Separate Opinions and Appointments in Investment Arbitration: Signalling and Crystallisation
Introduction
Are separate and dissenting opinions in international investment arbitration proceedings a way for the ‘opiniated’ arbitrators to ‘market’ themselves in view of further appointments?
There are reasons to think either way. On the one hand dissenting opinions add to the public record of an arbitrator, putting his or her name on the map, and assisting the due diligence of counsel and parties to pick the ‘best’ arbitrator to support their case.
On the other hand, penning an unduly harsh dissenting opinion (or an overly enthusiastic concurring opinion) could backfire: the opinion’s author might paint her or himself into a corner, resulting in decreased chances of being re-appointed, and increased chances of challenge. (There is anecdotal evidence for this thesis, although, of course, it is never clear whether arbitrators in these circumstances were never re-appointed or just refused to be re-appointed.)
Fortunately, there is probably now enough data on arbitral appointments to answer this question, at least to an extent. In what follows, I study two hypotheses:
- That arbitrators who publish an individual opinion are appointed more frequently three years down the line. (Answer: yes, it seems)
- That separate opinions help to ‘crystallise’ the profile (pro-investor or pro-state) of an arbitrator. (Answer: not really, on the contrary)
But first, a note on this article’s methodology and some raw (and updated) numbers about arbitrators and tribunals.
Methodology
The code used to analyse data for this article is available on this Jupyter Notebook. Data is on file with me.
I collected data on appointments mostly from the database of the International Centre for Settlement of Investment Dispute (ICSID) for ICSID disputes and from UNCTAD’s Investment Dispute Settlement Navigator. (Although I also collected appointment data on annulment committees, I focus only on original or resubmitted proceedings in this article.)
This resulted in a database of some 827 disputes with at least one appointed arbitrator. The arbitrators themselves numbered 576, with around half of them (n=302) appointed only once up to January 2018. The distribution is as follows (notice the bump at 94: this is Brigitte STERN):
I also noted the roles of the arbitrators (i.e., as a claimant- or respondent-appointed arbitrator, or a chair) for each appointment. Remarkably, the arbitrators’ first appointment was equally shared between claimants and respondent:
I also collected the dates of each appointment. This data is partly available with ICSID, but that still left a hole in the dataset for non-ICSID cases and cases where ICSID itself was silent. I also lacked information on the date of the start of the proceedings for around 200 cases, although I knew at least the year.
Based on the data I could collect, I learned that the median timeframe to constitute a tribunal is 170 days from the start of the proceedings,* while the median time to appoint any individual arbitrator is 150 days (but only 118 days for the wing arbitrators).
To remedy the missing appointment times, I (i) interpolated the exact start of the proceedings based on the disclosed year; and (ii) assumed that the date of appointment was the starting date (real or interpolated) + 150 days.
Finally, I needed to know the list of individual opinions written in investment proceedings and their authors, with the date thereof. I collected this data from Italaw.** I found 64 arbitrators penning a total of 103 dissenting or separate opinion. (In what follows, I do not account, however, for the fact that the text of some opinions was disclosed only some time following its nominal date.) The distribution of dissent is remarkably similar to that of cases alongs the years.
Finally, a caveat that bears repeating again and again: the following data is based only on public information and disclosed arbitrations. Many further arbitrations, likely involving the same arbitrators, remain confidential; meanwhile, many expected appointments were maybe secretly thwarted for a reason or another. As such, the conclusions below can only be taken cautiously.
First Hypothesis: More opinions, more appointments down the line
Loosely speaking, signalling is the process through which an agent communicates some of its attributes to an external audience fluent in this communication.
In the realm of international investment arbitration, arbitrators could — consciously or uncounsciously — signal their values, preferences, and overall inclination towards the party that appointed them by way of an individual opinion. In this model, compared to other potential picks, the opinion-writer becomes a relatively safer bet for the party looking for an arbitrator to appoint.
It is however difficult to test this hypothesis.
At a superficial level, arbitrators who indulged in putting their individual opinion on paper have had, on average, twice as many appointments in total as of January 2018. But this does not tell us that the opinion caused the higher frequency of appointments.
Indeed, since the two variables (existence of an opinion, number of appointments) are not truly independent, it is hard to test whether one additional opinion helps for future appointments. The reverse relationship — more appointment, more chances to write an opinion — could hold instead.
The alternative I followed for this piece is as follows: looking at a given arbitrator publishing an opinion (‘Arb_X’) at any given time, does Arb_X receives more appointment in the next three years than an average arbitrator Arb_Y, averaged over a section of peer arbitrators with the following characteristics:
- Being appointed and active at the same time as Arb_X wrote its opinion; and
- Having nearly as many total appointments (i.e., being in the same tier) as Arb_X?
In the event, I found that yes: in the three years following her of his first individual opinion, Arb_X will have another 2.06 arbitral appointments compared to 1.6 for his peers (embodied in the average Arb_Y).
If I repeat the analysis over all individual opinions (n=103, compared to n=65 for first opinions), the ratio still hold: Arb_X will have another 3.03 appointments in the future three years, whereas the average Arb_Y will have only 2.25.
Given that most arbitrators are appointed only once, however, I repeated the analysis taking into account only the arbitrators (be it Arb_X or the set of peers composing Arb_Y) who would go on to get at least one appointment following the opinion’s date. This is the third column below.
So, there seems to be a small edge for arbitrators who write an individual opinion compared to the average of a set of comparable arbitrators, especially for arbitrators who are not appointed only as a ‘one-off.’
Second Hypothesis: Crystallisation of arbitrator profile
A second hypothesis would be that Separate Opinions could crystallise the “profile” of the arbitrator, either as ‘pro-State’ or ‘pro-investor.’ (It is more uneasy to market yourself as a good Presiding arbitrator with the help of an individual opinion — presumably the chair is already writing the award, and individual opinions written by chairs are extremely rare.)
To test this hypothesis, I retraced the appointments of the 23 arbitrators who had had at least 3 appointments before and after signing their first individual opinions.
In the graphs that follow, appointment are retraced chronologically (‘Cl’: ‘Claimant’, ‘Re’: ‘Respondent’, ‘Pr’: ‘President’), with the middle line representing the first individual opinion. The middle line is blue for opinions issued as claimant-appointed arbitrator, red for respondent-appointed arbitrator.
With the caveat that, obviously, the dataset is limited, it seems that the crystallisation hypothesis does not hold. On the contrary, following the release of an individual opinion in a case, a given arbitrator’s appointment profile is if anything more volatile.
The reason for this finding might be that issuing an opinion contributes to the ‘seriousness’ of an arbitrator, eventually enhancing his or her value in the eyes of both types of litigants.
Conclusion
I found that opinions are likely correlated with further appointments down the line, and a slightly higher diversity in future appointments (again, accounting for the limited dataset, and my own limited skills at analysing it). What to make of this?
One tentative conclusion is that writing a dissenting opinion is likely beneficial in terms of future appointments, maybe by elevating the author’s profile in the world of the investment arbitration community.
Previous research noted that the arbitral appointment market is characterised by assymetries of information, which mainly benefits insiders. In the competition between insiders, however, writing an individual opinion might be a way to get an edge by enhancing your profile at all levels (with insiders and outsiders).***
This in turn would mean that parties and counsel are generally more interested in the ‘prestige’ of an arbitrator, and not on her or his inclinations towards a party or another. This is an hypothesis that makes sense: a prestigious arbitrator is more likely to convince at least one other member of the tribunal of his or her opinion. There is no point in appointing a persistent objector, or an arbitrator to committed to one side.
Of course, this is not suggesting some cynical intention on the part of the opinion writer. Reasons to pen an individual opinion are numerous, and they can make important contributions to international investment law. Nonetheless, in a context where most arbitrators are appointed only once whereas some star repeat players collect appointments, this is an interesting correlation to keep in mind.
Footnotes:
* This is in line with past findings, such as this piece from WilmerHale which notes that it takes “ an average of nearly seven months to constitute [an ICSID] arbitral tribunal.”
** I was also greatly helped by the brilliant database of international economic disputes set up by Wolfgang Alschner and Aleksander Umov: see here.
*** Malcolm Langford, Daniel Behn, Runar Hilleren Lie, ‘The Revolving Door in International Investment Arbitration’ (2017) 20 Journal of International Economic Law 2, 301–332, citing notably Catherine Rogers, ‘The Vocation of the International Arbitrator’ (2004) 20 American University International Law Review 957, at 968–969.