Balanced investment treaties are often sold as a solution to the legitimacy crisis of Investor-State Dispute Settlement (ISDS). States now incorporate detailed clauses, right-to-regulate provisions, and carve-out clauses to strike a balance between domestic regulatory space and investor rights. Yet in practice, these balanced treaties often produce the same outcomes. This essay presents an alternative explanation for the “new treaties, old outcomes” phenomenon in ISDS. Structural elements, such as high discretion and low veto possibilities, combined with a 2–1 voting system, create a risk that the tribunals reinterpret these provisions in ways that dilute their protections. I term this phenomenon a conversion risk: the repurposing of existing rules without textual change, exemplified by the tribunal’s award in Eco Oro v. Colombia”.
Existing accounts have presented self-contained legal arguments for why treaty reform might be neutralised in adjudication by most-favoured-nation (MFN) clauses and customary international law. While the text of a treaty is undoubtedly a contributing factor, this essay will argue that it is not the main causal driver. The wording on the page matters, but it does not explain why newer “balanced” treaties still often yield familiar investor-friendly outcomes.
I argue that the phenomenon is better understood as a result of gradual institutional change by agents under the structure of ISDS. By gradual institutional change, I mean small, case-by-case shifts in interpretation and practice that, over time, repurpose the regime without any formal amendment to its rules. By structure, I mean the basic design of the system: how arbitrators are appointed, how they vote (for example, by 2–1 majorities), what standards they apply, and how limited state control is once a claim is filed. By agency, I mean the choices arbitrators make within that design: how they interpret open-textured standards, how they use precedent, and how they respond to incentives such as repeat appointments. Neither structure nor agency alone provides a complete explanation: it is the interaction between a highly permissive structure and arbitrators’ interpretive agency that generates conversion risk.
At the same time, the conversion risk thesis is not a prevalence claim. I do not claim that conversion always occurs, or that it occurs in X% of the cases. My aim is only to provide a more accurate causal account of this outcome when it occurs. The conversion risk thesis, which attributes these outcomes to gradual institutional change, can be falsified if we consistently observe tribunals reading minimum standard of treatment (MST) narrowly and uphold exceptions that block liability.
For this post, I use the approach of Kathleen Thelen and Sven Steinmo, called “historical institutionalism”: Institutional, rather than purely legal-doctrinal, analysis allows us to step back from the text and take a broader view of the structure of ISDS, both procedural and substantive law, along with the norms, incentives, and expectations of its key players. Applying HI insights to ISDS has been a lively space in recent years; however, an application of historical institutionalism to the “new treaties, old outcomes” phenomenon is missing. This essay aims to provide that account in brief.
Kathleen Thelen and James Mahoney, for instance, predict that conversion is likely when two structural elements coincide: high interpretive discretion and low veto possibilities. ISDS corresponds exactly to these features. Arbitrators enjoy substantial interpretive discretion, and states lack veto power once an investor files a claim. Specifically, only a 2-1 majority is needed to support a particular outcome. Additionally, party appointees often align with their appointing party. In edge cases, party-appointed arbitrators frequently diverge; the presiding arbitrator often becomes a pivotal change agent. Therefore, these outcomes are better understood as a result of agency operating within structures. Institutional structures incentivise certain choices while constraining others; however, in the end, agents still have a choice. In other words, while structures may decide the menu, it is still up to agents to determine what to eat.
Structural Element One – Arbitral Discretion in Eco Oro and Red Eagle
A detailed analysis of the Eco Oro and Red Eagle cases is beyond the scope of this essay. But these cases highlight one key aspect: arbitral discretion in action.
Both tribunals considered whether environmental actions taken by Colombia violated the MST/fair and equitable treatment (FET) standard under Article 805 of the Canada-Colombia Free Trade Agreement. The relationship between FET and the MST remains controversial. The tribunals reached different conclusions. The Eco Oro tribunal concluded that the MST has evolved to encompass the protection of legitimate expectations, importing elements often associated with broader treaty-based FET standards into the customary law baseline (paras 743–755). Conversely, the majority in Red Eagle adopted a narrower, more traditional view, finding that the MST under customary international law does not encompass legitimate expectations akin to treaty-based FET, and thus found no breach on this ground (paras 292–301).
Upon finding a breach, the Eco Oro tribunal proceeded to the following line of inquiry: whether the General Exception under Article 2201(3) was applicable and, if so, what effect it had on liability. The tribunal found, controversially, that the exception did not preclude the payment of compensation, stating that “it would require explicit and unequivocal text to restrict a State’s obligation to pay compensation” (paras 826–837). Here, the tribunal relies on the award in Bear Creek v. Peru. In this case, the line of reasoning suggests that the exceptions in Article 2201 of the Canada-Peru FTA were likely inapplicable. In a thinly reasoned obiter dictum, the tribunal states, even if 2201 applied, “the exception in Article 2201 does not offer any waiver from the obligation in Article 812 to compensate for the expropriation.” It does not offer any authority to support this conclusion (para 477). Therefore, a thin alternative reading becomes grounds for conversion in Eco Oro. As a result, precedent remains a vital tool for conversion; in future cases, tribunals may cite both cases to support the same conclusion—a clear case of gradual institutional change that further facilitates conversion.
Structural Element Two – The 2-1 Majority and Sociology of the Change Agent
Reaching different conclusions about the exact legal text under similar circumstances is a common enough occurrence in legal adjudication. However, in ISDS, this is exacerbated by its structure, which allows each party to appoint one arbitrator. As Todd N. Tucker notes in his monograph, Judge Knot (p. 41), “the practice of two-on-one arbitration awards by the tribunal became particularly important in facilitating arbitrators’ incremental expansionist strategies in subsequent investment treaty disputes.”
In Eco Oro and Red Eagle, the parties’ appointed arbitrators adopted interpretations favourable to their appointees on multiple issues. Interpretation, therefore, is dependent of arbitrators’ incentives, given that we can reliably predict that, in edge cases, arbitrators are likely to side with the party that appointed them. In most cases, they were precisely appointed for that purpose. Therefore, conversion often hinges on the presiding arbitrator’s interpretation, rendering it a persistent rather than a one-off risk.
The sociology of investment arbitrators also supports this dynamic. Thelen and Mahoney predict that an institution with high interpretive discretion and low veto possibilities is likely to attract opportunists. These are agents who, rather than openly defying rules, exploit the institutional gaps and ambiguities available to them. The existing literature already suggests that repeat appointments and prestige are drivers of arbitrator behaviour. Therefore, the institutional structure of ISDS is vulnerable to opportunists acting in their own interests with little accountability. The expansionist interpretations of the ISDS tribunals support this conclusion.
Structural Element Three – Low Veto Possibilities
The third structural element, which creates the conditions for conversion risk, is low veto possibility. Once a tribunal passes an award, the states have very little control and effectively no “veto” possibility over the outcomes. In particular, ICSID arbitrations (such as Eco Oro and Red Eagle) are unique in that domestic courts cannot refuse recognition and enforcement on grounds of public policy under Article 54 ICSID Convention, and review is limited to the narrow annulment grounds outlined in Article 52. Therefore, inventive and even questionable interpretations of treaty language cannot be vetoed by the state parties ex post.
Conclusion
Existing work, such as that by Wolfgang Alschner, has analysed the “new treaties, old outcomes” phenomenon driven by the tools available to tribunals, including precedent, MFN clauses, and customary international law. This essay presents the alternative thesis that ISDS’s structure creates a persistent conversion risk. Even precise text leaves discretion at the MST baseline, as once a treaty refers to the customary MST, tribunals still decide what that standard requires in practice, leaving them significant room to expand or narrow state obligations. With limited review and the pivotal role of presiding arbitrators, a conversion risk persists. While text helps, structure decides.
If conversion flows from (i) broad interpretive discretion, (ii) who decides, and (iii) weak ex-post veto, then credible reform must target those elements. This is easier said than done, as arbitrators’ discretion, the parties’ ability to appoint their preferred arbitrators, and the finality of ISDS awards are all foundational features of the regime. Therefore, addressing the conversion risk and, by extension, the “new treaties, old outcomes” phenomenon requires a shift towards an architecture that no longer resembles contemporary ISDS.
Harsh Manohar is a Lecturer at Jindal Global Law School. He combines insights from Institutional Political Economy, TWAIL and other critical approaches in his writings.