Investor-State Dispute Settlement (ISDS), and Arbitration before the International Centre for the Settlement of Investment Disputes (ICSID) in particular, has been under fierce scrutiny from academia, civil society, and political actors alike. To name but one critique, significant damages awarded to investors, or the threat thereof, are considered to influence policymaking to the detriment of the public interest. Placing private interests over public ones, such a regulatory chill is observed to affect climate policy and human rights. Driven by a broader range of discontents, the so-called backlash against investment arbitration arose in the late 2000s.
This blog focuses on one central yet controversial characteristic of ISDS: party appointment. Party appointment most commonly entails that each party may select one arbitrator, and the presiding arbitrator is appointed by consensus. Although concrete figures are elusive, disputing parties are known to spend significant time and resources on their selection. The effort expended is warranted, as the choice of arbitrators is considered pivotal to the success of the arbitration. Beyond extensive liberty to shape the procedure of the proceedings, the arbitrators are not bound by precedent, nor are awards subject to appeal, while grounds for annulment are limited.
A perception of bias amongst such central actors undermines the legitimacy of ISDS as a whole. One frequently articulated criticism is that arbitrators are often private practitioners who wear a ‘double hat’ as both arbitrator in one proceeding and counsel in another. Double hatting is perceived as potentially compromising independence and is thought to predispose arbitrators to the investor’s interests. Considerations of arbitrator’s impartiality are a recurring theme at UNCITRAL Working Group III, the prime venue for discussions on ISDS reform.
One essential tool for refuting such allegations, and thus key to the legitimacy of ISDS, is greater transparency. Increased data availability, such as through UNCTAD’s efforts, has enabled mapping of arbitrator appointments in light of theories of how bias may emerge in ISDS. However, the success of such research depends on the reliability of the data used. Identifying a shortcoming in the reporting of arbitrator appointments, this post considers one important element currently undermining the reliability of appointment data and proposes an alternative that enhances transparency.
The Issue at Stake
The specific subject of this post is the reporting of appointments made using the ICSID list and ballot procedure (the list procedure) as appointments by the parties, thereby inaccurately reflecting ICSID’s role in the establishment of tribunals.
The list procedure is used when the disputing parties either fail to reach consensus on the presiding or sole arbitrator or when they proactively request it. The secretariat provides a shortlist of potential arbitrators, from which the disputing parties may strike undesired candidates and rank the remaining. ICSID consequently appoints the highest-ranking individual.
Used in place of a direct appointment by the appointing authority, the procedure offers advantages such as preserving party autonomy. However, reporting such appointments as party appointments does not accurately reflect the degree of institutional involvement. The solution to this problem could be straightforward: if one introduces a new appointment procedure, one should equally create a new reporting category.
The scale of the issue can be illustrated using the 2023 Annual Report. Annual reports provide an account of ICSID’s activities, including statistics on cases, panel nominations, and arbitrator appointments. At page 30, ICSID reports that it made 23% of the 173 appointments that year. The percentage corresponds to forty appointments, yet the report separately notes that the appointing authority made only thirty appointments. Where are the missing ten appointments? These appointments, implicitly mentioned as ‘ICSID appointments based on the agreement of the parties’, are instead registered as party appointments.
Most of these appointments will result from the list procedure, replacing almost a quarter of the instances where a traditional institutional appointment would otherwise have taken place. The recently released 2025 report is even more opaque, as it omits the final number of appointing authority appointmentsthe above calculation.
How the List Procedure Expands Arbitrator Choice
The current construction can be understood in the context of the ICSID Convention. Most fundamentally, ICSID is limited in its appointments to the Panel of Arbitrators. Article 38 of the Convention states that at the request of either disputing party, and in consultation with the parties, the Chairman shall appoint the arbitrators not yet appointed. Article 40 consequently limits such appointments to the Panel of Arbitrators. This Panel is established based on nominations by Member States and an additional ten nominations by ICSID. Currently, the Panel of Arbitrators contains nearly three hundred names. Yet, the previous Secretary General of ICSID and academic work have indicated that in many instances it is difficult to find suitable candidates within it.
These considerations are what motivated the introduction of the list procedure. ICSID suggests that, with the consent of the disputing parties, an off-roster appointment should also be possible. Article 37 of the ICSID Convention defines the default method of tribunal constitution in the absence of an agreement between the parties. Inversely, it can be interpreted as allowing the disputing parties to specify the method desired. The article is considered equally applicable to amending an ongoing tribunal constitution process, thereby opening the door for introducing the list procedure.
The Blurred Lines Between Institutional and Party Appointments
The qualification of such appointments as party appointments is somewhat awkward. For example, the list procedure is often not requested by the parties but proposed by the secretariat in response to a request for the appointment of an arbitrator. In principle, such a request should trigger Article 38. Additionally, the procedure is noted as the default method when a request for institutional appointment is received, which conveys a more authoritative and institutionalised character than an ad hoc party request or party agreement.
ICSID’s list procedure is furthermore nearly identical to the default procedures for appointing authorities included at Article 8(2) of the PCA Arbitration Rules, and Article 8(2) and 9(3) of the UNCITRAL Arbitration Rules.
Interestingly, footnote 10 of chapter four of the Guide to the PCA arbitration rules notes that the Permanent Court of Arbitration’s list procedure failed only twice during its application. In contrast, at its introduction, ICSID’s approach was successful in only a quarter of the cases. Corresponding to the frequency suggested by the 2023 Annual Report, there is a stark and interesting difference between the institutions; alas, further inquiry is currently impossible, as we do not know in which cases the list procedure was applied.
Initially introduced in 2009, the list procedure was incorporated into the 2022 ICSID Arbitration Rules at Rule 17, formalising its application. The rule states that parties may jointly request the Secretary-General to assist with the appointment of the President of the Tribunal or a Sole Arbitrator. Perhaps most importantly, disputing parties have consented to this practice for over 15 years now, remain free to reject it, and appear to welcome the procedure.
Why ICSID’s Reporting Needs to Catch Up with Its Practices
The point of this post is not that ICSID illegitimately evades the bounds of the Panel of Arbitrators. Its purpose is more modest: ICSID registers such appointments as party appointments, which is inaccurate and easy to fix.
The reasons why this should change are twofold. Firstly, ICSID’s duty to publish information on cases is rooted in its Administrative and Financial Regulations. Regulation 26 requires ICSID to maintain and publish a register containing all significant data related to the dispute. The regulation explicitly mentions that the method of constituting the tribunal is among such significant data.
The only remaining question is whether specifying the list procedure constitutes ‘significant data’ in relation to the existing categories. Given the secretariat’s substantial involvement in such appointments and the procedure’s widespread application, the answer should be unequivocally in the affirmative.
Secondly, given the importance of arbitrators, this issue undermines the transparency and legitimacy of ISDS by removing from scrutiny a quarter of appointments with extensive institutional involvement. These appointments are of particular interest, as they reveal institutional and party preferences beyond the Panel of Arbitrators. Therefore, these appointments are crucial for verifying claims by ICSID and for enhancing understanding of the institution’s role more broadly.
For example, paragraph 92 of this UNCITRAL WG III report states that the lack of diversity of arbitrators undermines the legitimacy of ISDS. ICSID maintains elsewhere that the list procedure has increased such diversity. Yet, the current reporting standards make this claim impossible to verify. Using the available data, one can observe that appointments by the appointing authority include a broader range of nationalities than those appointed by a single disputing party. Furthermore, appointments by agreement of the parties show a similar increase, yet to a lesser extent. Is that ‘diversity’ the result of the list procedure or the disputing parties?
Elsewhere, it is suggested that polarisation in arbitration precludes party agreement on a presiding arbitrator, leading to a greater reliance on the appointing authority. However, the share of reported appointments made by the appointing authority has been declining. One cannot currently verify whether this hypothesis is thus false, or whether the use of the list procedure currently obscures a growing institutional influence over appointments.
The reliability of ICSIDs’ appointment statistics is compromised, as it is impossible to distinguish the arbitrators appointed through the list procedure from those appointed by party agreement.
Transparency in reporting is not only an academic concern, and ICSID appears aware of its importance. The 2022 ICSID Rules amendment took several steps to increase transparency, and the ICSID secretariat notes on page 857 of this document that transparency enhances the public legitimacy of ISDS.
A Simple Fix: Creating a Separate Reporting Category
As the former Secretary-General of ICSID, Meg Kinnear herself, said in a blog post elsewhere, “At the end of the day, policy decisions on the form of ISDS are for States to make.” And indeed, the ICSID Convention leaves states considerable freedom to shape arbitral proceedings.
That said, States, academics, and the public require appropriate access to information to make informed decisions. The list procedure is likely to be a positive addition to the methods of appointing arbitrators, without the need for complex amendments to a convention that is nearly 60 years old. However, its use should not undermine adherence to appropriate reporting standards, as the current lack of transparency has negative ramifications for the practice’s legitimacy.
To that end, I propose introducing a new reporting category: ‘Appointed via the List and Ballot Procedure’. This minor change would align ICSID’s reporting with its evolving practices and provide the transparency needed for accurate evaluations of appointment practices.
The author would like to thank Andreas Sennekamp and Kendra Magraw for their valuable feedback on earlier drafts, as well as the anonymous peer reviewer and the editors of Völkerrechtsblog for their helpful comments and suggestions.
Bram Goede is a PhD Candidate and Teaching Assistant at the Geneva Graduate Institute of International and Development Studies (IHEID), in Geneva, Switzerland. For his PhD dissertation, Bram is writing an empirically grounded analysis of the appointment of arbitrators in ICSID arbitration.