’.Footnote ⁵⁴ While this phrase might even suggest deduction of the rule from facts, actually the ICJ was deducing the customary rule of uti possidetis from the other customary rules governing independence. In the Qatar and Bahrain Maritime Delimitation, the ICJ set aside evidence of practice as inconclusive and instead deduced that ‘low-tide elevations cannot be equated with islands … or other land territory’ as they are understood under prevailing rules on territorial sovereignty.Footnote ⁵⁵ Thus, in these cases another norm of customary international law formed the framework for assessing the norm at issue. And while it is debatable whether the ICJ was identifying a new rule or clarifying an existing one, it appears to apply a teleological interpretation of other rules of customary international law. This approach should not be too shocking, because domestic courts sometimes take the same approach. In the case of A v. Attorney-General, the Swiss Federal Criminal Court held that the accused was covered by the same immunity enjoyed by a head of state because, as minister of defence, he was a member of the unusual collegiate body – the HCE, or Haut Comité d’État (High State Council) – that discharged presidential authority. The Federal Criminal Court did not examine state practice and opinio juris to determine whether a collegiate presidential council would enjoy the same immunities, but reasoned that logically it would, based on the purposes of head-of-state immunity.Footnote ⁵⁶

However, the ICJ has also deduced the existence of customary rules from other customary rules, not by logically deriving the rules from other rules but by analogy. In this approach, the ICJ is more clearly focused on applying customary international law rules that apply in other situations to a case where they clearly do not, and thus is more likely identifying a new rule rather than clarifying an existing one. For example, in the Libya and Malta Continental Shelf case, the ICJ identified a customary rule permitting 200 nautical miles of continental shelf by drawing an analogy with the exclusive economic zone.Footnote ⁵⁷ However, this approach is not without criticism. In the Arrest Warrant case, the ICJ deduced the existence of immunities from arrest for a foreign minster by considering the functions of the office.Footnote ⁵⁸ Unlike the situation of a head of state mentioned above, which clearly benefitted from a rule of immunity and needed clarification, this case asked whether the official had any immunities at all. Judge Van den Wyngaert critiqued the approach of identifying the new rule by deduction, calling it ‘a mere analogy with immunities for diplomatic agents and Heads of State’ without proper reliance on evidence of state practice and opinio juris.Footnote ⁵⁹

The ICJ also deduces customary international law from legal principles. There are at least two versions of this approach. One possibility, a more conservative one, is to deduce the existence of certain principles from treaties or other customary law and then take one more step and deduce the existence of rules from those principles. This is the approach that Talmon terms ‘triangular reasoning’.Footnote ⁶⁰ For example, in the Corfu Channel case, the ICJ identified the principle underlying the Hague Convention VIII of 1907 (i.e. the obligation to notify ships of minefields).Footnote ⁶¹ The treaty did not specifically apply to the situation of peacetime, and yet the Court stated that such a duty also applied in peacetime.Footnote ⁶² Similar to the practice of deduction by analogy, this practice appears to be focused on identifying new rules of customary international law.

One method for reasoning the existence of customary international law from abstracted principles, themselves potentially drawn from other treaties and customary law, is the necessity of the rule. As the ICJ said in the Gulf of Maine case, some rules are ‘vital’ or ‘ensur[e]’ international cooperation and coexistence, and in the Gabčikovo-Nagymaros Project case the ICJ stated that the rule was ‘require[d]’.Footnote ⁶³ The ICJ has stated that the reason for this view is that it should be reluctant to find customary international law restraining state action ‘[w]hen the stakes are not as high’,Footnote ⁶⁴ – that is to say, when the rules would be less necessary. That being said, the ICJ has been willing to deduce rules that are not absolutely necessary. Some rules are merely ‘important’Footnote ⁶⁵ and perhaps even ‘logically connected’ to other principles.Footnote ⁶⁶ In fact, some rules are simply reasonable expectations from other principles, such as the ‘elementary considerations of humanity’ cited in both the Nicaragua caseFootnote ⁶⁷ and the Corfu Channel case.Footnote ⁶⁸

The ICJ also occasionally applies an even more liberal approach, not necessarily identifying a treaty or customary law from which to draw legal principles before deducing the existence of customary international law. This is the case in Jurisdictional Immunities, where the ICJ noted the important principle of sovereignty and its derivatives sovereign equality and territorial sovereignty,Footnote ⁶⁹ and in Certain Documents, when the ICJ took note of the principles protecting the integrity of legal proceedings.Footnote ⁷⁰ Such practice can even be identified in the Lotus case, where the ICJ stated that ‘[a] corollary of the principle of the freedom of the seas is that a ship on the high seas is assimilated to the territory of the State the flag of which it flies, for, just as in its own territory, that State exercises its authority, upon it, and no other State may do so’,Footnote ⁷¹ and in the Pulp Mills case the ICJ observed that ‘a precautionary approach may be relevant in the interpretation and application of the’ treaty and then stated that the treaty ‘has to be interpreted in accordance with [what] … may now be considered a requirement under general international law to undertake an environmental impact assessment’.Footnote ⁷² Perhaps the ICJ could have identified treaties and other custom that evidenced such principles as applicable, but that step was omitted.

In other cases, the ICJ does not deduce the existence and content of a customary rule from a treaty or other source of law, nor deduce principles from those sources, but instead deduces customary rules from the logic and reasoning, the underlying purpose, of the rule itself. As such, the particulars of the rule applicable in a given case might be merely characterized as explaining the content of the rule in more detail, rather than identifying an entirely new rule. In the Western Sahara and Wall advisory opinions, the ICJ deduced from the existence of the rule of self-determination under customary international law that its full scope of application was erga omnes,Footnote ⁷³ again with little regard to state practice and opinio juris.

One the one hand, the ICJ appears to use deduction of rules quite often, yet, on the other hand, insists that it does not simply deduce rules.Footnote ⁷⁴ A possible explanation is twofold. First, in many of these cases the ICJ is not necessarily deducing the existence of new rules but rather interpreting the content, scope, and application of existing rules. For example, in the Arrest Warrant case, one could characterize the rule as being immunity for certain high state officials, with head of state being a clearly established example and foreign minister a less clear example. The ICJ did not discover a new rule of immunity of foreign ministers, but instead clarified that the scope of high state official immunities covered them. State practice and opinio juris was not necessary for these narrow cases, because there was already well-established state practice and opinio juris for the functional category in which foreign ministers fall. In a similar manner, the extension of customary international humanitarian law to cyber operations might not call for supplementary evidence of state practice and opinio juris, because there is already solid evidence that these rules apply to armed conflict through whatever mode it is manifested.Footnote ⁷⁵ This view can be justified by recalling the distinction between the identification and the interpretation of customary international law. Where the ICJ is identifying a new and distinct rule, perhaps it must locate state practice and opinio juris, but when it is merely interpreting the content of the rule, it is freer to use other legal interpretative techniques such as analogy and the context of other legal rules. That being said, the line between identifying a new rule and clarifying the content of the rule is so fine that the distinction is admittedly difficult to draw, potentially opening the door to more flexible interpretative techniques. Given the ICJ’s affirmations of the state practice-opinio juris rule, we could conclude that when it applies deduction in this manner, it is presuming what state practice and opinio juris it would expect to find, had it looked.

Second, we can also recall the distinction between the formation of customary international law and evidence for the formation. Several authorities have argued for the use of deduction when determining whether there is sufficient evidence for customary international law. The ICJ could be deducing the probable existence of certain rules and then lowering the threshold of evidence to establish the rule (perhaps even reducing it to zero in clearly obvious cases), rather than deducing the existence of customary international without any regard to state practice and opinio juris. In essence, some norms are simply more obvious,Footnote ⁷⁶ because they are logical or necessary for international cooperation. In his dissent in the North Sea Continental Shelf cases, Judge Lachs stated that deduction is an alternative to induction when the norms are difficult to prove.Footnote ⁷⁷ Certainly, the ICJ appeared to take this approach in the Asylum,Footnote ⁷⁸ Nicaragua,Footnote ⁷⁹ and Qatar/Bahrain Maritime Delimitation cases.Footnote ⁸⁰

However, the ICJ does not confirm that either of these explanations is correct. More often, it dispenses with a lengthy explanation, or perhaps any explanation at all, of the evidence on which it based its decision, leaving obscurity over whether deduction was the sole basis for the rule, a supplementary basis, or a reduction in the burden of proof through inductive means.Footnote ⁸¹ For example, in the Arrest Warrant case the ICJ simply stated ‘it is firmly established that’.Footnote ⁸² Use of the term ‘established’ can probably be assimilated to ‘proved’, although even if it was a synonym for ‘constituted’, the ICJ is still referencing confidence in its existence, which is an evidentiary matter. In the Gulf of Maine case, the ICJ expressly stated that there were two classes of rules of customary international law, proved by induction and by deduction.Footnote ⁸³ In the Corfu Channel case, the ICJ quite clearly drew on considerations of humanity as the reason for its conclusion, not as a basis for deducing the scope of another rule of law.Footnote ⁸⁴ If it truly was simply a lower evidentiary burden, one would expect the ICJ to use different phrasing to describe what it was doing. While we might attempt to overlook the Corfu Channel approach because it emerged during an era when international law might not have been well developed, the Gulf of Maine case was decided in 1984.

Most likely, the best explanation is that, in principle, state practice and opinio juris are required in all cases. However, when a norm is necessary (or at least logical or sensible) for the international legal system, or a parallel treaty is in place that governs the parties, or the question is really just a clarification of the scope of an already existing general rule of customary international law, then the ICJ can deduce that it is safe to presume state practice and opinio juris, and an inductive analysis with evidence is not necessary. The rule is simply very obvious in such cases. Other, more novel, situations demand stronger evidence of state practice and opinio juris. While the distinction between these two approaches is more likely a spectral one, reducing the need for compelling evidence in proportion to the reasonableness that the rule exists, the ICJ in Gulf of Maine identifies the two ends of the spectrum. Thus, the ICJ is not necessarily inventing customary international law from whole cloth by pure deduction, assertion, or arbitrariness, but instead identifying rules that are more likely than not to exist, and more likely than not to have certain content, when operating within an environment of other rules of international law.

3.3 Deduction of Customary International Law within the Application of the Elements

Of course, this analysis means that the ICJ still makes use of inductive reasoning. In many cases, it does apply state practice and opinio juris as elements that are used to identify legal rules. Following from Section 3.2, where logic and reason could be justifications for an abridged inductive analysis, this section discusses the assessment of evidence for customary international law in the inductive process, and again we find the use of logic throughout the process.

Initially, it could be that the states disputing a matter might simply agree on the existence of a rule of customary international law.Footnote ⁸⁵ If they stipulated this rule, then it would govern the dispute and, especially if used as the basis for a decision by the ICJ, would control the outcome. This outcome would stand, even if the norm of customary international law later appeared to be unfounded in state practice or opinio juris. After all, some of the votes of the ICJ on questions of the existence of a rule of customary international law have been rather close.Footnote ⁸⁶ For its part, the European Court of Justice has also taken the position that the parties must contest the existence of a customary international law rule or it can be presumed to exist.Footnote ⁸⁷ Pellet and Gaja have both argued that acceptance of a rule of customary international law is not required by states, so whether states accept the existence of such a rule in a dispute between them is not a requirement.Footnote ⁸⁸ The question in this chapter, however, is the reverse. While acceptance is not a prerequisite for the existence and application of the rule, in the sense that refusal could block creation and application of the norm, acceptance does appear to be constitutive of the norm, at least insofar as the norm is applied in the dispute and serves as precedent for disputes that follow. Pellet argues that acceptance is not constitutive, but instead merely eases the burden of establishing the existence of the rule in the dispute.Footnote ⁸⁹ Here again, the distinction between the existence of a rule and the evidence of a rule appears.

In a similar fashion, for some states the views of their foreign ministry might preclude any ascertainment of customary international law. A state’s internal determination by its foreign ministry on the existence of the customary rule might be determinative that a rule exists under customary international law, be it in operationalizing its foreign policy or for settling a dispute of international character within its domestic legal system.Footnote ⁹⁰ Again, such a determination might not be supported by adequate state practice and opinio juris. Such a decision by a domestic court might in turn be used elsewhere as evidence of the existence of the norm. The ILA concluded that the differing views of the executive or legislature on a customary international law norm would have different weight, but that neither was necessarily excluded.Footnote ⁹¹ However, when the question of the norm is raised within a domestic situation, the executive’s position might be binding on a court.

When states disagree over the existence of a rule, then the ICJ may take a more scientific inductive approach, but this approach should not blind us to the use of deductive logic within the application of induction. The dispute serves to frame the question about the rule and, as such, already impacts the form of the rule, as states suggest competing hypotheses and the decision-maker needs to adopt one or the other hypothesis.Footnote ⁹² Important to note, however, is that the competing hypotheses are themselves not necessarily the outcome of inductive logic but could be based on deduction from other principles or aims. Evidence of state practice and opinio juris that does not support one or the other hypothesis can be disregarded.Footnote ⁹³ In this regard, Judge Van den Wyngaert’s dissent in the Arrest Warrant case is important, for there she challenged the framing of the hypothesis by the ICJ:

Whether there is any presumption in favour of or against the formation of a rule of customary international law is also deduced logically. In principle, we might apply the Lotus presumption against states binding themselves, but the ICJ has suggested that UNGA resolutions on point can create a presumption in favour of customary international law.Footnote ⁹⁵ Instead of a presumption, it might simply be that UNGA resolutions are weighty evidence.Footnote ⁹⁶ Also consider in the Navigational and Related Rights case, where the ICJ found the existence of particular customary international law because Nicaragua failed to object to a long-undisturbed practice.Footnote ⁹⁷ It is difficult to understand how the ICJ believed a Lotus presumption against custom had been overcome in Navigational and Related Rights with such weak opinio juris and – as it admits – not well-substantiated practice. In essence, the ICJ used logic to interpret the meaning of silence.

Logical deduction is also largely the basis for determining the types and forms of evidence that the ICJ and other authorities accept for the purpose of identifying customary international law. As a preliminary matter, the limitation of practice to acts of states, rather than international organizations or other actors, is largely traditional. The ILC special rapporteur cited little actual practice of states to support his conclusion that international organizations contribute to customary international law ‘[i]n certain cases’ and that we should approach the question with ‘caution’.Footnote ⁹⁸ In addition, whether actors within a state engage the state’s role in constituting customary international law is often assessed by analogy to either the Draft Articles on State Responsibility rules on attributionFootnote ⁹⁹ or the Vienna Convention on the Law of Treaties rules on capacity to bind the state.Footnote ¹⁰⁰ There is no clear classification of which acts are ‘practice’,Footnote ¹⁰¹ and states often suggest which acts qualify as practice in yearbooks or digests on practice.Footnote ¹⁰² Indeed, sometimes the ICJ appears to reduce the evidentiary burden for proving customary international law by invoking strong, convincing authorities such as the ILC.Footnote ¹⁰³ One would imagine that a state would need to bring quite compelling evidence to convince the ICJ to determine that a provision in the Draft Articles on State Responsibility was not customary international law. Beyond these basic conditions, the ICJ and other authorities do not perform any inductive analyses of the secondary rules on the nature and quality of the practice being ‘sufficiently widespread and representative, as well as consistent’.Footnote ¹⁰⁴ Again, we can note the lack of inductive analysis also on the part of the ILC special rapporteur in reaching his conclusions on these points.Footnote ¹⁰⁵

After offering competing hypotheses, the ICJ must weigh the evidence. There does not appear to be any inductive analysis on the burden of proof or weight of evidence for customary international law.Footnote ¹⁰⁶ The ICJ has only issued very vague statements that evidence is weighed and contradicted.Footnote ¹⁰⁷ The Special Tribunal for Lebanon used the expression ‘beyond any shadow of doubt’ when considering whether a customary rule had been proven, though it did not indicate whether this was the burden of proof it applied or whether this was merely a rhetorical flourish.Footnote ¹⁰⁸

Having produced a final pool of data on the basis of all of these logical steps, the ICJ will perform an inductive inference to identify the applicable rules (or least says that it will). The usual approach, in line with an inductive logic, is to employ a sampling survey. The actual sampling conducted by the ICJ is not random or following typical scientific approaches for induction.Footnote ¹⁰⁹ For example, the practice is not simply quantified based on the number of states in alignment, but certain acts by certain states might be weighted more heavily.Footnote ¹¹⁰ For example, the judges in the Legality of the Threat or Use of Nuclear Weapons advisory opinion disagreed, not about the methodology but about relative significance of the practice.Footnote ¹¹¹ In both situations, the judges are applying logic to the value of practice.

Some cases will result in insufficient or inconsistent evidence. Talmon has observed that in such cases the ICJ might simply deduce the existence of the rule. He reads the Gulf of Maine case to say that the lack of evidence could lead the ICJ to rely on a deductive approach.Footnote ¹¹² The present author, however, considers that the chamber in the Gulf of Maine case might have only identified a certain class of norms that are reasonable as to their content and, as a result, imposed a lower burden of proof.Footnote ¹¹³ That is to say that the determination of the content of the rule preceded the identification of method,Footnote ¹¹⁴ whereas Talmon reads the chamber judgment as holding that the method of proof came first, and that only when it failed was an alternate method adopted.

Lastly, although it is agreed that, in principle, one applies inductive reasoning to infer the existence of a rule from examples of practice and opinio juris, the precise form of the inference remains somewhat unclear. The inference could be an enumeration of cases and inverse inference, concluding that the frequency of a certain characteristic in the sampled population is the same as in the general population.Footnote ¹¹⁵ The inference might instead be a predictive inference. This approach attempts to use past examples of actions to predict likely behaviour in the future,Footnote ¹¹⁶ or uses past practice to identify an underlying characteristic that gives cause to the behaviour and predicts behaviour when that characteristic is present.Footnote ¹¹⁷ Or, in a similar approach, the investigator could engage in universal (hypothetical) inference, proposing a rule that applies to the general population, due to it being exhibited in the sample pool.Footnote ¹¹⁸ In the usual adversarial approach, a court practices Bayesian confirmation, where it identifies which of the hypotheses is the better explanation for the evidence,Footnote ¹¹⁹ though it is certainly possible that a court or other authority could develop its own hypothesis. Some language from the ICJ suggests that it is applying Bayesian confirmation of a universal inference. For example, in examining the evidence in Jurisdictional Immunities case, the ICJ merely observed that practice had not been contradicted with other evidence of practice.Footnote ¹²⁰ In any event, one cannot use the sample pool merely to describe the likely state practice exhibited and opinio juris held by states; it must identify what custom has been prescribed.Footnote ¹²¹ These inductive approaches can only result in descriptive outcomes, and one must still make a normative inference from the descriptive outcome.

2.1 ‘Control’ in International Law

The best-known test for determining whether a state has enough control to be held responsible over the actions of other entities is the ‘effective control’ test set by the ICJ. In all cases where the ICJ has discussed the issue of attribution to states of actions committed by other entities, it has found that states could only be held responsible for the actions of actors that they effectively controlled. In the case of Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. USA), the ICJ established what it considered to be the threshold of control for attribution of conduct to states based on the rule expressed in Article 8. Nicaragua had turned to the Court claiming that the United States was responsible for supporting armed actions by a non-state armed group known as the Contras in Nicaraguan territory. The Court concluded that the United States should be held responsible for supporting the Contras. However, it also found that it could not hold the United States responsible for illegal actions committed by the Contras in their armed operations because, although the United States played a relevant role in the Contras’ operations, it did not have effective control over them.Footnote ⁹ The ICJ later reiterated its understanding that attribution under Article 8 of the ARSIWA would require evidence that a state has effective control over the actions of a non-state actor in the Armed Activities in the Congo caseFootnote ¹⁰ and in the Bosnia v. Serbia case.Footnote ¹¹

Other tribunals read the ICJ’s interpretation of control as the standard customary law test for attribution of conduct under Article 8 of the ARSIWA.Footnote ¹² However, the solution found by the ICJ was not the only possible one. For example, those operating in specialized fields of international law have found that it was sometimes necessary to adopt a lower threshold of control in light of the legal and practical circumstances of the cases they were handling.

This flexibility is illustrated by the International Criminal Tribunal for the Former Yugoslavia (ICTY), which attributed the conduct of armed forces to states by making use of what became known as the ‘overall control’ test. In the Tadić case, the ICTY acknowledged the ‘effective control’ test developed by the ICJ, but found that its high threshold was incompatible with the strict hierarchy under which military operations take form. According to the ICTY: ‘The requirement of international law for the attribution to States of acts performed by private individuals is that the State exercises control over the individuals. The degree of control may, however, vary according to the factual circumstances of each case.’Footnote ¹³

Following this logic, the ICTY concluded that the ‘effective control’ test set by the ICJ was suited for situations where it was necessary to determine whether the actions of a private individual (or group of individuals) are attributable to a state. In these cases, the ICTY stressed, there must be evidence that such an individual received from a state specific instructions to act as its de facto agent.Footnote ¹⁴ However, such detailed examination would not be necessary when the state already has authority over an individual or group that operates under an organized hierarchy. Organized and hierarchical structures, such as military units, operate under more predictable, preexisting logics and strategies and therefore do not require a threshold as high as the one expressed by the ICJ in Nicaragua v. USA. The ICTY clarified that:

The ICJ responded specifically to the Tadić test in its Bosnia v. Serbia judgment, suggesting that the approach taken by the ICTY in this case was ‘unpersuasive’ and dismissing the possibility of applying it in the context of state responsibility.Footnote ¹⁶ The ICTY, however, kept following its own test as set in Tadić,Footnote ¹⁷ with defenders of this approach arguing that both tests can coexist because they are designed to deal with different circumstances.Footnote ¹⁸

The peculiarities of state-owned enterprises may also justify a specific approach to establish that a state controls them. In its commentary to Article 8 of the ARSIWA, the ILC discussed the attribution of conduct of enterprises that are owned and controlled by the state as an example of a situation that could warrant a threshold different from that used in ICJ case law. According to the ILC, the conduct of a state-owned enterprise should be attributed to the state pursuant to Article 8 of the ARSIWA when there is ‘evidence … that the State was using its ownership interest in or control over a corporation specifically in order to achieve a particular result’.Footnote ¹⁹

The ILC’s commentary relied on an example from the case law of the Iran–US Claims Tribunal (IUSCT) on the attribution of the conduct of state-owned enterprises. The IUSCT, a tribunal created as part of a set of measures to manage the political crisis between the United States and Iran arising from the 1979 hostage crisis and the subsequent freezing of Iranian assets by the United States, decided several cases that involved the attribution to the state of the actions of state-owned enterprises. These decisions were grounded on interpretation of Article VII(3) and (4) of the Claims Settlement Agreement, which provided for the attribution to Iran and the United States respectively of the acts of the entities they controlled.

In most cases, the IUSCT attributed the actions of state-owned enterprises to the state only if there was proof that the latter was using its controlling stake to direct the actions of the enterprise.Footnote ²⁰ However, in cases such as Foremost Tehran v. Iran the tribunal’s decision on attribution rested on the corporate characteristics of an enterprise – namely, ‘the identity of its shareholders and the composition and behaviour of its board of directors, which must be examined together’,Footnote ²¹ rather than on the specific interactions between the state and the corporations.Footnote ²² The reason for such variation was the position taken by the minority of judges, as stated in Judge Khalilian’s opinion in the case of SEDCO v. Iran:

The ILC, when examining the IUSCT case law, focused on the majority cases where the tribunal looked for evidence that the state had used its corporate power to push the state-owned enterprise towards adopting specific actions.Footnote ²⁴ The ILC found these majority cases representative, not only of the IUSCT practice for dealing with state-owned enterprises but of the position of international law in general.Footnote ²⁵

These examples show how there are multiple ways of interpreting and applying Article 8 of the ARSIWA. The term ‘control’, in the way that it is used in the ARSIWA, or even the ‘effective control’ terminology used by the ICJ, are polysemic terms – they do not encompass a single type of relationship between states and non-state entities. It is up to those applying Article 8 to examine whether each situation they face meets the conditions for attribution under that provision. However, this case-by-case approach does not mean that Article 8 should be interpreted indiscriminately or inconsistently. In fact, all bodies examined in this section have sought a consistent application of Article 8 when dealing with similar sets of facts. Consistency in these cases is achieved not by applying the exact same text, but rather through constant judicial dialogue. This logic can be seen in the intense dialogue between ICJ and the ICTY about whether there should be a single control test, and if so, which should be the correct one. Both concluded that there were different possible interpretations for the term ‘control’ and that the test that each of them had been applying so far was the correct one for the specific circumstances they were facing.Footnote ²⁶ Similarly, in the IUSCT the judges that were in the minority in cases such as SEDCO engaged in internal judicial dialogue when dissenting from its own case law.Footnote ²⁷

3.1 Caution towards State-Led Initiatives in the Development of International Investment Law and Investment Arbitration

Investment arbitration is often guided by a general cautiousness towards state-led initiatives. International investment law was developed around the principle of investment neutrality, which assumes that state involvement in investment relations may generate political risks that harm foreign investment.Footnote ⁴⁷ The term ‘political risk’ is not defined legally but is used in investment and business circles to refer to risks associated with political fluctuations such as changes of power and policy. Many of the protections contained in investment agreements are aimed at reducing political risks that foreign investors may face in the host state – for example, most-favored-nation and fair-and-equitable-treatment clauses. Most-favored-nation clauses prevent the host states’ relationships with other states from interfering with investment relations by allowing foreign investors to claim most-favored-nation treatment.Footnote ⁴⁸ Fair-and-equitable-treatment provisions require foreign investors to be treated ‘fairly and equitably’Footnote ⁴⁹ – terms that have been interpreted as a blanket protection for foreign investors against host state actions that seem to be politically motivated or ‘irrational’.Footnote ⁵⁰

This logic of limiting state power also applies to investment arbitration. Investors often feel that the host states’ domestic courts – the default forum for them to settle these disputes – may be biased against foreign investors or lack the expertise to deal with investment treaty claims. Even before clauses providing for arbitration as a means to settle disputes started to be introduced in investment treaties, states and investors were already reaching out to organizations such as the World Bank to request informal assistance to settle their disputes.Footnote ⁵¹ At the same time, the international dispute settlement options that existed then were limited and politically sensitive. To start a case against the host state, investors had to convince their home states to exercise diplomatic protection. This forced investors to spend significant time and political resources without any guarantee that the case would be picked up by their home states or lead to a desirable outcome.Footnote ⁵² Investment arbitration as a mechanism of dispute settlement that investors can use directly, without having to appeal to their state of origin, was thus considered a more politically neutral option than both domestic dispute settlement and diplomatic protection.Footnote ⁵³

In addition to depoliticizing the access of investors to international dispute settlement mechanisms, there is also the goal of depoliticizing the resolution of disputes themselves. Investment law often deals with subjects that are highly sensitive to states,Footnote ⁵⁴ and one of the central goals of investment arbitration is to reduce state influence over litigation outcomes.Footnote ⁵⁵ This includes, for example, prohibiting investors’ home states from directly engaging in investment disputes, as provided in Article 27 of the ICSID Convention.Footnote ⁵⁶

The ICSID Convention also prevents states from making use of investment arbitration as investors themselves. The idea of having investor-states as parties to investment arbitration was raised and rejected during the negotiations of the ICSID Convention. Many of the states involved in negotiations feared that allowing states to access investment arbitration as investors would facilitate political capture of the system, which would defeat the purpose of creating a depoliticized system.Footnote ⁵⁷ Article 25(1) of the ICSID Convention, which establishes ICSID jurisdiction, therefore excludes the possibility of interstate disputes:

Similarly, the early development of investment arbitration seems to have embedded in its discussions the idea that, although formally separate from states, state-owned enterprises would probably not be eligible claimants in investment arbitration because they operated under state control. Early ICSID drafts determined that, for jurisdictional purposes, state-owned enterprises should be equated to states.Footnote ⁵⁸ The final draft of the ICSID Convention also allowed states to consent to foreign investors filing claims directly against their state-owned enterprises, therefore presuming that those enterprises were part of host states themselves.Footnote ⁵⁹ Even today, there are investment treaties with provisions limiting the protections afforded to state-owned investors and, as a consequence, their capacity to resort to investment arbitration.Footnote ⁶⁰

However, the creators of this dispute settlement system were aware that these preconceptions were subjective and could not reasonably justify a blanket prohibition against state-owned claimants making use of investment arbitration. When faced with the question of whether state-owned enterprises should be excluded from filing investment arbitration cases, Aaron Broches, former general counsel of the World Bank and one of the main sponsors of the ICSID Convention, acknowledged that they could play distinct roles in an economy, but seemed to believe that truly independent and market-oriented state-owned enterprises were an exception. In Broches’s words, ‘[t]here would appear to be no reason why a state-owned enterprise, if it elected to assimilate itself to a private enterprise rather than a government agency’Footnote ⁶¹ should be treated differently from a privately owned corporation.

Most of the discussions on state-owned enterprises in investment law were, however, informal and were not translated into explicit treaty provisions. Broches’s writings on the ICSID negotiations indicate that the drafters of these treaties may have thought that it was not necessary to add such a level of detail on state-owned enterprises, since they shared a ‘common understanding’ on the subject. This understanding was that although state-owned enterprises could not be automatically equated to states, most of them presented a risk for a free market economy because they operated under strict state supervision. This notion, however, covered only the political behaviour and characteristics of state-owned enterprises.

It is worth noting that this tendency to overgeneralize treaty terms on state-owned enterprises also prevails in modern investment treaties. With the exception of the EU–China Comprehensive Agreement on Investment, which specifically lists the entities covered by its provisions,Footnote ⁶² the few treaties that contain provisions on how to distinguish state-owned from other corporations or from states still do so in very broad terms, such as ‘when it exercises any regulatory, administrative or other governmental authority’.Footnote ⁶³

While the legal instruments shaping international investment law do not provide much guidance on how to deal with legal questions involving state-owned enterprises, they show that stakeholders engaged in developing international investment law and investment arbitration were cautious towards recognizing rights of state-owned enterprises as investors. As stressed by the tribunal in Maffezini v. Spain, the overall framework of international investment law presumes that state-owned enterprises will, as a general rule, operate alongside states and even as part of their structure.Footnote ⁶⁴ In the extreme, relying on such notions may lead to decisions such as the one taken by the tribunal in Bayindir v. Pakistan, which concluded that it could attribute to a state the actions of any enterprise that it owns, regardless of there being specific evidence showing that the state indeed exercised control over that entity.Footnote ⁶⁵

3.2 Using ‘Effective Control’ as a Flexible Standard

A second element shaping how investment arbitration approaches the attribution to states of state-owned enterprise actions is the flexibility with which investment tribunals have approached the term ‘effective control’. Tribunals have relied on the terminology of effective control in various contexts and referring to different thresholds of control, including corporate control. In the context of investment arbitration, the term ‘effective control’ has been used to indicate that a tribunal must look at the facts of each case, rather than relying on general formulas that presume control.

The legal issue with regard to which the tribunals show the greatest flexibility in the interpretation of the standards of effective control is the determination of the nationality of foreign investors. As stated in Article 25(1) of the ICSID Convention, modern investment arbitration is supposed to be between a state and a national of another state. Therefore, a corporate foreign investor will be allowed to file an investment claim on its own only if it proves that it is a national of a state party to a relevant treaty. However, it is not easy to determine the nationality of an investor, since many host states require investments to be made through local subsidiaries that would not normally qualify as foreign investors. In general international law, the nationality of a corporate investor is determined on the basis of its place of incorporation,Footnote ⁶⁶ which makes it impossible for foreign shareholders to file claims seeking relief for damages suffered through a local subsidiary. Aware that this would hamper the access of transnational companies to investment arbitration,Footnote ⁶⁷ the drafters of the ICSID Convention included in Article 25(2)(b) the possibility for states to agree to treating local subsidiaries controlled by foreign corporations as foreign nationals:

Many investment treaties thus include in their dispute settlement chapters provisions that also treat as foreign investors subsidiaries that are controlled by a national of a foreign state.Footnote ⁶⁹ However, such provisions require proof that the claimant parent company has corporate control over the local subsidiary.Footnote ⁷⁰ Although the most obvious form of corporate control would be to hold more than half of the subsidiary’s shares, there are many ways of controlling a corporation – for example, through special voting rights, golden shares, or joint ownership, to name but a few. Perhaps to avoid being blindsided by looking only at majority shareholding, tribunals have treated as nationals of a state claimants that are under the ‘effective’ corporate control of an entity established in that state.Footnote ⁷¹

Despite the use of terminology identical that used by the ICJ in the Nicaragua v. United States case, the test used by arbitral tribunals to determine the nationality of investors could not be more different from the one employed by the ICJ. Whereas the ICJ test focused on whether a state had de facto control over another entity’s specific actions, the test used by arbitral tribunals to determine the nationality of investors involves an inquiry into the company’s corporate control mechanisms. In the case of Rompetrol v. Romania, the tribunal examined the successive changes in the ownership of Rompetrol since 1993 to determine who had ‘effective control of the corporate entities in question’.Footnote ⁷² The Vivendi I annulment committee adopted a similar approach to examine whether a transfer of control over one of the claimants could affect its nationality, stressing that a company could be considered a foreign national only if it was ‘effectively controlled, directly or indirectly, by nationals of one of the contracting parties’.Footnote ⁷³

Neither of the tribunals in the cases of Rompetrol or Vivendi I referred or drew connections to Nicaragua v. United States or to Article 8 of the ARSIWA. This raises the question of whether these two investment tribunals had the intention to apply and interpret the ICJ’s effective control test through the lens of investment law when adopting such terminology. Nevertheless, the use of the term ‘effective control’ has led other tribunals to treat such case law as part of a common analytical framework. In Pan American Energy and BP Argentina v. Argentina, the claimants and the tribunal referred to case law on the determination of nationality and on attribution of conduct to justify the use of corporate control as a criterion to determine nationality.Footnote ⁷⁴ The resulting award thus combined the case law on ‘effective control’, thereby creating another new test for attribution of conduct under these criteria.Footnote ⁷⁵

Investment tribunals therefore approach the interpretation of the term ‘control’ under Article 8 of the ARSIWA from a rather unique perspective. In the absence of a distinction between ‘corporate nationality control’ and ‘control over non-state actor’s conduct’, the concept of ‘effective control’ ceases to be a specific threshold for control in the context of investment arbitration. Rather than using it as a reference to a threshold of control, investment tribunals seem to treat the ‘effective control’ standard as translating an obligation to examine each case individually and to evaluate different legal approaches to reach a solution that is appropriate to the specific circumstances of the case.

3.3 Judicial Reasoning as a Compass for Determining a Threshold for Control

The flexibility with which tribunals have approached the ‘effective control test’ is not exceptional in investment arbitration. As argued elsewhere, investment arbitration tribunals approach situations on a case-by-case basis, tailoring their legal solutions to the characteristics of the agreements at hand and the specific needs of the parties to the dispute.Footnote ⁷⁶ Faced with the question of attribution of state-owned enterprise conduct to states based on Article 8 of the ARSIWA, investment tribunals will therefore determine what threshold of control to apply in each case by calibratingFootnote ⁷⁷ standards used in general international law and the specific characteristics of the international investment law applicable to the case at hand.

An examination of the way in which investment arbitration tribunals applied Article 8 of the ARSIWA raises the question not only of the threshold of control that they apply but even more so the elements that they take into consideration to determine the threshold of control applicable to each case. A tribunal that rules on whether to attribute the actions of a state-owned enterprise to a state pursuant to Article 8 of the ARSIWA will have to consider whether the (very high) threshold of control set by the ICJ is appropriate to the circumstances of the case at hand,Footnote ⁷⁸ including the rules agreed upon by the treaty parties; the common practices in the field;Footnote ⁷⁹ the treaty parties’ laws on state-owned enterprises;Footnote ⁸⁰ and evidence of the involvement of the treaty party in the state-owned enterprise or in events related to the case.Footnote ⁸¹

The highest threshold of control that a tribunal could adopt would be the one applied by the ICJ in the Nicaragua v. United States case, which holds states responsible only for actions that they have specifically controlled by, for example, ordering or directing the state-owned enterprise to act in such a way. The lowest threshold would be to attribute to the state the actions of any entity in respect of which it has participatory decision-making power, regardless of whether it is indeed involved in the decisions taken concerning that entity. Between these two extremes there is a range of other possible thresholds that may be adopted depending on the context of a case. A tribunal may, for example, hold states responsible for the actions of non-state entities when they have evidence that the state formally cleared or authorized those actions. All of these approaches are in principle allowed by international investment law and investment arbitration, as arbitral tribunals have the power to interpret and apply international law to solve the conflict they have at hand.

Tribunals’ powers to interpret and apply open-ended rules such as the one expressed in Article 8 of the ARSIWA is tied to an obligation to provide reasoning. Article 48(3) of the ICSID Convention requires tribunals to state the reasons upon which they have made their decisions. Article 34(3) of the UNCITRAL Arbitration Rules contains a similar obligation, which applies unless the parties have agreed otherwise. In ICSID proceedings, an arbitral award that lacks an account of the tribunal’s reasoning may even be annulled pursuant to Article 52(1)(e) of the ICSID Convention. The amount of detail that a tribunal must give to fulfill its obligation to provide its reasoning will vary from case to case. As noted by the annulment committee in Wena, there is no single compulsory format in which the tribunal must present its reasons, and it can even leave its reasons implicit, ‘provided that they can be reasonably inferred from the terms used in the decision’.Footnote ⁸² For example, we can assume that a tribunal dealing with Article 8 of the ARSIWA that applies the exact same threshold for control as the ICJ did in Nicaragua did so because it saw no specific reason to recalibrate the general international law standard. However, if a tribunal issues a decision that seems to depart from previous understandings, it should consider explaining why it decided to do so.Footnote ⁸³

Although tribunals enjoy considerable freedom when interpreting open-ended rules such as that expressed in Article 8, references to different approaches in a single award suggest that, in investment arbitration, the term ‘effective control’ refers more to a standard of evaluation than a specific threshold of control.Footnote ⁸⁴ Take the tribunal in Karkey Karadeniz, for instance, which justified its decision on attribution on the basis of elements that refer to corporate control, de facto control, and specific instructions – namely (1) the state-owned had acknowledged that it was owned and controlled by the government of Pakistan; (2) the government of Pakistan had approved the contract related to the dispute; (3) the government of Pakistan was the guarantor of that contract; and (4) the buyer was not allowed to interfere in licenses granted in the contract, which was interpreted as a sign of bad faith.Footnote ⁸⁵ The diverse interpretations, each unique to a particular award, demonstrate the elasticity the term ‘effective control’ has within this system.

In the absence of a specific threshold of control to be applied uniformly across cases, a correct application of Article 8 of the ARSIWA relies on judicial reasoning. Rather than simply looking at the case law for possible tests and picking one, investment arbitration tribunals must explain their reasons when ruling on what would be the correct threshold of control to be applied in the case before them. As tribunals are not limited in the reasoning that they need to present to justify their decision, they have the power and freedom to decide what elements they will consider relevant in each case and how they weigh each of those elements. However, those reasons must be clear and rational from the stance of international investment law.

3.1 The CJEU and Interpretation of EU Law: The Basics, in Brief

Unlike traditional international courts, whose authority to interpret international law is on an equal footing with that of national courts and which in principle do not have compulsory jurisdiction nor operate as part of a co-ordinated scheme, the CJEU is entrusted with the authentic interpretation of EU law and there is clear division of labour between it and the member states’ national courts.Footnote ⁴² Also, the CJEU exercises exclusive and compulsory jurisdiction over matters falling under the EU treaties, having in this respect a role resembling to that of states’ constitutional or supreme courts.

As a result, the CJEU has rapidly affirmed its interpretative practices and methods, which are somewhat different from those in international law. It applies methods that are similar to the interpretive techniques of a domestic constitutional court,Footnote ⁴³ albeit with variations on a number of points. As early as the Van Gend en Loos case, the CJEU identified three key interpretative methods it employs – namely, giving consideration to ‘the spirit, the general scheme and the wording of [the] provisions [of EU law]’Footnote ⁴⁴ – which can be readily equated with, respectively, teleological, systemic, and grammatical interpretation of the law (generally applied in that order of priority in CJEU practice).

3.2 The CJEU Interpreting CIL: Exploring and Understanding the Patterns

The intellectual process consisting in interpreting the law must be distinguished from the identification of a CIL rule (i.e. determining the existence of such a norm). In accordance with Merkouris’s fundamental distinction, interpretation is a deductive process after a given CIL rule has come into existence, whereas identifying a CIL rule is more of an inductive exercise.Footnote ⁴⁵ Once a CIL norm has come into being (i.e. has implicitly been endorsed by the international community through state practice and opinio juris), its subsequent application in particular cases in the course of its lifespan will be subject to the deductive process of interpretation. What is mistakenly called ‘interpretation’ of state practice (or of both constitutive elements of a CIL norm) – an exercise that would be better termed the ‘qualification’, ‘assessment’, or ‘evaluation’ of the formative elements of CIL – should not be conflated with, and thus needs to be distinguished from, interpretation strictly speaking subsequent to the rule’s emergence.Footnote ⁴⁶ ‘In this manner, interpretation focuses on how the rule is to be understood and applied after the rule has come into existence.’Footnote ⁴⁷ However, some grey zones and overlaps persist. Let us take the example of exceptions under a given CIL rule. It might be debated whether delimiting the scope of particular exceptions is a purely deductive exercise, as such qualifying as interpretation; or whether recourse to the two-element test (assessment of state practice and opinio juris) is again necessary to determine whether such an exception exists as a self-standing rule, an exercise qualifying as identification of a CIL norm that is incidentally an exception to the main customary rule.Footnote ⁴⁸ Another grey area is the dynamic relationship between a rule and its exceptions, which ‘may well change and evolve over time or even on a case-by-case basis’.Footnote ⁴⁹

When the CJEU interprets CIL norms forming an integral part of the EU legal order, its interpretation – given that its rulings are binding erga omnes – also binds the member states, and consequently influences their practice and opinio juris. In other words, CJEU judgments engaging with CIL rules have a ‘multiplying’Footnote ⁵⁰ or ‘snowball’Footnote ⁵¹ effect. Below we discuss four examples of CIL norms dealt with by the CJEU that well illustrate its approach to the interpretation of customary international law.

2.1 The Indispensable Parties Principle and CIL

As custom forms a broad corpus of law, the present analysis will be confined to the principle of indispensable parties, as articulated by the International Court of Justice (ICJ) in the Monetary Gold Removed from Rome case.Footnote ¹⁸ Aware of the difficulties surrounding the history and legal origins of this principleFootnote ¹⁹ and the objections that exist as to its nature as a rule or principle of customary international lawFootnote ²⁰ the present author nevertheless argues that it is precisely this confusion that makes it worth examining, as the role played by the international judiciary in its evolution offers ample scope for discussion. This, in turn, provides the opportunity to scrutinise the work and analysis carried out and to open a discussion on the role of activism and restraint in relation to custom in general.Footnote ²¹

Recognised in international law as a governing principle and a piece of conventional wisdom,Footnote ²² this principle essentially affirms that a court or tribunal will not decide a case if the decision will affect the legal rights of a third state which is indispensable to the resolution of the dispute before it. This principle is most often framed as an absolute,Footnote ²³ yet without any particular source attributed to it.Footnote ²⁴ Considering that different principles may be derived from ‘different sources of international law at different times and even the same principle may be simultaneously shaped by various elements of the legal order’,Footnote ²⁵ we can see how the principle can be the result of a variety of sources, one of which could arguably be CIL.Footnote ²⁶

As this analysis sets out to demonstrate, the indispensable parties principle has by now not only acquired general acceptance within the international legal system but is also called upon quite frequently in practice.Footnote ²⁷ The principle affords an opportunity to analyse both its identification and its subsequent application as a customary rule: while Monetary Gold may or may not reflect a rule of custom that existed at the time of the ruling, subsequent international courts have allowed us to understand it as such today. Furthermore, as the principle has a procedural character,Footnote ²⁸ this enables a direct link between judicial effectiveness and judicial activism, because the very need for effective procedures is directly linked to the effective administration of justice.Footnote ²⁹

4.1 A Definition

Before commencing the analysis, it is important that the terms ‘judicial activism’ and ‘judicial restraint’ be defined. As former CJEU judge Pescatore noted ‘[w]hat is described by one as activism is seen by another as a just and necessary safeguard’.Footnote ⁵² Therefore, the determination of a judicially activist or restraint decision will always be a matter of degree.Footnote ⁵³ Despite its relatively widespread usage, the term ‘judicial activism’ has no unified meaning. Although judicial activism is commonly attributed to the law-creating nature of the role of judges, and restraint to rejection of that law-creating role, these attributions provide little as to the nature or content of these principles. Indeed, it has even been asserted that ‘judicial activism is simply the label used for decisions one does not like’.Footnote ⁵⁴

International legal scholars have nevertheless attempted various definitions of judicial activism, which offer some indication as to how it might be discerned. Hugh Thirlway identified two types of activism: activism in the ‘formal sense’, which arises when the judge ‘deals with an issue the decision on which is not essential to the resolution of the dispute before him … [but wishes] to “enrich and develop” the law’;Footnote ⁵⁵ and ‘substantive activism’, which arises when the judge is ‘dissatisfied with existing law … [and] … will be ready to indulge in something close to open law-creation in order to base his decision’.Footnote ⁵⁶ Former ICJ judge Kooijmans objected to both activism and restraint before the ICJ, steering away from the ‘destructive trap’ of activism and the ineffectiveness of restraint, ‘which closes windows’,Footnote ⁵⁷ and introducing instead the concept of ‘proactive judicial policy’.Footnote ⁵⁸ To him, the judicial role is ‘to utilise those aspects of the case which have a wider interest or connotation in order to enrich and develop the law’.Footnote ⁵⁹ Lara Pair considered the ‘ultimate criterion for judicial activism’Footnote ⁶⁰ to be the focus on ‘avoidance of an unjust result’ and ‘creative reasoning rather than strict application of the law’.Footnote ⁶¹

Judicial restraint has been even more illusory, yet some helpful definitions can also be found. Lauterpacht described a court that exercises restraint as one which ‘refrains from explaining rules in detail’Footnote ⁶² as it ‘may create … the impression the court was laying down new rules of law’.Footnote ⁶³ Judge Kooijmans similarly qualified restraint as the court that ‘wilfully abstain[s] from using the opportunity … to provide clarification on a matter which […] is highly controversial’.Footnote ⁶⁴

Former ECtHR judge Mahoney defined the two concepts in parallel. He did not approach activism and restraint as two ‘conflicting theories as to how judges should go about decision-making’.Footnote ⁶⁵ Instead, he referred to ‘making interpretation choices’, with judicial activism being ‘neutralised’ by the exercise of judicial self-restraint.Footnote ⁶⁶ Therefore, activism involves ‘[modifying] the law from what it previously was or was stated to be in the existing legal sources’,Footnote ⁶⁷ while restraint is a technique necessary in order to move ‘proceedings forward by incremental steps in specific contexts rather than by dramatic leaps in the dark’.Footnote ⁶⁸ These two forces ultimately work together within judicial decision-making in a ‘coherent path of continuity’.Footnote ⁶⁹ Activism and restraint are ‘two sides of the same coin’,Footnote ⁷⁰ both aimed at giving the law a coherent structure.

Judge Mahoney’s description seems to be the one best suited to the nature of the international judicial function and the legal system within which it acts. Viewing the two notions in terms of their complementarity is an effective way of understanding activism and restraint, transcending traditional black-and-white definitions to better reflect the true nature of the judicial process. After all, ‘rules are not picked from trees’:Footnote ⁷¹ even if the legal principles are clear and the law is clearly enunciated, ‘someone’ will have had to develop them – a competent law-determining agency, a court.Footnote ⁷²

4.2 Identifying Activism and Restraint within the Overall Context of International Law

With the above in mind, identifying activism and restraint remains a difficult exercise, as observed through the lack of any such theoretical framework within the literature on international law.Footnote ⁷³ Further, identifying activism and restraint must be placed within the overall context of the international legal system, whose characteristics are inseparably linked to how and why certain decisions may or may not be considered instances of judicial activism.

One must first recall that the international legal system is considered to be a system of states:Footnote ⁷⁴ the settlement of disputes remains largely dependent on the consent of the states involved,Footnote ⁷⁵ and the exercise of jurisdiction by international courts and tribunals is not integrated into an institutionalised and mandatory legal system as at the domestic level.Footnote ⁷⁶ It must also be recalled that states have the upper hand by choosing whether to create a court in the first place, by deciding on the nature of the agreement that will institute a court or tribunalFootnote ⁷⁷ and the substantive law which will applyFootnote ⁷⁸ and, in most instances and particularly important here, by choosing to be a party to a case.

As a result, the first indisputable factor pointing to judicial activism would be a ruling that explicitly goes against the will of the parties and the principle of state consent. Ruling against the accepted law in force would be an example of this, but so too would be deciding against what states have previously explicitly accepted as the law, be it in treaties and conventions, statements, public opinions or judicial proceedings.Footnote ⁷⁹ Next, and flowing from this first factor, judicial activism involves modifying the law from what it was previously accepted to be. The focus here would be on the legal rules or principles themselves and their amendment, as opposed to the will or the consent of the state. It could involve departing from previous interpretations of a principle that were already accepted as law in such a way as to create incoherence or uncertainty. These are all indications of a judge who is ‘making sense’Footnote ⁸⁰ of new law, possibly due to dissatisfaction with existing law, the law which was, by and large, already recognised and accepted by that system of states. A third indication of activism, which is somewhat related to the former two, would consist in changing a judicial process or procedure in flagrant opposition to accepted customary practice.

Another characteristic of the international legal system, which leads to our next criterion, is the material incompleteness of international law.Footnote ⁸¹ This is the presumption that international law, by its very nature, has not evolved or become detailed enough to be seen as a complete system of laws, rules or norms,Footnote ⁸² thus leaving huge gaps in legal regulation. Despite this, legal certainty through judicial decisions is required. The act of deriving legal reasoning from abstract ideas or principles without any necessary legal certainty as to their true definition could reflect judicial activism but also restraint, as it might lead the judge to refrain from deciding.Footnote ⁸³ However far-fetched they may be, judicial decisions are expected to ensure certainty, and coherent reasoning is needed to support the ultimate decision rendered. Equally, a clearly result-oriented decision, either lacking in judicial interpretation or pulling a rabbit out of its hat, can be a characteristic of both activism and restraint, since the result sought could indicate the mental state of the court in taking the decision and its intention to either expand or retract the law. Finally, refusing to interpret or expand upon a legal point raised by the parties or to even carry out any interpretation would be an indication of judicial restraint, as Lauterpacht and Kooijmans noted.

6.1 The Principle Applied

As we have seen, one measure of judicial effectiveness is to consider whether the decisions of international courts are harmonised at a systemic level, meaning that principles of law accepted and endorsed by some courts become accepted as applying to other judicial institutions across the international stage. Telling, in this regard, are the decisions concerning the indispensable parties principle before other tribunals and courts.

Notable is the case of Larsen v. Hawaiian Kingdom,Footnote ¹²⁸ which is the only case in which the principle was applied in such a manner as to prevent an international tribunal from exercising its jurisdiction.Footnote ¹²⁹ The arbitral tribunal, constituted under the auspices of the Permanent Court of Arbitration (PCA), ultimately dismissed the case for lack of authority to ‘exercise jurisdiction over a State unless that State has given its consent to that jurisdiction’.Footnote ¹³⁰ It upheld this test because the tribunal ‘operates within the general confines of public international law’,Footnote ¹³¹ thus extending the application of the principle beyond the ICJ’s mandate and beyond what the ICJ had pronounced on the principle. The tribunal can be said to have added to the definition of the Monetary Gold case, going beyond what the ICJ had previously considered to be a matter of the ICJ Statute, in a hint of judicial activism. Another possible consideration to take into account here is the fact that the parties’ real motive may have not been to resolve the dispute between them but rather to raise publicity for the cause of the Hawaiian nationalist movement by having the tribunal render an opinion on the legal status of Hawaii.Footnote ¹³² The tribunal’s choice to uphold the principle may have been heavily influenced by the need to reject the case in any way possible.

The indispensable parties principle was also raised in the South China Sea Arbitration (Philippines/China),Footnote ¹³³ where an intervening third state, Malaysia, asked the tribunal to reject jurisdiction on the basis that it might affect its rights and obligations as an absent third state. Without providing any explanation beyond the decision of the ICJ in Nicaragua, the tribunal took the view that Malaysia ‘overstated’ the Monetary Gold principle, as the ‘more expansive reading would impermissibly constrain the practical ability of courts and tribunals to carry out their function’.Footnote ¹³⁴ Remarkably, in contrast to the narrow interpretation the ICJ applied in East Timor, this tribunal saw an expansive reading as limiting the judicial function. Clarifying that the relevant issues in the case concerned ‘areas to which Malaysia makes no claim’,Footnote ¹³⁵ it held the latter’s interests were not affected. Interestingly, the difference in reasoning between the ICJ’s shaping of the rule as we know it today and the tribunal’s approach to is suggestive of activism, leading to dis-harmonisation and, by consequence, ambiguity.

Mention must also be made of the M/V Norstar case before the International Tribunal for the Law of the Sea (ITLOS),Footnote ¹³⁶ where the tribunal ‘acknowledge[d] that the notion of indispensable party is a well-established procedural rule in international judicial proceedings’.Footnote ¹³⁷ It nevertheless refrained from applying the principle in that specific case,Footnote ¹³⁸ leaving an unexplained reference to the principle, insufficient for deeper examination.

6.2 Lack of Relevance of the Principle

Although the indispensable parties principle has, by and large, been accepted as well established, the role of the ICJ in creating a new rule of law remains questionable, as its role as a rather reluctant lawmaker arises. This is particularly pertinent when one considers that certain other international courts have rejected the relevance of the principle. In this vein, we can see the complete rejection of the principle in the opinion of Advocate General Wathelet in the Western Sahara Campaign v. United KingdomFootnote ¹³⁹case before the Court of Justice of the European Union (CJEU), who rejected its application as it would allegedly automatically preclude the possible examination or review of international agreements between the EU and third states.Footnote ¹⁴⁰ The Advocate General’s lack of analysis suggests that he was exercising judicial restraint. Prior to that, a World Trade Organization (WTO) panel also held that there is no concept of ‘essential parties’ in WTO law,Footnote ¹⁴¹ despite examining the case law of the ICJ and using it as evidence to elaborate on the content of the principle.Footnote ¹⁴² Most recently, in 2021, the International Criminal Court refused to apply the principle when requested, on the grounds that its jurisdiction did not concern inter-state disputes but dealt instead with natural persons.Footnote ¹⁴³ Without a doubt, these decisions indicate the failure of the Monetary Gold principle to achieve support within the international judicial community both in terms of fostering harmonisation and as a new rule of international law.

The European Court of Human Rights (the ECtHR) has also taken a differentiated approach to this principle. In the case of BankovicFootnote ¹⁴⁴ the ECtHR ruled out any need to decide on an objection raised that was based explicitly on the Monetary Gold principleFootnote ¹⁴⁵ as it declined jurisdiction over other issues in the case that apparently took precedence over any rights deriving from the absent third state principle. Similarly and maybe particularly tellingly is the fact that the same court has had no difficulty providing detailed consideration of the conduct of states not parties to its proceedings or sometimes not even members of the Council of Europe.Footnote ¹⁴⁶ This, for instance, occurred when the facts of the case pointed to a possible violation of the Convention, as most famously occurred in Soering,Footnote ¹⁴⁷ where the court decided against the extradition of the applicant to a third state because it was quite possible that that absent third state would violate the Convention. Interestingly, the question of the essential third party was not even raised, despite the court having explicitly taken decisions in that regard. It is thus evident that this was a result-oriented decision made with the intention of solidifying the law and as such is indicative of judicial activism.

7.1 Consolidation of the Indispensable Parties Principle

As seen in the Introduction, the interaction between the use of certain principles of customary international law and judicial activism can yield an uncomfortable relationship. This is one where the inherent tension between custom’s recognition of unwritten law through the identification of state practice and opinio juris and judicial activism’s ‘creative reasoning rather than strict application of the law’ seek to co-exist.Footnote ¹⁴⁸ An analysis of whether the indispensable parties principle as identified in Monetary Gold qualifies as custom lies outside the confines of the present discussion. However, it is interesting to note the overlap between custom and activism in the judicial process that enables the principle to be characterised as one of customary law. Although the principle had appeared in international decisions prior to Monetary Gold, the ICJ seems to have pulled a rabbit out of its hat when applying it in that case. The existence of prior references to the rule, of which the court made no mention, makes its approach a form of continuity in the judicial interpretation of this principle rather than a new apparition altogether. However, its vague terminology left the source of the rule openFootnote ¹⁴⁹ and gave way to its subsequent evolution – a clear lack of coherence and consistency in the decision.

Additionally, it cannot be denied that, despite not having been applied by tribunals, the rule has been relied on repeatedly by states, which indicates its acceptance as law and in practice.Footnote ¹⁵⁰ Indeed, it has also been argued that the lack of objections to a string of judicial decisions may signify support for a positive customary rule.Footnote ¹⁵¹ What this analysis has shown is a judicial process of thought which, through activism and restraint, has been able to create what can best be described as a principle of customary procedural law.

7.2 The Judicial Function and the Indispensable Parties Principle

Returning to the effectiveness or ambiguity of international courts in the application of the indispensable parties principle, and since its re-introduction in international legal discourse through the Monetary Gold case, there are several observations that can be made.

There is a very important distinction that need to be drawn between, on one hand, the cases where the principle was applied, and, on the other, the cases where the principle was recognised but not applied. In Monetary Gold we saw that the ICJ sought to uphold an indisputable public function, which was to protect the interests of third states not parties to the proceedings. It did so, even if this meant rejecting the entire case and despite the parties having consented to its jurisdiction. We must also consider the utility of leaving dispute unaddressed. A case in point is East Timor, where questions of erga omnes and self-determination arose. As former ICJ judge Kooijmans established, restraint is reflected in instances where the court refuses to decide on a matter which is highly controversial, as happened in this case. The ICJ did not even attempt to provide an answer to these long-standing questions, which would have benefitted the international legal scene – something it was perfectly qualified to do, given its wide jurisdictional powers. Even though reference was made to Article 59 of the ICJ Statute, according to which the decision would be binding only upon the states parties to the proceedings, the ICJ was nonetheless unwilling to decide upon the legal interests of any third state. The refusal to decide was clearly a policy decision, as it was also for the PCA tribunal in Larsen, where a decision would have promoted the wrong political gains.

By contrast, and quite interestingly, we can also see a public function in deciding on the dispute, as was apparent in Nicaragua and Certain Phosphates of Nauru, so much so that in this latter case the court considered that even though other states could also be held liable, it was sufficient that liability arose for that one state party alone. This was an interesting reasoning, outside the bounds of what had been previously accepted. Notable also are the approaches taken in other courts. Advocate General Wathelet’s comment in Western Sahara was indicative of the very bounds of the principle, yet also demonstrates the restraint implicit in a refusal to decide. What if a dispute arising out of a convention with a third party only affects the legal interests of a member state of the EU? Similarly, the rejection of the principle in the South China Sea arbitration and by the ECtHR in Bankovic show a public function quite different from the one assumed by the ICJ in those two cases in accepting the indispensable party principle. In truth, we can see how the reasoning of the tribunals alluded to greater public benefit in deciding rather than not deciding the cases.

As already observed, judicial effectiveness also calls for systemic harmonisation at the international level – something the indispensable party principle also fails to meet. If there is any consistency in the decisions examined, it ironically lies in their disharmony, for the manner in which the principle is applied, whether within a single court or across several courts. This is evidenced, for example, in the lack of coherent or continuous reasoning across the decisions of the ICJ. At the same time the principle was openly accepted by most international courts and tribunals as an established rule of international law, despite not being applied. If we come back to the utility of the principle and its effectiveness at the international level, the lack of its use could be indicatory of its lack of any effectiveness. If the idea is to create a form of ‘transnational public order’Footnote ¹⁵² whereby all the courts and tribunals share a set of principles, even at the procedural level, creating predictability and coherence at the international level, then the indispensable party rule certainly does not meet this standard.

Turning finally to the judge’s role as reluctant lawmaker, the question arises as to whether, in adjudicating Monetary Gold, or even East Timor, the ICJ has made law. The wording of the Monetary Gold judgment alludes to a principle already accepted in international law – something that could be reflected in previous decisions of international courts and tribunals, as seen above, despite the fact that the court did not mention any. Conceivably, the principle may have existed as one of custom, without further substance as to its content. Going beyond the acceptance of the principle, whatever its source might be, subsequent ICJ cases attempted to create a methodology through which this principle would be applicable. However, this methodology, or ‘law’, has been rejected repeatedly by other international courts and on most occasions by the ICJ itself. This is where the question of methodology in custom becomes even more important: does the occasional recognition, but not application, of a rule, make law?

2.1 Blurred Lines between the Identification and the Interpretation of Customary International Human Rights Law

2.2 Judicial Dialogue as a Tool in the Identification and Interpretation of Jus Cogens Norms Protecting Human Rights

The ILC’s special rapporteur Dire Tladi recalled that if ‘a State has objected to a rule of customary international law while that rule was in the process of formation, the rule is not opposable to the State concerned for so long as it maintains its objection’.Footnote ²³ However, the ILC emphasised that ‘the persistent objector rule does not apply to peremptory norms of general international law (jus cogens)’.Footnote ²⁴ This conclusion ‘flows from both the universal application and hierarchical superiority of [these norms] … that apply to all States’.Footnote ²⁵ In its 2019 report on the peremptory norms of general international law, the ILC affirmed that ‘[c]ustomary international law is the most common basis for peremptory norms of general international law (jus cogens)’.Footnote ²⁶ adding that ‘[t]reaty provisions and general principles of law may also serve as bases for peremptory norms of general international law (jus cogens)’.Footnote ²⁷ It also observed that ‘[d]ecisions of international courts and tribunals, in particular of the International Court of Justice, are a subsidiary means for determining the peremptory character of norms of general international law’,Footnote ²⁸ as is the case with the determination of CIL. In the ILC’s 2022 report Dire Tladi provided the most recent list of norms considered to be part of jus cogens and which reflect fundamental values for the international community as a whole. They include the prohibition of aggression; the prohibition of genocide; the prohibition of crimes against humanity; the basic rules of international humanitarian law; the prohibition of racial discrimination and apartheid; the prohibition of slavery; the prohibition of torture; and the right of self-determination.Footnote ²⁹ It is interesting to note that this list partially overlaps with customary rules identified as such in legal scholarship, examples of which are ‘the prohibition of genocide and torture, the prohibition of slavery and piracy, the rules on State responsibility, the principle of non-refoulement and the no-harm rule’.Footnote ³⁰

In Conclusion 14, the special rapporteur pointed out that ‘[a] rule of customary international law not of a peremptory character ceases to exist if and to the extent that it conflicts with a new peremptory norm of general international law (jus cogens)’.Footnote ³¹ This statement could be relevant in future cases dealing with the relationship between legal norms that are not at an equivalent hierarchical level. It could determine the effect of a jus cogens norm when there is a conflict with a customary rule, such as the normative interaction between the prohibition of torture and the rule on state immunity before foreign jurisdictions. An extensive interpretation of Conclusion 14 would lead to neutralisation of the CIL immunity rule’s precedence over the peremptory norm when the application of the first rule renders ineffective in its scope the customary rule that has jus cogens status. However, for that to be the case, one must first consider that there is an existing conflict between, on the one hand, international customary law on immunities and, on the other, erga omnes obligations and jus cogens norms. The main uncertainty lies in the relationship between customary rules related to state immunities and the hierarchical character of jus cogens norms that should entail effective legal consequences. The ILC’s special rapporteur stated:

The ECtHR affirmed that ‘the right under Article 3 not to be subjected to torture or to inhuman or degrading treatment or punishment enshrines one of the fundamental values of democratic society’ and that ‘[i]t is an absolute right, permitting of no exception in any circumstances’.Footnote ³⁵ If the special rapporteur was indeed recalling the well-established case law of the European Court, it could have been criticised for its contradictions and limited legal consequences in terms of state accountability.

This precedent established by the ECtHR has led to a legal loophole in the case of grave human rights violations that occur extraterritorially and implicate a foreign state. Even in such exceptional circumstances, the customary law on immunities exercises a constraining effect on the legal consequences in terms of state responsibility, as individuals will have no access to justice. Indeed, as affirmed by the ECtHR itself:

The dissent of Judges Rozakis and Caflisch from the Al-Adsani v. the United Kingdom solution, which was adopted by only nine votes in favour and eight against, showed the division among the judges at the time. The following words in the dissenting opinion show that the adoption of a radically different interpretation could have been possible:

On the one hand, the ICJ followed the ECtHR’s precedent and ruled in its Jurisdictional Immunities of the State caseFootnote ³⁸ that there was no conflict between the rule on state immunities in civil proceedings and jus cogens norms, as the former is of a procedural nature while the latter are of a substantive nature. This case dealt with both a debate around the identification of an exception to state immunities where there are violations of jus cogens norms, as affirmed by the Italian Supreme Court in its Ferrini case,Footnote ³⁹ and the resolution of a possible conflict of norms between the customary rule on immunities and the absolute prohibition of torture. An analysis of domestic and international case law led the ICJ to reject any exception to the regime of state immunity from foreign jurisdiction, even in the case of violations of peremptory norms. As a result of this restrictive interpretation, the prohibition on torture becomes devoid of substance in an extraterritorial context: its peremptory nature is illusory and impracticable, its legal effects and effective sanctions existing only in a theoretical imaginary, given that state immunity prevails as a sine qua non prior to any judicial proceedings.

On the other hand, with regard to peremptory norms, the case law of other international courts and tribunals has also been used to identify the emergence of customary international legal rules determined by other courts that are binding on all states and not limited in their application. In the Furundžija case, the Trial Chamber of the International Criminal Tribunal for the Former Yugoslavia (ICTY) found that the prohibition of torture had not only become part of customary international law but had also reached the status of a jus cogens norm.Footnote ⁴⁰ The ICTY has also stated that the prohibition of rape within the context of an armed conflict had become a rule of customary international law despite the lack of a universal definition of rape in international law,Footnote ⁴¹ the tribunal using the principle of the protection of human dignity to define the act of rape after analysing contradictory national legislations.Footnote ⁴² The tribunal also defined the criteria of torture in armed conflict after recalling that ‘[t]he broad convergence of … international instruments and international jurisprudence demonstrates that there is now general acceptance of the main elements contained in the definition set out in article 1 of the Torture Convention’,Footnote ⁴³ applicable in times of both peace and armed conflict.

It is also interesting to note that when the case law of international tribunals refers to the prohibition of torture as a customary norm that is part of jus cogens, the formula used is usually ‘prohibition of torture’ and not the universal treaty’s broader wording, which reads ‘the prohibition of torture and other cruel, inhuman or degrading treatment or punishment’.Footnote ⁴⁴ Therefore, it is not clearly stated whether the proscription of inhuman or degrading treatment characterised by a minor gravity as compared with torture also fall under customary international law and could be qualified as a peremptory norm. For instance, in Questions Relating to the Obligation to Prosecute or Extradite case the ICJ stated that:

The court recalled that although Article 4 of the Convention against Torture requires states parties to criminalise such acts, ‘the obligation to prosecute the alleged perpetrators of acts of torture under the Convention applies only to facts having occurred after its entry into force for the State concerned’.Footnote ⁴⁶ Nevertheless, it is not clear whether only the prohibition of torture is part of customary international law as a jus cogens norm or whether this prohibition necessarily also entails procedural obligations strongly linked to the effectiveness of the prohibition as a customary norm. Furthermore, as regards the Convention against Torture, the definition of acts of torture is strongly linked to states obligations as it requires involvement of a public official, whereas this is not the case with torture defined as a crime against humanity involving the liability of individuals.Footnote ⁴⁷

Finally, in the Prosecutor v. Furundžija case, the ICTY analysed treaty provisions of international humanitarian law and human rights law,Footnote ⁴⁸ recognised the numerous states parties to those treaties and, acknowledging that ‘torture in time of armed conflict is prohibited by a general rule of international law’, observed that ‘[i]n armed conflicts this rule may be applied both as part of international customary law and – if the requisite conditions are met – qua treaty law, the content of the prohibition being the same’.Footnote ⁴⁹ With respect to the content of the prohibition of torture under conventional and customary international law, the tribunal remarked that states had the obligation not only to prohibit and punish torture but also to prevent any potential breaches of the prohibition of torture as well as any inhuman or degrading treatment, as had been ‘authoritatively held by the European Court of Human Rights’Footnote ⁵⁰ in the Soering case.Footnote ⁵¹ The ICTY concluded that ‘international rules prohibit not only torture but also (i) the failure to adopt the national measures necessary for implementing the prohibition and (ii) the maintenance in force or passage of laws which are contrary to the prohibition’.Footnote ⁵² This conclusion is reinforced by the jus cogens nature of the prohibition of torture and the fact that it imposes erga omnes obligations designed to produce a ‘deterrent effect’.Footnote ⁵³ For the tribunal, the peremptory character of the prohibition of torture has effects at the inter-state level, such as de-legitimising any legislative, administrative or judicial act authorising torture,Footnote ⁵⁴ and at the individual level, enabling every state ‘to investigate, prosecute and punish or extradite individuals accused of torture, who are present in a territory under its jurisdiction’.Footnote ⁵⁵

Analyses of the international case law related to the prohibition of torture have shown that its customary nature and content flow from the coherence of general principles of international law and treaty provisions applicable in international humanitarian and human rights law, analysed in the light of the international courts’ case law. It also appears that the customary prohibition of torture has distinct content and scope, binding all states as a peremptory norm of international law and giving rise to erga omnes obligations.

2.3 Cross-References and the Crystallisation of a New Customary Rule: The Example of the Binding Nature of Provisional Measures

Consideration of other international instruments and the associated case law resulting from their interpretation by other courts can also lead to the identification of new customary rules, as exemplified by the binding nature of provisional measures. In the case of Mamatkoulov and Askarov v. Turkey before the ECtHR, the applicants alleged that their extradition to Uzbekistan by the Turkish authorities exposed them to a real risk of ill-treatment proscribed by Article 3 of the European Convention on Human Rights (ECHR). The court had indicated an interim measure in application of Rule 39 of its Rules of Court. The provisional measure requested a suspension of the extradition proceedings for the purpose of establishing whether the risk of ill-treatment existed. Turkey’s failure to comply with the measure indicated by the Court raised the issue of the mandatory nature of interim measures and whether there was a breach of the effective exercise of the right of individual application guaranteed by Article 34 of the ECHR.

The LaGrand caseFootnote ⁵⁶ before the ICJ dealt with a dispute related to alleged violations of the Vienna Convention on Consular Relations of 24 April 1963 on the grounds that the applicants had been tried and sentenced to death without having been informed of their rights, as required under Article 36(1)(b) of the Vienna Convention. The ICJ indicated provisional measures, calling upon the United States to suspend the execution of the death penalty, but the respondent state failed to comply with the court’s order made under Article 41 of the ICJ Statute. The court had therefore to determine whether the provisional measures had a binding effect. It reached the following conclusion:

The ECtHR has referred extensively to the LaGrand case decided by the ICJ,Footnote ⁵⁸ in which the mandatory nature of provisional measures was established after decades of doctrinal debates.Footnote ⁵⁹ It should be noted that in its Cruz Varas caseFootnote ⁶⁰ the ECtHR took a different position as it considered that its procedural rules were not equal to a conventional instrument approved and adopted by states, making it impossible to affirm the binding nature of interim measures in comparison with other systems. Nevertheless, in the Mamatkoulov case the ECtHR emphasised that ‘in the light of the general principles of international law, the law of treaties and international case-law, the interpretation of the scope of interim measures cannot be dissociated from the proceedings to which they relate or the decision on the merits they seek to protect’.Footnote ⁶¹ The ECtHR stressed the specificity of interim measures within the inter-American system, where the power of the Inter-American Court of Human Rights (IACtHR) to order such measures has an explicit conventional basis.Footnote ⁶² The ECtHR also recalled, however, that the ICJ in its LaGrand case, the IACtHR, the Human Rights Committee and the Committee against Torture

Although there was no change to the ECtHR’s procedural rules over the years, nor any amendment to the ECHR, cross-references to the case law of other international bodies allowed the ECtHR to establish the mandatory nature of its provisional measures and to conclude that ‘a failure by a respondent State to comply with interim measures will undermine the effectiveness of the right of individual application guaranteed by Article 34 and the State’s formal undertaking in Article 1 to protect the rights and freedoms set forth in the Convention’.Footnote ⁶⁴

In their joint partly dissenting opinion, Judges Caflisch, Türmen and Kolver criticised the court’s reasoning. They noted that the ICJ, for instance, was interpreting Article 41 of the ICJ Statute, a provision of a constitutive treaty, whereas the ECtHR was interpreting its procedural rules, to which states have not given their consent.Footnote ⁶⁵ They stressed that by relying on international instruments and precedents from other international bodies, the ECtHR departed from its own case law and exercised ‘a legislative function, for the Convention as it stands nowhere prescribes that the States Parties to it must recognise the binding force of interim measures indicated’.Footnote ⁶⁶ As regards crystallisation of the binding nature of provisional measures on the basis of normative convergence in the international legal order, the dissenting judges finally concluded:

It is interesting to note that the ECtHR departed from its own precedent, set previously in the Cruz Varas case. It founded its legal reasoning, notably, not on state practice or opinio juris but on the general principles of international law, the law of treaties and international case law,Footnote ⁶⁸ and on the fact ‘the right of individual application is no longer dependent on a declaration by the Contracting States’.Footnote ⁶⁹ Even though not affirmed explicitly, the court acknowledged this as a departure from its previous case law, the practice of international courts and tribunals having revealed the binding nature of interim measures to be now part of general international law.

In this example from the ECtHR, cross-references to the case law of other courts were used as a tool to counteract the negative consequences arising from so-called absolute state sovereignty and to bypass normative constraints. The co-ordination and harmonisation of international case law minimises divergent interpretations of a common legal rule such as the binding nature of provisional measures. Interim measures ordered by judicial bodies are aimed at preventing irreversible damage to the rights of parties during executive, legislative and political decision-making processes. As a matter of fact, the binding nature of provisional measures is linked to their function, as the purpose of judicial proceedings is to ensure effective protection of human rights by state authorities required to act in accordance with the rule of law.

Therefore, the growth of multidimensional dialogue between international bodies protecting human rights, between national judges and between international bodies and national judges contributes to the emergence of a network within the international judiciary.Footnote ⁷⁰ Although operating against the backdrop of distinct and independent legal systems in the international legal order, the international judiciary shares common judicial practices that help to define norms of reference and minimum standards of protection. Nevertheless, these norms and standards need to be incorporated into national legal orders for full applicability and effectiveness, since the primary responsibility to ensure the respect of human rights is devolved to national authorities. Even in the absence of an explicit written rule, international bodies, by applying and interpreting human rights obligations and by sanctioning violations implicating the international responsibility of states, aim to place limits on the exercise of power by framing executive, legislative and political process within the rule of law.