Transparency in international investment arbitration refers to the extent to which the public can access arbitral proceedings and information pertaining to those proceedings. The ICSID Arbitration Rules and the UNCITRAL Transparency Rules offer two different transparency frameworks. From a purely substantive rule-based perspective, ICSID’s rules provide a comparatively lower degree of transparency in relation to the public disclosure of documents and non-party access to proceedings than the UNCITRAL rules. The caveat is that the UNCITRAL rules can only deliver their more robust transparency policy provided parties have opted into their applicability, save for qualifying treaties under the Mauritius Convention. By contrast, ICSID’s rules apply by default in ICSID arbitration, guaranteeing at least a minimum level of transparency, however basic.
Transparency in investment arbitration is desirable for a number of reasons explored in this post. As such, ICSID’s proposed amendments to its rules, the substance of which appears to closely align with UNCITRAL’s more robust transparency mandate, are welcome. Should the proposed amendments be implemented, ICSID proceedings will require a greater degree of transparency with the added benefit of applying by default. Ultimately, investment arbitration is likely to gain from the practical benefits of standardising transparency rules and an increase in public confidence.
This post begins by outlining the substantive differences between the ICSID and UNCITRAL rules as pertains to information disclosures and access to proceedings. It then explores why different approaches to transparency have emerged to begin with. This post next considers the potential alignment of the ICSID and UNCITRAL transparency frameworks in light of ICSID’s ongoing amendment process, culminating in a discussion of the positive effects of the implementation of these reforms on investment arbitration.
Confidentiality: ICSID’s Comparatively Lower Standards of Transparency for Information Disclosures
There are notable differences between the ICSID and UNCITRAL rules on disclosure of information (or ‘confidentiality’). ICSID’s party-led and consent-based approach to public disclosure of documents provides a comparatively lower standard of transparency than UNCITRAL’s “presumed and compulsory” approach. By using the word “shall” in various provisions, the UNCITRAL rules “impose an absolute duty on the tribunal to deliver its transparency policy”. By contrast, the ICSID rules allow the extent of the disclosure of information to be guided by party consent. This effectively means that the ICSID regime leaves room for parties to agree to a largely confidential procedure. Absent party agreement, the tribunal may be engaged to decide on the matter via the issuance of provisional measures; in practice, however, ICSID tribunals have been conservative in deciding on issues of transparency.
This conservative attitude was exhibited in Biwater Gauff Ltd v Tanzania in relation to access to documents during the proceedings. In this case, the claimant requested provisional measures on confidentiality as a result of the unilateral disclosure of the minutes of a tribunal meeting. The tribunal agreed to the requested measures, stating that the disclosure of some documents should not be allowed in principle since it would jeopardise the procedural integrity of the arbitral process. The documents referred to included not only those which contained business secrets, but also “information which aggravates disputes before investment tribunals”, considerably constraining the kinds of documents that may be published. Similarly constraining is ICSID Arbitration Rule 48(4) which requires party consent for the publication of the final award. The presumption of non-disclosure is telling of ICSID’s more restrictive take on transparency as a matter of principle. Admittedly, in practice, ICSID tribunals have tended toward publishing excerpts of the legal reasoning underpinning awards; however, it is worth noting that, while this is a step further than the rule’s previous iteration, the current rule still limits what the Secretariat may unilaterally publish.
By contrast, UNCITRAL’s Article 3 imposes comparatively greater disclosure obligations by requiring written submissions, transcripts of hearings, a list of exhibits of documents, expert reports, witness statements presented in the proceedings, and awards to be made available to the public. Against this backdrop, Biwater and ICSID Arbitration Rule 48(4) demonstrate that, unless parties agree to a higher degree of transparency, one should not expect ICSID arbitrations to be conducted with the same level of transparency as those subject to UNCITRAL rules, at least not via prescribed rules or arbitrator initiative.
Privacy: ICSID’s Comparatively Lower Standards of Transparency Regarding Non-Party Access to Proceedings
There is also significant divergence between the institutional rules on active and passive non-party access to proceedings (or ‘privacy’). ICSID’s discretionary approach to non-party access provides a lesser degree of transparency than UNCITRAL’s duty-imposing approach. For instance, in regard to passive participation, UNCITRAL tribunals have a duty to arrange for open hearings (Article 6(1)). By contrast, ICSID tribunals have a discretion to allow open hearings, provided no party objects, after consulting with the Secretary-General, and logistical arrangements permitting (Arbitration Rule 32(2)). Thus, both the discretionary and conditional nature of third-party access in ICSID differentiates it from UNCITRAL’s equivalent provision.
In regard to active participation, both ICSID’s Arbitration Rule 37(2) and UNCITRAL’s Article 4 nearly identically enable tribunals to discretionarily accept submissions by amici, subject to particular considerations. However, the difference is that UNCITRAL’s additional Article 5 goes further by imposing a duty on the tribunal to accept submissions by non-disputing treaty Parties on matters of treaty interpretation. There is no equivalent provision in the ICSID rules. While this in itself has not necessarily been a barrier to accepting amici submissions, the fact that the ICSID rules do not provide the same level of security and predictability on paper as do the UNCITRAL rules speaks to a lesser commitment to making proceedings accessible to the public.
Causes of Dissimilarity and Future Prospects of Uniformity
The substantive differences between the ICSID and UNCITRAL rules may be owed to the fact that there is no strict consensus as to the appropriate balance between squarely conflicting notions of confidentiality/privacy and transparency in international investment arbitration. On the one hand, compromising confidentiality peels away one main attraction of international arbitration. On the other, maintaining levels of transparency is of significant import in the context of investment arbitration because disputes often involve matters of particular public interest and consequence.
Notwithstanding, three proposed changes to the ICSID rules as exhibited in Working Paper No. 4 may align them with the UNCITRAL rules. First, on the confidentiality front, the proposal (via Rule 64(1)) to introduce a duty on the tribunal of disclosure of documents filed during the proceedings with the consent of the parties aligns it closer with UNCITRAL’s Article 3. That is, whereas previously both party consent and tribunal discretion was needed, the proposed amendment makes information disclosure easier and more seamless in light of the need only for party consent. Second, on the privacy front, the proposal (via Rule 68) to introduce a duty to accept submissions from non-disputing treaty Parties mirrors UNCITRAL’s Article 5. Finally, the proposal (under Rule 65(1)) to amend the existing Arbitration Rule 32(2) on oral procedure to impose a duty – previously a discretion – on the tribunal to enable non-parties to observe hearings, absent party objection, also brings the approach to third-party access closer to UNCITRAL’s Article 6(1).
The implementation of these contemplated amendments would be a welcome development in international investment arbitration. It has the potential to achieve greater normative uniformity across institutions and boost the subjective acceptance of the regime, altogether increasing the legitimacy of investment arbitration. First, standardisation helps reduce uncertainty as to “how calls for transparency will be resolved in any particular case”. This is beneficial for states and investors alike who can rely on a single body of practice. Second, implementing a more robust set of transparency rules could increase public confidence in investment arbitration by virtue of its openness. Investor-state disputes may be of significant interest to potentially affected communities, such as those involving environmental or human rights concerns, state concessions over natural resources or approvals of the privatisation of public services, and disputes resulting in a state’s liability for which payment may be absorbed by public tax money. In this context, greater demands for transparency are justified and to be expected. In addition, transparency in arbitral decisions contributes to the development and drafting of new treaties and increases the predictability of investment law, consequently leading to greater “participation and confidence in the system particularly of the less knowledgeable investors and host States”. Moreover, increasing access to decisions also enhances the quality of decisions as tribunals and parties learn from the experience of their predecessors.
As it stands, the substance of the ICSID and UNCITRAL rules on transparency differs significantly. The existence of the differences between the ICSID and UNCITRAL rules may be explained by the absence of an international consensus on the appropriate balance between the conflicting concepts of confidentiality/privacy and transparency. From a rule-based perspective, ICSID provides a comparatively lower standard of transparency on both confidentiality and privacy fronts. However, due to UNCITRAL’s opt-in condition of applicability, ICSID’s default procedure is arguably better placed to deliver a minimum standard of transparency. In light of this, ICSID’s proposed amendments to its rules, which may soon align with UNCITRAL’s more transparent policy, show promise in harnessing a greater standard of transparency with the added benefit of applying automatically in ICSID proceedings. This is a welcome development as it will lead to greater uniformity and increased public confidence in the regime, ultimately maintaining arbitration’s procedural appeal without undercutting “effective democratic participation, good governance, accountability, predictability and the rule of law”.