A Framework for the Weight of Treaty Body Output: The ICJ’s Right to Strike Advisory Opinion

On 21 May 2026, the International Court of Justice delivered its advisory opinion on the right to strike under ILO Convention No. 87, holding by ten votes to four that the right is protected by that instrument. The Opinion clarifies a long-running issue between workers’ and employers’ groups. But it is also significant as a methodological case. The Court majority concludes that the ordinary meaning of the terms of the treaty, read in good faith, in the context, and in light of the object and purpose of the treaty indicates that ILO Convention No. 87’s freedom of association provision encompasses the right to strike. Participants throughout the proceedings cited various other material to support their competing claims, including a variety of ILO treaty body output, which is in focus here. In responding to these claims, the Court majority constructs a framework for situating pronouncements of the ILO treaty supervisory bodies in the landscape of Article 31(3) and Article 32 of the Vienna Convention on the Law of Treaties. The Court’s approach will likely enlighten the debate on the doctrinal weight of such material also in respect of the output by compliance bodies in other treaty systems.

The majority’s approach to this issue is threefold. First, pronouncements of treaty supervisory bodies are not, in themselves, subsequent practice under VCLT Article 31(3)(b). Article 31(3)(b) requires the agreement of the parties, and a body of independent experts (or a tripartite supervisory architecture) cannot, by its own pronouncement, supply that agreement (para. 83).

Second, supervisory body pronouncements may operate under Article 31(3)(b) derivatively, where they give rise to or refer to State practice that itself establishes agreement (para. 83). The position seems to align with Conclusion 13(3) of the ILC’s 2018 Draft Conclusions on Subsequent Agreements and Subsequent Practice (as per Nolte, Special Rapporteur). Pronouncement and practice are two different things. What engages Article 31(3)(b) is the practice generated by States based on pronouncements, not the pronouncements themselves.

Third, the Court admits supervisory body output as supplementary means under Article 32, treating them as falling within an open-ended reading of that provision. In doing so it extends the “great weight” formula from Ahmadou Sadio Diallo – originally developed for the UN Human Rights Committee – to the ILO supervisory design (paras. 116–119). The Court preserves its Diallo stipulation – that it is not legally required to follow such output. The present comment suggests that the third move is the most interesting of the three.

Diallo concerned the Human Rights Committee, an independent expert body whose composition (all members are in practice lawyers) suggested ascribing weight to its output. The ILO supervisory system – the output of which is dealt with by the Court in this Opinion –is structurally different. The Committee of Experts on the Application of Conventions and Recommendations (CEACR) is to some degree comparable to the HRC. The Committee on Freedom of Association (CFA) is tripartite – composed of representatives of governments, employers, and workers, whose interests are opposed in many of the cases before it. The International Labour Conference is a political organ. Commissions of Inquiry are constituted ad hoc.

The Court seems to justify the extension by reference to convergence (the bodies have, over time, reached substantially the same position, see para. 119) but also function (each performs a monitoring role analogous to that of the HRC under the ICCPR Optional Protocol). The mutatis mutandis qualification (para. 117) is however notable and should inform the transposition of the Court’s statement to other supervisory bodies, whether in the realm of human rights or in other sectors. The Court’s approach is conservative.

The framework applies, with little modification, to other treaty supervisory bodies, in particular in the human rights sector of international law. The UN human rights treaty bodies – the HRC, CESCR, CEDAW, CERD, CRC, CAT, CRPD and the others – produce General Comments, Views in individual communications, Concluding Observations and Statements that the framework now accommodates. None is, in itself, Article 31(3)(b) practice. Each may operate derivatively under that provision where it generates of reflects State practice meeting the threshold of agreement. Each must be regarded as “supplementary means” under Article 32 and be ascribed “great weight” only insofar as there is merit to doing so. Where a treaty body has itself adopted an interpretive methodology that departs from the VCLT approach, the case for ascribing “great weight” to its output is correspondingly weaker  The majority does not address this directly (see, however, para. 118 of the opinion), but its framework does not avoid it either (for one illustration, see Emberland, 23(2) HRLR (2023), on the CRC’s Syrian Camps decisions).

Outside the UN human rights system, the framework may apply, mutatis mutandis, to compliance committees in other treaty regimes, such as the Aarhus Convention Compliance Committee, the Advisory Committee under the Framework Convention for the Protection of National Minorities, and The European Committee of Social Rights. Whether the framework applies in unmodified form to such bodies is, on the face of the Opinion, an open question – they differ in composition, mandate, and the form of their outputs – but the central methodological question is the same, and the analogical extension is available.

The European Court of Human Rights has since Demir and Baykara v. Turkey drawn extensively on UN and ILO supervisory body material in interpreting the ECHR. The Strasbourg Court should be expected to invoke the advisory opinion in defence of its existing practice, but with the methodological refinement of it.

The Opinion does not conclude the conversation. Some of the dissenters individually and forcefully disagree with the majority’s inclusion of treaty body output as VCLT Article 32 “supplementary means”, see in particular Dissenting opinion of Judge Tomka paras. 82–84, Dissenting opinion of Judge Abraham paras. 56–66), Dissenting opinion of Judge Hmoud paras. 46 et seq., and Dissenting opinion of Judge Xue para. 64. Even among the majority there are varying views on the methodological approach taken, see, especially Judge Nolte’s Separate opinion, and Judge Tladi’s Separate opinion. This enables further debate. The next phase should, however, focus less on whether supervisory body output is admissible (it is in principle, but only under VCLT Article 32) and more on the criteria that determine how much weight, if any, particular categories of such output should carry. The ILC’s ongoing project on subsidiary means and the CAHDI work on non-legally binding instruments may supply resources for that phase.