The Unexplained Discount

When an investor’s own misconduct contributes to its loss, how much of the damage should the host state have to pay? International investment tribunals answer this question all the time, yet they rarely explain how they reach their conclusions in apportioning responsibility. Indeed, when arbitral tribunals find the contributory fault of the investor, they tend to apply their margin of discretion for deciding the apportionment of responsibility rather than calculating the causal effect of that misconduct on damages. Even though the amount of compensation increased in new cases, arbitral tribunals did not change judicial practice on the margin of appreciation and did not oblige themselves to reason for the conclusion.

From Chorzów Factory to the BITs Era

Causation sits at the heart of state responsibility, yet international courts and tribunals apply it in very different ways. Under the classic “but-for test”, harm is treated as a chain of separate, dependent events: a mechanical model borrowed loosely from science and reshaped for legal use. Katalin Sulyok shows (29–30) in her work on how judges engage with scientific reasoning that legal causation and scientific causation are not the same thing. International adjudicators often borrow the language of science without adopting its methods, which is part of why causal reasoning in international law can look so inconsistent from one tribunal to the next.

The starting point is nearly always the same case. In the landmark Factory at Chorzów judgment (at 47), the Permanent Court of International Justice held that reparation must, as far as possible, wipe out the consequences of an unlawful act and restore the situation that would probably have existed but for the breach. As Borzu Sabahi notes (at 47), this standard has become the most frequently invoked authority in investor-state arbitration, even though it was written for a dispute between two states. James Crawford argues (at 131) that the ILC’s Articles on State Responsibility (ARSIWA) (art 28) remain “residual” in nature. An adjudicator must first examine the relevant BIT, and if the treaty encompasses matters of responsibility as lex specialis, the ILC rules would not apply, as indicated in Article 55 of the ARSIWA.

Given that bilateral investment treaties contain very few provisions, they rarely regulate reparation in any detail. For this reason, tribunals routinely import the Chorzów standard, and with it the ‘but-for test’, to establish the legal basis of a state’s breach as a concrete duty to pay the individual investor. The borrowing of this standard is plausible, but in practice, it often encourages framing that treats state conduct as the sole engine of harm. Since the Chorzów standard originated in an inter-state realm, arbitral tribunals frequently treat the investor’s conduct as a matter of quantum reduction rather than scrutinizing its causal effect at the liability stage.

Contributory fault entered this picture almost by accident. In his Eighth Report (at 13–14) to the ILC, Special Rapporteur Roberto Ago proposed treating it as a circumstance precluding wrongfulness, in effect a full defense. In the final text of the ARSIWA, Special Rapporteur James Crawford and the ILC relegated contributory fault under Article 39  to a factor that simply reduces the amount of reparation owed. This shift, from a question of liability to a question of remedy, has a significant practical impact. Instead of rigorously examining whether an investor’s misconduct broke the causal chain for a specific breach, investment tribunals often implement what Martin Jarrett calls (at 162) “the default practice of intuitively picking a percentage”.

“Rough Justice” in the Awards

Two well-known cases show this pattern clearly. In Occidental v. Ecuador (para. 825), the arbitral tribunal held that Ecuador’s termination of an oil contract breached the fair and equitable treatment standard. In quantifying damages, the arbitrators found that Occidental’s failure to obtain required ministerial authorization constituted contributory fault, reducing liability by 25 percent. This yielded an initial award of roughly USD 1.769 billion—at the time of its issuance in 2012, the largest sum ever ordered in ICSID history. Yet rather than tying Occidental’s misconduct to any specific loss, the tribunal deemed the percentage “fair and reasonable” without further quantification.

The Yukos shareholders’ case (paras. 1633-1637) against Russia followed the same script. In what remains the largest arbitral award in history, the arbitral tribunal found that Russia breached its obligations under the Energy Charter Treaty but also held that the claimants’ aggressive use of low-tax regions constituted contributory fault under Article 39. The arbitrators applied the very same 25 percent cut, again describing the figure only as ‘fair and reasonable in the circumstances’ (para. 1637). Without tying this reduction to specific damage heads or lost value, the award effectively wiped out roughly USD 16 billion on a purely intuitive basis—demonstrating how ‘rough justice’ operates on a massive scale.

Rudolf Dolzer, Ursula Kriebaum, and Christoph Schreuer (at 428) emphasize in their treatise by referencing the Yukos tribunal’s own dictum (para. 1603) that the principle of contributory fault should not be confused with the investor’s duty to mitigate its losses. Even though the arbitral tribunal was aware of this distinction, it relied on discretionary power and implemented the round-number discount.

A Question of Evidence

The problem is not only about numbers. It is also about process. When international investment tribunals jump straight to a round-number discount, they also skip over questions about who carries the burden of proving contributory fault, and to what standard of proof.

Contributory fault findings are among the least reviewed parts of an award, precisely because they are framed as discretionary matters rather than as legal findings subject to a defined standard. This characterization is exactly the kind of outcome annulment committees are reluctant to disturb, since a “rough and ready” apportionment is, almost by definition, not a manifest excess of power. Part of the problem, as Joachim Knoll and Tania Singla (at 93) argued, is that most investment treaties say almost nothing about causation, leaving tribunals to apply a variety of standards and approaches generally borrowed from domestic tort law that were never designed for a relationship between a private investor and a sovereign state.

When Arbitral Tribunals Explain Themselves, and When They Do Not

The real inconsistency is not about which causal test arbitral tribunals use. It is about how openly the tribunal explains its choice. The Occidental v. Ecuador case is of particular importance in this respect. Its findings constitute a model case for a detailed examination of the principle of contributory fault. After stating that the contribution must be material and significant, the arbitral tribunal said it had a wide margin of discretion in apportioning responsibility. However, this margin of discretion was based solely on a previous arbitration decision (see MTD v. Chile (Decision on Annulment, para. 101)) without presenting the legal reasons behind it. (Occidental v. Ecuador, para. 670) Subsequently, the arbitrators noted that the respondent state’s issuance of the Caducidad Decree was equivalent to expropriation, but concluded the claimant also contributed to the outcome through its own wrongful actions. The arbitral tribunal stated that determining the weight of the actions that led to this decision in the causal relationship was a difficult task for it (paras. 678-685). Given the difficulty of determining the causal effect of numerous actions on expropriation, it decided that apportioning responsibility between the claimant and the respondent at 25% and 75% was fair and reasonable (para. 687).

Consequently, the arbitrators recognized the claimant’s contributory negligence and determined that it had an impact on the occurrence of the damage but, based on the difficulty of establishing a causal link between the actions and the result and relying on the broad scope of their discretionary power, determined the amount of compensation at their discretion without performing any calculation. In this regard, it would be reasonable to examine the MTD case, which the tribunal considered as the basis for its decision.

In MTD v. Chile (Decision on Annulment, paras. 99–101), the ad hoc annulment committee accepted a 50 percent reduction for the investor’s own negligence comparatively contributing to the loss, and the tribunal had a margin of estimation in weighting the apportionment. More interestingly, the annulment committee referenced the Costa Rica Packet case, decided more than a hundred years ago, to endorse the view that international tribunals often do not give “any exact explanation” regarding the calculation process. Such tolerance for imprecision, once endorsed at the annulment stage, has arguably encouraged later tribunals, as we can see in the Occidental case, to determine round apportionment numbers instead of a more careful calculation of loss. In earlier decades and centuries, international arbitration was far less developed than it is today.

Legal processes and treaties were very few, calculation methods were not developed, and arbitrators were implementing equity principles as a useful legal tool for avoiding non liquet under these circumstances. Therefore, relying on old international arbitration practices would be problematic in the modern era. Katharina Diel-Gligor observes (at 103–105) several flaws in the investor-State dispute settlement regime and particularly the ICSID system. Accordingly, she argues that international investment jurisprudence suffers from a high degree of inconsistency and that, without some mechanism for harmonizing tribunal reasoning across cases, this sort of inconsistency will keep repeating itself.

Similarly, Jean-Michel Marcoux and Andrea Bjorklund (at 888) draw attention to the incoherent application of the contributory fault principle by international investment tribunals and identify several of the main reasons behind this lack of coherence. Ambiguous reliance on the principle of contributory fault, the non-existence of defined standards for assessing the actions or omissions of the investor, and the common absence of justification concerning the apportionment of responsibility may be significant reasons for this inconsistency in jurisprudence.

A materiality threshold, anchored more explicitly in Article 39 of the ARSIWA, could bring some predictability to arbitral jurisprudence that currently rewards a margin of discretion over rigorous analysis. It would also give annulment committees a clearer benchmark against which to test a tribunal’s reasoning, rather than deferring to whatever apportionment a tribunal is willing to call fair and reasonable.

Toward More Principled Reasoning

In summary, the matter of contributory fault has increasingly come to the fore in high-value disputes over the last two decades. However, as noted in the MTD case, historical reliance on equity-based methods for assessing compensation in past centuries has been taken as a precedent by tribunals in modern proceedings. Modern advancements in compensation assessment methodologies have been neglected out of convenience.

While international investment tribunals invoked the breadth of their discretionary powers as justification, this trend has also been influenced by the lack of explicit language and the absence of a fault-based structure in Article 39 of the ARSIWA, as well as the lack of detailed provisions in BITs. Addressing this analytical deficit through treaty renegotiation or the formal amendment of the ARSIWA remains practically unfeasible. Instead, a viable solution requires a systemic shift toward rigorous judicial reasoning and procedural accountability within the existing arbitral framework. First, arbitral tribunals must substantiate their reduction in awards with concrete economic evidence and verified interruptions in the causal chain, rather than relying on arbitrary, unreasoned percentage discounts under the guise of equitable discretion. Second, Annulment Committees should enforce detailed reasoning standards, treating a tribunal’s unreasoned or speculative apportionment of fault as a failure to state reasons under Article 52(1) (e) of the ICSID Convention. Third, institutional bodies such as UNCITRAL or ICSID should issue technical guidelines on causation and contributory fault as soft law documents to standardize valuation methodologies and constrain subjective arbitral discretion.

Acknowledgments: This post draws on a paper first presented at the “State Responsibility in Crisis” workshop at Bristol University Law School. I am grateful to Dr. Kathryn Allinson and the workshop’s other organisers and participants for their valuable comments, and especially to Dr. Vladislava Stoyanova, who served as discussant on my panel, for her insightful feedback. I would also like to thank the editors of the Völkerrechtsblog for their valuable comments and suggestions.