Workers of the World, Unite… and Strike?

On 21 May 2026, the International Court of Justice (ICJ) issued its long-awaited Advisory Opinion (AO) on the right to strike, by ten votes to four, affirming that the right to strike is protected under the International Labour Organization’s (ILO) Freedom of Association and Protection of the Right to Organize Convention, 1948 (Convention No. 87). The AO, accompanied by four separate opinions, four dissents, and two declarations, puts––or at least purports to put––a formal end to a decades-long conflict that had paralyzed the ILO’s supervisory machinery. While the AO itself offers a great deal for legal analysis, I will focus on two, perhaps the most pertinent, aspects of the AO’s interpretive architecture––the ICJ’s characterization of the pronouncements of supervisory bodies, as well as the Court’s treatment of the travaux préparatoires as a supplementary means of interpretation under Article 32 of the Vienna Convention on the Law of Treaties (VCLT).

Both issues generated considerable disagreement among the judges and raised questions that extend well beyond Convention No. 87, touching on the very foundations of how international courts engage with treaty silence and the interpretive authority of institutional bodies. As the AO was awaited by a huge array of different actors, including trade unions, labour lawyers, and industry representatives, due to its influence on industrial relationships across the globe, I will also offer a few thoughts on the social implications of the Court’s analysis of the issue. Namely, that it can neither be reduced to a straightforward triumph for labor rights nor to a missed opportunity to grant a more meaningful and material protection to workers across the globe. It can and should be seen as both.

The Pronouncements of ILO Supervisory Bodies: An Answer in a Non-Answer?

The AO’s journey started on 10 November 2023, when the ILO Governing Body adopted a resolution requesting the ICJ to rule whether Convention No. 87––despite its express silence on the matter––encompasses a right to strike. This question was the root of the irreconcilable tension within the ILO for decades. The ILO’s supervisory bodies, notably the Committee of Experts on the Application of Conventions and Recommendations (CEACR) and the Committee on Freedom of Association (CFA), for many decades, had treated the right to strike as an implicit corollary of freedom of association. But the Employers’ Group contested both the existence of that right under the Convention and the very authority of the supervisory bodies to adjudicate it. The latter issue even featured in the first draft of the ILO resolution requesting the AO, but was ultimately removed (see para. 59 of the AO).

The Court found that the ILO supervisory bodies, the CEACR, CFA, and the Commissions of Inquiry, have progressively converged in recognizing the right to strike as protected under Convention No. 87 (para. 83). It then held that, while such pronouncements do not in themselves constitute subsequent practice under Article 31(3)(b) of the VCLT, they may nonetheless be accorded “great weight” as a supplementary means of interpretation (paras. 117-119). How the majority dealt with the pronouncements, be it their status, importance, or the place in the internal ILO dispute, caused a lot of discussions in the dissents, separate opinions, and declarations.

Judge Bhandari, for instance, accepted the general proposition that supervisory body pronouncements deserve weight, but took issue with the specific calibration. The terminology of “great weight,” he argued, is drawn from a context (i.e., human rights treaty bodies with quasi-judicial authority) that does not map onto the ILO’s tripartite supervisory framework. The CEACR and the CFA operate within an institutional structure characterized by government, employer, and worker representation; their pronouncements do not carry the same degree of judicial authority. They should, thus, be granted “due weight,” which would be more appropriate to the institutional character, expertise, and degree of consensus underlying each pronouncement––a case-by-case assessment, so to say.

Judge Tomka was far more skeptical. In his view, the supervisory bodies, particularly the CEACR, did not engage in a rigorous exercise of treaty interpretation at all. Their positions were “repeatedly and persistently contested” by a significant number of states, and thus, a practice so challenged cannot be elevated into “authoritative guidance”. This was quite a harsh critique, especially given the context that the question about the CEACR’s interpretative authority was removed from the final request, and in Tomka’s view, the Court still implicitly answered that question anyway by ascribing “great weight” and thus, validating the authority that was never put to the CEACR.

Judge Tladi also offered a nuanced contribution. He observed that the ICJ has, on numerous previous occasions, considered and relied upon supervisory body pronouncements without feeling the need to categorize them within a specific provision of the VCLT. By explicitly placing them under Article 32, the AO takes a novel classificatory step. Tladi’s concern was that this categorization may be premature, especially given that the International Law Commission confirmed that pronouncements cannot, as such, constitute subsequent practice under Article 31(3) (para. 22 of his opinion).

What remains to be seen is how the Court’s findings will be internally digested at the ILO. My intuition is that the AO will not be welcomed as a clarificatory gift to the Workers’ and Employers’ Groups in their disputes concerning the authority of the supervisory bodies.

The “Inconclusive” Result of the Travaux Préparatoires

If a legal instrument is silent on a certain issue, we, as lawyers and judges, must resort to the rules of treaty interpretation. Accordingly, the Court had to engage in a rigorous assessment of Articles 31 (general rules of interpretation) and 32 (supplementary rules of interpretation) of the VCLT. To assess whether the right to strike was encompassed in the Convention No. 87, the ICJ studied the ordinary meaning of the terms “activities” and “programmes” in Article 3(1), read in conjunction with Articles 2 and 10. It found that they are sufficiently broad to encompass strike action as one of the means by which workers’ organizations further and defend their members’ interests (paras. 69-74). Having reached that conclusion under Article 31, the Court turned to the supplementary means of Article 32, including the travaux.

The Court’s survey of the preparatory work is remarkably candid. It acknowledged that the drafting process was primarily preoccupied with the right to strike of public officials, and that when the general right to strike was mentioned, several states stated (leading the International Labor Office to take a similar view) that it was irrelevant to the proposed Convention (paras. 108-110). The Court’s conclusion was that the travaux led to “an inconclusive result” (para. 111).

Several judges, however, had a different position on this element of “inconclusiveness”. Judge Tomka relied on the fact that on the only time the right to strike briefly surfaced in the negotiations it was deemed tangential. This thus “leaves little doubt” that no such right was ever intended for the Convention, meaning that “inconclusiveness” was not a correct outcome to reach. The AO, he argued, “glosses over these elements, cherry-picking those that suit its conclusion.” Vice-President Sebutinde, by contrast, went in a wholly different direction: she found the Article 31 analysis so clear-cut as to make recourse to Article 32 entirely unnecessary. In her declaration, she reasoned that Articles 3, 10, and 11 of Convention No. 87, read together and in good faith, are sufficiently broad to encompass strike action without any need for supplementary confirmation. The protection of the right to strike, she argued, is a necessary corollary of freedom of association, since absent such protection, that freedom would be deprived of much of its practical effect. Thus, the travaux’s inconclusiveness is simply irrelevant.

Perhaps, the Court’s decision to engage with the travaux anyway had some transparency advantage as it addressed the historical argument head-on instead of dismissing it. The travaux were the Employers’ Group’s best weapon. The entire ILO dispute was driven by their contention that Convention No. 87 was never meant to cover the right to strike. Thus, by surveying the drafting record candidly and only then finding the result “inconclusive,” the Court could reassure the losing side that their evidence was examined but does not bear the weight they placed on it. Nevertheless, whether or not the ICJ intended it as a precedent, it has supplied a template for other judicial bodies that can now reach an Article 31 result first and then neutralize the preparatory work as “inconclusive”. The interesting question will be whether the Court has quietly licensed itself to find travaux inconclusive whenever the text already points where it wishes to go.

Historic Win or Missed Opportunity?

Whether the Court read the Convention correctly is an important question for lawyers who will grapple with it for years to come. But it was not the only question the Opinion raises. Behind the machinery of Articles 31 and 32 stands a right exercised, or suppressed, in the working lives of hundreds of millions of people, and it is on that plane that the AO is hardest to assess. The ICJ’s approach leaves a lot of uncertainties on the social effects of this, as judge Xue pointed out, “exercise of human rights advocacy.” Will it lead to the resolution of the long-standing disagreement among the tripartite constituents? Was it a missed opportunity to cement the right through indicating its scope, conditions, and limits? Or was it a historic win for the trade unions and allied actors?

To analyze these questions, a short recourse to the social context is needed. The right to strike is violated in 87% of countries. The ICJ’s affirmation of this right under Convention No. 87 could thus go far beyond the interpretive technicalities of Articles 31 and 32 VCLT. Judge Cleveland, in her pleasantly surprising declaration (in which she also pointed at the customary nature of the right to strike), said that: “Workers do not form organizations to play cards. They associate to try to enhance their collective power in order to secure better wages, hours and working conditions. Without the right to strike, employees lose the leverage needed to overcome power asymmetry in the workplace and at the bargaining table.” The Court confirmed this point in paragraph 73, holding that strike action is “one of the main activities engaged in and tools used by workers and their organizations to promote their interests and improve conditions of labour.”

Despite this, the Court’s conclusion “does not entail any determination on the precise content, scope or conditions for the exercise of that right” (para. 140), which somewhat confirms the existence of the right, but leaves unclear what follows from it. The practical consequences of this ambiguity are quite significant. What operational effect does the confirmation of the right have on national legislative provisions, such as mandatory notice requirements, essential services restrictions, or cooling-off periods? Without a definition of what strikes are, and without determining the conditions under which the right may legitimately be curtailed, there is no meaningful protection. The ILO supervisory bodies might continue to develop these conditions through their pronouncements, now validated as carrying “great weight,” but still without binding force.

The structural problem, which the Court––procedurally correct––did not to address, is that the question it was asked was necessarily incomplete. The Employers’ Group’s core objection was never purely about whether the right exists, but about who has authority to define and enforce it. By confirming the right without delimiting its content, the ICJ has given workers a symbolic victory, but it has also granted employers a continuing platform for contestation––which might, legitimately, be a source of bitter feelings. The Court has on other occasions (e.g. in its Chagos AO or Occupied Palestinian Territory AO), addressed questions it was not directly asked and engaged with the broader implications of its holdings––despite these being politically charged cases. But this time, it chose restraint.

Nevertheless, this AO is, unambiguously, a historic milestone. For the first time, the principal judicial organ of the UN has authoritatively confirmed that the right to strike is protected under international law’s major freedom of association instrument, and it deserves celebration. Moreover, for us, international lawyers, the ICJ’s engagement with the full panoply of Article 31 and Article 32 interpretive tools will render this Opinion a standard reference in treaty interpretation doctrine for years.