One Size Does Not Fit All: Institutional Diversity and the International Law Commission’s Limits in Codifying Judicial Precedent

[Lena Raxter is an Adjunct Professor at American University, Washington College of Law]

Introduction

The International Law Commission (ILC), during its seventy-sixth session in May 2025, provisionally adopted the first-reading draft conclusions on subsidiary means for the determination of rules of international law. Among them, Draft Conclusion 6 [7] addresses the absence of legally binding precedent in international law, providing that:

Decisions of international courts or tribunals may be followed on points of law where those decisions address the same or similar issues as those under consideration. Such decisions do not constitute legally binding precedent unless otherwise provided for in a specific instrument or rule of international law.

In a recent piece to EJIL: Talk!, Mehmet Emin Büyük argues that the Commission’s use of permissive language obscures the burden of justification that international adjudicators actually impose upon themselves when departing from established jurisprudence, and proposes replacing “may” with a uniform rebuttable presumption of consistency. While this argument has some merit, Draft Conclusion 6 [7] succeeds precisely because it preserves the distinction between a self-imposed methodology and a legally formalized one.

This response advances four arguments. First, codifying a self-imposed rebuttable presumption in the Commission’s operative text elevates that methodology into a formal rule, creating the functional equivalent of stare decisis—i.e., a system of legally binding precedent. Second, a uniform rebuttable presumption of consistency cannot coherently apply across regimes that handle precedent in structurally incompatible ways. Third, the normative content that “may” is said to leave unaddressed is located in the commentaries, precisely where the Commission’s methodology dictates it should reside. Fourth, the Commission’s progressive development mandate cannot serve as a vehicle for such a change.

The Stare Decisis Paradox

Büyük’s proposal exposes a fundamental paradox: operative text, designed to describe existing practice without infringing on the prohibition of stare decisis, can nevertheless inadvertently create a binding system of legal precedent. Büyük proposes revising Draft Conclusion 6 [7] to the following: “decisions should be followed unless there are compelling reasons not to”, placing the justificatory burden on the departing adjudicator. This formulation closely tracks stare decisis as defined by the Max Planck Encyclopedia of Public International Law: a court must follow the legal principles established in its own decisions and those of courts of equal or higher rank on “litigated and necessarily decided issues”, except where departure is warranted. The structural parallel is unmistakable. Both formulations establish a presumption of adherence, place the burden on the departing adjudicator, and permit deviation only upon sufficient justification. The proposed rebuttable presumption thus differs from stare decisis in name, but not in operative effect.

According to Büyük, the explanation requirement is “a matter of judicial methodology, not of source hierarchy” and, on that basis, a rebuttable presumption “does not transform judicial decisions into authoritative interpretations of primary sources.” Yet this distinction does not survive the act of codification. Judicial methodology and source-of-law hierarchy remain separate only where adjudicators develop interpretative habits without those habits becoming binding. A draft conclusion, however, is not a description of practice; it is a normative statement instructing future adjudicators what to do. As Mr. Forteau warned, the Commission’s final output has significant practical effect, “accord[ing] … legitimacy to certain [legal sources] that could increase their importance”. Stating a methodological presumption in operative text necessarily reclassifies it from a discretionary practice to an external rule that adjudicators are expected to follow. For example, the ILC’s “Draft conclusions on the identification of customary international law”, while formally non-binding, carry sufficient weight to function as an authoritative normative reference point—as Judge Iwasawa’s separate opinion in Question of the Delimitation of the Continental Shelf between Nicaragua and Colombia confirms. Consequently, once the Commission codifies judicial methodology in operative text, the distinction Büyük relies upon collapses because the text itself becomes the hierarchy-conferring instrument, elevating the self-imposed methodology into the functional equivalent of stare decisis.

Indeed, the ILC members were aware of this risk. For example, Mr. Galindo, Mr. Oyarzábal, Mr. Nesi, and Mr. Patel each cautioned against importing conceptions of precedent into international law via unclear drafting language. Mr. Forteau also expressed concern that a provision in the Special Rapporteur’s second report “amounted to creating a rule of precedent with binding effect”. Mr. Sall likewise noted that State-consent constraints on jurisdiction and adjudicatory powers form fundamental “structural obstacles to the establishment of doctrine at the international level”.

To navigate this problem, the Commission intentionally adopted language preserving the line between informal practice and formal obligation. The original negative formulation—courts and tribunals “do not normally follow” prior decisions—was recast as a positive one: courts and tribunals may follow prior decisions, with some exceptions. In doing so, the Commission balanced clarity against the prohibition on stare decisis in international law. The “may” language reflects the reality that following prior legal reasoning is a possibility, not an obligation.

Institutional Diversity and the Threshold Problem

Whether international adjudicators treat prior decisions as more than merely persuasive varies by institutional context. In investor-State dispute settlement, for example, the practice of following prior decisions remains unsettled, with many arbitrators selectively citing prior decisions. Standing courts, by contrast, more readily treat prior decisions as presumptively binding. The International Court of Justice (ICJ), for example, will not “depart from its settled jurisprudence unless it finds very particular reasons to do so” (Croatian Genocide, paras. 52 to 56). The Secretariat’s Memorandum on relevant case law documents similar practices in other standing courts.

A rebuttable presumption formulation would essentially codify the practice of standing courts into a uniform operative rule. The question, however, is whether this standard can be stated uniformly across the full range of institutions that the draft conclusion must address.

It cannot, for two related reasons. The first is institutional structure. Standing courts develop settled jurisprudence through consistent bench composition and institutional memory. A settled line of reasoning may emerge from a handful of decisions or even a single decision of particular weight. By contrast, ad hoc tribunals operate with rotating panels and frequently adopt parallel, even conflicting, reasoning on the same legal questions without justifying the departures. Although numerous investment awards include statements that tribunals should adopt prior legal reasoning absent compelling contrary grounds, considerable debate exists over fragmentation, inconsistency, and incoherence in investment arbitration. The Commission cannot impose a uniform operative rule without obscuring these fundamental structural differences.

The second is the threshold question a rebuttable presumption creates: how many prior decisions must exist before the presumption attaches? This question has no settled answer. Standing courts have no clear standard for when the “settled jurisprudence” threshold is reached. Likewise, in investment arbitration, where scholars broadly recognize the emergence of jurisprudence constante, no consensus exists on which lines of reasoning should be followed or how many decisions are required to establish them. Precedential value also varies by institution. As Mr. Asada explained:

“the decisions of permanent courts might … have more value as precedent than those of ad hoc tribunals”

which may explain why the ICJ generally refrains from citing arbitral awards. Within investment arbitration itself, some arbitrators are considered more authoritative than others. A uniform operative text cannot accommodate this variation without distorting actual practice.

Draft Conclusion 6 [7]’s “may” language, on the other hand, better accommodates these variations in institutional structure and practice. While arguably weaker than the Croatian Genocide standard, it allows each adjudicatory body to acknowledge the weight of prior decisions without prescribing a uniform threshold for when that weight becomes determinative. The Commission can then clarify the nuances of actual practice in the commentaries, rather than the operative text. For example, the Commission could build upon nuances already described in the second and third reports: “the reasoning and conclusions of earlier cases remain important for other cases”; a presumption exists that “the Court will start from its prior correct statement of the law”; some courts are considered so authoritative that their decisions may have a “ broader legal effect … felt by all States”; and following prior reasoning provides:

“legal security, consistency and predictability to States, while offering legitimacy to the Court”.

These statements suggest that adjudicators deviating from established reasoning already informally bear some justificatory burden. Nevertheless, in such a varied system, including a one-size-fits-all  rebuttable presumption as an operative rule would not describe international practice—it would distort it.

Where the Substance Lives: The Role of Commentaries

Draft Conclusion 6 [7] cannot be assessed solely on its operative text; doing so overlooks the integral role commentaries play in the Commission’s work product. The Commission’s methodology of pairing concise operative text with detailed commentaries is deliberate and well-established, applied consistently across topics from State responsibility to identification of customary international law. Operative text must be general enough to accommodate the diversity of international legal practice; the commentaries then provide the granular analysis, supporting jurisprudence, and interpretative guidance that give the text its full meaning. Indeed, the supporting jurisprudence Büyük invokes in favor of a more demanding operative rule is precisely the material the commentaries are designed to address.

States engaging with the Commission’s work in the United Nations General Assembly confirmed this function, with multiple delegations welcoming further elaboration in the commentaries. The Drafting Committee Chair’s statement also references specific examples of practice to be addressed in the commentaries.

Importantly, the publicly available initial draft commentaries provide precisely the granular guidance that the operative text deliberately withholds. They identify the practice whereby “certain courts or tribunals” follow previous decisions “unless there are ‘convincing reasons’ or ‘compelling reasons’ not to do so”—the Croatian Genocide standard—and describes this as “a functional equivalent to a rule of binding precedent” even though it creates no formal legal obligation. This language is significant: it acknowledges the practical weight of prior decisions without elevating that practice to a binding rule. Accordingly, the normative content Büyük seeks is located exactly where the Commission’s methodology dictates it should go: in the commentaries.

Progressive Development and Its Limits

The ILC’s Statute tasks the Commission with promoting both progressive development and codification of international law. Büyük argued that the progressive development mandate justifies the proposed rebuttable presumption. However, this misrepresents the powers granted through the progressive development mandate.

Progressive development—defined in Article 15 as:

“the preparation of draft conventions on subjects which have not yet been regulated by international law or in regard to which the law has not yet been sufficiently developed in the practice of States”

is a tool for crystallizing emerging practice, not overriding established principles. The exclusion of stare decisis is not an underdeveloped area awaiting normative support; it is settled doctrine. As former ICJ President Guillaume confirmed, “the stare decisis rule has been excluded” from international law since 1922. Operative language that functionally imposes stare decisis would not be progressive development; it would be legislation, which is beyond the Commission’s mandate.

The Commission’s record confirms its approach was deliberate. Both the Special Rapporteur’s first report and initial draft commentaries characterize the topic as “reflect[ing] primarily codification and possibly elements of progressive development of international law”. The Drafting Committee chose “may” because following prior legal reasoning is “not an obligation but rather a possibility” and modified the provision’s title from “absence of a rule of precedent in international law” to “absence of legally binding precedent in international law” to anchor the conclusion in existing law. As Mr. Oyarzábal and the initial draft commentaries explain, international adjudicators follow prior decisions not because those decisions are binding, but for reasons of convenience, policy, legitimacy, persuasiveness, consistency, and procedural economy. The “may” language does not reflect an abdication of the progressive development mandate; it accurately reflects a practice too complex to be condensed into a single operative rule.

Conclusion

Draft Conclusion 6 [7] is carefully calibrated to preserve doctrinal accuracy in its operative text, while deferring to the commentaries to address the nuances seen in practice. The “may” language reflects the established rule that judicial decisions do not constitute binding precedent, even though international adjudicators routinely engage with and build upon prior legal reasoning. Indeed, the initial draft commentaries already acknowledge the functional weight those decisions carry, identifying their role as a “functional equivalent to a rule of binding precedent” without elevating it to a formal obligation. The operative text’s apparent sparseness is a principled allocation of analytical work: Draft Conclusion 6 [7] accurately describes the current state of international law, while the commentaries supply the detail clarifying its application. In doing so, the Commission successfully preserves the distinction between a self-imposed methodology and a legally formalized one.