The New York International Arbitration Center’s (“NYIAC”) annual Grand Central Forum took place on 13 July 2022. The event coincided with the 55th session of the United Nations Commission on International Trade Law (“UNCITRAL”) held in New York which, among other topics, focused on the recent fourth draft of the joint ICSID-UNCITRAL Code of Conduct for Adjudicators in International Investment Disputes (“Code”) issued for informal consultation in June 2022 and currently under discussion in UNCITRAL Working Group III (“WGIII”).
Under the guidance of Donald Donovan (Arbitration Chambers, New York), the panel composed of Andrea Bjorklund (McGill University, Montreal), Lauren Mandell (WilmerHale, Washington D.C.), Kate Brown de Vejar (DLA Piper, Mexico City), and Andrés Jana (Jana y Gil, Santiago) discussed how the Code attempts to regulate the issue of so-called “double-hatting,” i.e., where an arbitrator concurrently has a role as counsel or expert in a related proceeding. In particular, the panelists addressed the relationship between double-hatting and the concepts of independence and impartiality, the prohibited multiple roles and temporal scope of the ban, the proposed three-year tail and disclosure as a solution to the double-hatting challenge.
The Regulation of Double-Hatting in the Current Draft of the Code of Conduct
The conference opened with words by Louis B. Kimmelman (Independent Arbitrator, New York), Anna Joubin-Bret (UNCITRAL, Vienna), and Martina Polasek (ICSID, Washington D.C.) on the ongoing efforts of UNCITRAL Working Group III in drafting the Code and the significant innovations it offers to address concerns surrounding arbitrators in international investment disputes (“IID”).
The limits on an arbitrator acting in multiple roles are set by Article 4 of the draft Code (“Limit on Multiple Roles”). Article 4(1) requires an arbitrator to refrain from acting concurrently (and potentially for a period of three years “following the conclusion” of the proceeding) as a legal representative or an expert witness in another investment dispute involving the same measures, the same or related parties, or the same provisions of the same treaty; while Article 4(2) contains a prohibition on serving as arbitrator where she/he is acting as a legal representative or an expert in another case involving “legal issues which are substantially so similar” that accepting such a role would be in breach of the independence and impartiality obligations regulated in Article 3. Under the current draft, the disputing parties may exclude the application of Article 4(1), but not Article 4(2).
Independence and Impartiality
In his introductory remarks, Donald Donovan posited that the discussion of double-hatting is most properly framed in terms of independence and impartiality; that is, whether and to what extent taking on multiple roles compromises an adjudicator’s ability to decide a case in an independent and impartial manner. Viewed that way, the Code’s goal should be to achieve a practical consensus on how those fundamental principles apply in the particular context of double-hatting.
Andrea Bjorklund addressed the interplay between these principles and double-hatting in the draft Code, an issue that was further commented on by Lauren Mandell and Kate Brown. Article 3 tries to bypass the discussion of whether the independence and impartiality standards should be objective or subjective and instead seeks to find non-exhaustive proxies of situations the community might perceive as problematic, such as nationality bias. The current version of Article 4 attempts to do the same in the specific case of double-hatting, proposing a list of circumstances in which double-hatting by an arbitrator can be a potential source of concern.
The draft raises three important questions. First, whose perception is the Code trying to address? If it is the parties’, then the possibility of waiver under Article 4(1) is sensible; however, if the Code seeks to address the legitimacy of the IID system, then the rule may be inadequate. Second, if concerns with double-hatting are in essence related to predispositions an adjudicator might have, then why can an arbitrator take on multiple cases involving the same issue as long as they only act as an arbitrator and not take other roles (i.e., as counsel or an expert)? Third, is the list in Article 4 exhaustive or exclusive? Further discussion might be required to ensure the provision properly addresses the concerns raised by double-hatting.
Multiple Roles and Time Limitations
The panel then discussed the “tailor-made” ban the Code attempts to create, in a discussion headed by Lauren Mandell, with interventions by Kate Brown and Andrés Jana. Some delegations of WGIII have raised concerns that an outright ban of double-hatting would limit party autonomy, decrease the pool of available arbitrators, and adversely affect diversity. The Code attempts to minimize such effects by tailoring the ban to address only the conflicts of interest that are of greatest concern. But the current draft creates at least two issues in this regard.
First, it is unclear how the obligations in Article 4 can be adhered to in practice. It might not be possible for an adjudicator – or, for that matter, for counsel – to fully comprehend all the contours and minutiae of the dispute at the time of appointment. As a result, it will be difficult to assess, at the outset, whether a case will involve issues under Article 4(2) that are substantially similar with another case in which the adjudicator is involved.
The second concern deals with the duration of the ban under Article 4. If an adjudicator cannot act concurrently in multiple roles (and potentially for three years “following the conclusion” of the dispute), then when is the proceeding “concluded” and when does the three-year-tail commence? Many alternatives are arguably possible: when the time to petition for annulment elapses, when enforcement proceedings end, when an arbitrator resigns or is successfully challenged, among others.
If the Code is to provide a practical test, further work is required to bring certainty in relation to these issues, either by modifying the language of Article 4, by providing further commentary, or by leaving the issue to be developed by tribunals in the light of the obligations of independence and impartiality under Article 3.
The Three-Year Tail Under Article 4 of the Draft Code
Kate Brown then headed a discussion on the three-year tail, with interventions by Andrés Jana and Andrea Bjorklund. The likely goal of establishing a three-year tail introduced into Article 4 of the Code is to address the concern of whether counsel can advocate for a particular issue and then, within a short period of time, advance an independent position on that same issue as an arbitrator. But enforcement of Article 4 would be problematic. First, for the reasons already mentioned surrounding when the proceeding “concludes.” Second, although double-hatting by an arbitrator candidate can be controlled by way of a challenge, control over an arbitrator who subsequently takes a role as counsel or expert is more problematic. In the latter case, Article 4 seems to be self-enforcing and the consequences for breaching the rule are unclear.
A helpful comparison can be made with current treaties. The Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), the United States–Mexico–Canada Agreement (USMCA), the 2019 Netherlands Model Investment Agreement, the 2019 Slovakia Model BIT, and the 2012 South-African Development Community Model BIT support prohibiting an arbitrator’s concurrent role as counsel when the disputes involve the same agreement or any other international agreement. However, these instruments do not show a consensus among the States on whether an arbitrator can concurrently take a role as counsel in a non-IID proceeding, nor do any of these contain a three-year tail. The panel agreed that the rationale for, and utility of, the three-year extension of the prohibition is unclear.
Disclosure as a Solution to the Double-Hatting Challenge
Finally, the last portion of the debate, led by Andrés Jana and with primary interventions by Andrea Bjorklund and Lauren Mandell, focused on the shift from the disclosure requirement in earlier drafts of Article 4 toward the narrower prohibition in the current version. The current draft represents a compromise between, on one hand, states that support a full prohibition as double-hatting creates an appearance of bias that affects the legitimacy of the system and, on the other, states that advocate for full disclosure because double-hatting creates a problem of independence and impartiality.
A portion of the panel welcomed the departure from disclosure as a solution to the double-hatting challenge while others stressed that disclosure, captured in Article 10 of the draft Code, still has a role to play in the issue. Although the current draft shows a compromise between states, one must bear in mind that the drafting process was conducted entirely remotely, which led to fewer interventions than typically seen at in-person debates. One can hope that the in-person meeting of WGIII scheduled for September 2022 will help assess whether there is a consensus on the role of disclosure. In any event, the Code would benefit from clearer links between Articles 4 and 10, especially Article 4(2) which is (for now) not subject to party autonomy.
What Comes Next for the Code of Conduct
The current draft was released for public consultation and will be further discussed in the next Working Group III session, tentatively scheduled for 19 to 23 September 2022 in Vienna. The panelists anticipate preparing a submission to WGIII capturing the night’s discussions in an effort to drive consensus. The Code is expected to be finalized in 2023.