Introduction
The new Arbitration Rules of the International Centre for Settlement of Investment Disputes (ICSID) that came into force in July 2022, is a comprehensive revision of the 2006 Arbitration Rules brought with a view to bringing efficiency and cost-effectiveness in the investor-state resolution process. Major changes like mandating disclosure of third-party funding, providing different timelines to the tribunal to give its decision at various stages and expressly providing the power to grant security for costs align with broader concern for increasing transparency and efficiency in the system of investor-state dispute resolution. But particularly it is the purported omission of the provision allowing the investment tribunals to draw adverse inferences in the 2022 rules which raises questions.
Adverse inference refers to an indirect conclusion drawn by a tribunal as a sanction against a party that refuses to provide direct evidence. Essentially, it assumes that the withheld evidence would have been unfavourable to the party's case. For example, in Feldman v. Mexico, the claimant accused Mexico of discriminatory tax treatment against foreign taxpayers. The tribunal ordered Mexico to produce evidence against discrimination between foreign and domestic taxpayers. However, Mexico failed to provide this evidence citing confidentiality concerns. As a result, the tribunal inferred that the withheld evidence likely demonstrated unfair treatment of domestic taxpayers, which would have supported the claimant’s allegations.
Adverse Inferences are essential in international arbitration to ensure fairness and effectiveness by compensating for the tribunal's lack of coercive power, to compel evidence from sovereign states. For instance, unlike courts, tribunals cannot enforce contempt charges or impose penalties for failing to comply with orders for document production. Adverse inferences help address this limitation by drawing conclusions unfavourable to the non-complying party.
2022 Arbitration Rules & the Ensuing Conundrum
In the erstwhile 2006 rules, Rule 34(3) allowed the parties to request the tribunal to ‘take a formal note of the parties refusal’ to produce evidence as ordered by the tribunal. However, these words, i.e. ‘take formal note of the parties refusal’ contained in Rule 34(3) of the 2006 Rules have been omitted from the Rule 2022 Rules without any explicit explanation as to its deletion in the working papers. The only reference to adverse inferences is found in the comments given by China and Armenia. While it was proposed by China that the power to take adverse inferences should be formally omitted, Armenia argued in favour of the formal inclusion of this power. However, no discussion was recorded on these proposals. It becomes even more intriguing to see a similar omission from Note 13 on Document Production of UNCITRAL Notes on Arbitral Proceedings wherein an explicit reference to the power to draw adverse inference has been deleted in 2016 notes that was earlier contained in the erstwhile 1996 notes.
Although, there has long been a debate surrounding the potential risk of ‘false positives’ associated with reliance on such a mechanism, the recent revision prompts the question of whether the Tribunals still have the power to draw adverse inferences under Arbitration Rules 2022. To begin with, the general principle regarding use of inferences in international dispute resolution has been explained by ICJ in the Corfu Channel Case. In the said case, the ICJ was to decide upon the liability of Albania in laying the underwater mines. However, there was no direct evidence establishing that the mines were laid by Albania. Despite this, while holding Albania liable for the loss caused due to the mines, based on inferences, the court reasoned that when a state exercises exclusive control, the victim of an international law breach often lacks access to direct evidence to establish state responsibility. Therefore, the reliance of the victim on inferences and circumstantial evidence should be construed liberally, and the tribunal should be allowed to draw inferences when the party to the dispute fails to provide the required direct evidence (Page 18). This practice is now accepted across legal systems and is acknowledged by international courts and tribunals. For example, in the case of William J Levitt v. Islamic Republic of Iran, the Iran-US Claims Tribunal noted that it is free to draw inferences from the parties' non-compliance with its order to produce documents (¶ 61).
However, since the ICSID Arbitration Rules 2022 omits the reference to the power of an arbitral tribunal to take formal note of the failure to produce documents, which was earlier expressly contained in Rule 34(3) of the ICSID Arbitration Rules 2006, the question as to whether or not the ICSID tribunals continue to possess this power becomes relevant. Although, in most instances, the tribunals have assumed the power to draw adverse inferences without any justification, it has sometimes referred to power in Rule 34(3) as the source of the power to draw adverse inferences. Illustratively, in the case of Rompetrol v. Romania, the tribunal ruled that the power to take formal note under Rule 34(3) is the source of the tribunal’s discretionary power to draw adverse inferences. In RSM v. Saint Lucia, the tribunal ruled that the “Rule 34(3) reflects the common principle that a fact-finder can draw inferences from a failure to produce evidence” (¶ 56). Similarly, in the case of Feldman v. Mexico, the arbitral tribunal ruled that it is empowered to draw appropriate inferences from any party's failure to comply with the document production order (¶ 8).
At this juncture, it is important to look at the source of the words ‘take formal note of the refusal’ as contained in the ICSID Arbitration Rules 2006. A similar use of these words can be seen in the Statute of the International Court of Justice (ICJ), which, under Article 49 deals with the Evidence taking the power of the ICJ. It also uses the words ‘Formal note shall be taken of any refusal’ (to comply with the evidence production order), and commentators have interpreted this as the source of the power of ICJ to attach such consequences as it deems necessary for the non-compliance with the document production order.
In such a scenario, it can be argued that the omission of the power to take formal note of the refusal to comply with the document production order, is tantamount to the omission of the power to attach negative consequences to non-compliance with the document production order, thereby excluding the power of arbitral tribunal to draw adverse inferences from the non-production of documents.
A similar case against adverse inferences is also reflected in the UNCITRAL Notes on Arbitral Proceedings, wherein the explicit reference to the power to draw an adverse inference in its Note 13 on Document Production contained in the erstwhile 1996 notes has been deleted in the 2016 notes.
To the contrary, it can be argued that a tribunal possesses an inherent authority to draw adverse inferences from the non-production of the documents that is embedded in its power to determine the admissibility, relevance, and weight of the evidence presented. This view is also supported by Nathen D. O’ Malley, in his treatise “Rules of Evidence in International Arbitration.” Accordingly, it can be concluded that the power to draw adverse inferences remains intact despite the change in arbitration rules (¶ 7.37). A similar approach was also followed by the tribunal in the case of Sevilla Basheer B.V. v. The Kingdom of Spain, where the tribunal, while relying on the power to admit and weigh evidence, contained in Rule 34(1) of the 2006 Rules, concluded that it did possess the power to draw adverse inferences from the non-compliance with the document production order (¶ 550).
Under the 2022 Arbitration Rules, the tribunal continues to possess the power to admit and assign weight to the evidence under Rule 36. Here, the tribunal is empowered to admit direct and indirect evidence as the provision does not specifically omit the indirect evidence. It is important to note that this distinction, between the direct and indirect evidence, pertains to the weight of the evidence rather than its admissibility, thus, both the direct and the indirect evidence can be admitted. Since, adverse inferences, are, by their very nature, indirect evidence only, they should be admissible under Rule 36(1) of the ICSID Arbitration Rules 2022, and the only question for the tribunal to decide is regarding the weight to be attached to it.
Further, it is well-established that tribunals possess the authority to resolve procedural matters. Article 44 of the ICSID Convention grants tribunals the power to address any procedural issues not expressly covered by the ICSID Convention, Rules, or Regulations. This principle was reaffirmed in the case of Libananco Holdings Co. v. Republic of Turkey, wherein it was reiterated that the tribunal ‘must be regarded as endowed with the inherent powers required to preserve the integrity of its process – even if the remedies open to it are necessarily different from those that might be available to a domestic court of law.’ Additionally, this position is in consonance with the procedural laws of most legal regimes where courts can assess the value of any evidence. Even in international arbitration, arbitration clauses rarely address the issue of weighing evidence directly. It can hence be argued that the inherent and unquestionable authority to draw adverse inferences unless otherwise agreed by the parties, stems from the arbitrator’s wide discretion in admitting and evaluating the relevance of evidence, as well as their power to establish and manage arbitration procedures.
Further it cannot be argued that the power to draw adverse inferences results in a shifting of the burden of proof (onus probandi) since the request to draw adverse inference is often made by the opposing party, thereby violating the general principle in international arbitration of actori incumbit probatio. If an opposing party fails to provide evidence that challenges claimant’s case, then it will be a matter of procedural non-compliance and will not affect the burden of proof. As Jeremy K. Sharpe explains, referencing the arbitral award in Feldman v. Mexico, once the party bearing the ultimate burden of proof establishes a prima facie case, the burden of production (onus proponendi) shifts to the responding party to counter that evidence. In other words, if the party with the burden of proof presents evidence sufficient to create a presumption of truth, the burden shifts to the opposing party, which must then produce adequate evidence to rebut the presumption. This approach, adopted in other cases as well, as Sharpe rightly notes, does not shift the burden of proof itself but rather the burden of production or evidentiary burden.
Moreover, it is essential to recognise that the authority to draw adverse inferences has historically been a discretionary prerogative rather than a default sanction. Even IBA Rules present adverse inference as a possible sanction and subject this power to certain requirements that must be met under Article 3.3 and Article 4.10. Due to the discretionary nature of this power, it is drenched in subjectivity but applying such clear conditions and criteria can introduce greater objectivity. For example the standard of reasonableness, consistency with the facts in the record, logical nexus of inference and the missing evidence, which has been inspired by the scholarship of Jeremy Sharpe and Bin Cheng[3] and presently contained in IBA rules as well laid down by various tribunals like in Frederica Lincoln Riahi v. Government of the Islamic Republic of Iran. Arbitrators should explicitly show in their award that these criteria were followed and provide reasons for assigning or withholding weight to the adverse inference. They must also ensure that the defence rights are upheld throughout the process. Moreover, adverse inference as a form of indirect evidence in itself carries very limited evidentiary value. The Arbitral jurisprudence has developed several cautions before an adverse inference is taken. Therefore, in the opinion of the author, instead of omitting the said power itself, which has continued to be a tool balancing the subjective role of the tribunal in weighing evidence and the need for objective fairness in the process, the focus must be on better institutionalizing these safeguards.
Conclusion
In the light of the foregoing discussion, it can be concluded that the omission of the power to take ‘formal note’ of the parties' refusal to comply with the document production order does not conclusively take away the power of arbitral tribunals to draw adverse inferences in case of non-compliance with the document production order. Since no arbitral award based on the 2022 rules has discussed the power of arbitral tribunals to draw adverse inferences, it remains to be seen how the tribunals interpret this omission. However, in the opinion of the authors, the tribunals must rule in favour of power to draw adverse inferences because it is the most potent arrow in the quiver of the arbitral tribunal to enforce its document production orders, especially when it lacks the other sanctions available with the domestic courts to compel the production of evidence. This power with ICSID tribunals becomes even more important because investor-state disputes always involve a much more powerful sovereign state that possesses various key evidence for the fair adjudication of the disputes, and it is seldom possible to marshall the evidence against the sovereign.
[1] Khyati is a Third-Year BA.LLB. Student at Gujarat National Law University, Gandhinagar and can be reached at khyati22bal037@gnlu.ac.in
[2] Saransh is a Third-Year BA.LLB. Student at Gujarat National Law University, Gandhinagar and can be reached at saransh22bal069@gnlu.ac.in
[3] Bin Chen, General Principles of Law as applied by International Courts and Tribunals p. 333-335 (Cambridge University Press 2006).
Komentar