What could be more basic? Arbitrations begin with each side naming an arbitrator. References are occasionally made to “the fundamental right” to name one’s arbitrator.

But there is no such right. Moreover, if it existed, it would certainly not be fundamental. The original concept that legitimates arbitration is that of an arbitrator in whom both parties have confidence. It is quite unlikely that a party will have confidence in an unknown arbitrator selected by its unloved opponent.

We must confront an uncomfortable fact: as illustrated by two studies of ICC cases in different years earlier this decade, dissenting opinions were almost invariably (more than 95%) the work of the arbitrator nominated by the losing party. (See Alan Redfern, “Dissenting Opinions in International Commercial Arbitration: the Good, the Bad and the Ugly,” 2003 Freshfields Lecture, 20 Arbitration International 223 (2004); Eduardo Silva Romero, “Brèves observations sur l’opinion dissidente,” Les arbitres internationaux in Société de législation comparée at 179-186 (2005)).

This troubling record is at least matched in the newer field of treaty-based arbitrations brought by foreign investor against states. The posture adopted by ad hoc judges before the International Court of Justice, almost invariably espousing the theses of their appointing states, has over many years – with a very few notable exceptions – been similarly inglorious.

The common practice of unilateral appointments of co-arbitrators may be said to be popular, in the sense of being perceived as a valuable opportunity on which many parties insist. This seems clear from the frequency of explicit contractual stipulations to this effect, as well as from similar provisions found in rules devised by arbitral institutions. It may seem surprising to ask whether this is compatible with the very concept of arbitration. Yet it is a serious question. It touches on a simple but profound matter. The legitimacy of arbitration rests on the parties’ confidence in arbitrators. There is every reason to doubt that any party is willing to have its case decided by its opponent’s chosen arbitrator.

The two main purported justifications for the practice of unilateral appointments are hardly convincing. It is said that (i) even Homer nods and therefore three heads are better than one, especially when the stakes are high, and (ii) parties will have greater confidence in the selection of arbitrators selected for their special knowledge or skill. Yet each of these objectives may just as easily be met when all three arbitrators are chosen jointly or by an appointing authority. (The trustworthiness of appointing institutions will be considered at the end of this contribution.)

Let us think it through. We need not be concerned with troublefree cases that result in unanimous awards. (Fortunately, unanimity seems prevalent in international commercial arbitration; this is perhaps the great unrecognized achievement of the culture of arbitration.) Yet our model must also work in the difficult cases, which is of course where any system is tested, and thus exposed to criticism and disaffection. Is a 2:1 decision perceived by the losing party as more legitimate than a decision by a sole arbitrator, because “three heads are better than one”? That makes no sense – quite the opposite. The losing party in a difficult case is likely to consider that it appointed a “good” arbitrator, who has somehow been outvoted by the “bad” arbitrator appointed by an unscrupulous adversary and a feckless chairman misled by the “bad” arbitrator. QED: in the eyes of the losing party, the 2:1 decision is less legitimate than that of a sole arbitrator in whose selection the opponent had no more and no less than an equal say. (The pejorative assumptions – “unscrupulous”, “bad” – reflect the fact that we are ex hypothesi talking about difficult cases.)

Let us also consider the oft-repeated objective of enhancing each party’s confidence that it will get a fair hearing if one arbitrator (not biased, of course) “understands its legal system and culture.” This feature may increase the likelihood of consensual submission to a particular jurisdiction. Such a strategy for engendering confidence is understandable on the part of fledgling arbitral bodies. But why should mature institutions need such mental struts? The ICJ itself was conceived – not without controversy, including the severe condemnation of Hersch Lauterpacht (The Function of Law in the International Community, 424 (1933)) – on a similar premise: the establishment in the Court’s Statute of a right for any litigant to name a judge ad hoc of its choice in the event none of the 15 titular judges is of its nationality. (Cases before the ICJ are like arbitrations in that they cannot be brought in the absence of consent to jurisdiction.) The practice carries with it a risk of perverting arbitration’s original legitimating impulse.

This has nothing to do per se with the choice between a sole arbitrator and a tribunal comprising three or more arbitrators. In either case, when each arbitrator is chosen jointly by the parties, or is appointed by a neutral institution, each is invested with an equal measure of confidence and an equal claim to moral authority.

Not so when there are unilateral appointments. It seems obvious that a very different impulse is at work here, introducing an adversarial element into the very deliberation of the arbitral tribunal. Disputants tend to be interested in one thing only: a favourable outcome. They exercise their right of unilateral appointment, like everything else, with that overriding objective in view. The result is speculation about ways and means to shape a favourable tribunal, or at least to avoid a tribunal favourable to the other side – which is logically assumed to be speculating with the same fervour, and toward the same end. Forgotten is the search for the arbitrator trusted by both sides.

Yet all is not lost. The arbitration may be taking place in an environment where a sufficient number of individuals enjoy a solid reputation for impartiality and insight. To name a partisan arbitrator from outside that circle is to run the risk that the two other arbitrators will deliberate within an intellectual zone of shared confidence into which the partisan arbitrator has no access. For a party convinced in the merits of its case, it would be folly not to name one of the recognized impartial arbitrators. For a party not so convinced, it would be pointless to do so; the only avenue left, if the case cannot be settled on tolerable terms, is to try to seek to achieve damage limitation by skilled advocacy – and in this hypothesis that party may as well burnish its bona fides by selecting one of the recognized impartial arbitrators.

The only solution which will be reliable in all circumstances is that any arbitrator, no matter the size of the tribunal, should be chosen jointly or selected by a neutral body. This aspect of the process should no longer be misused as a sales argument for arbitration. Confidence-enhancement is properly focused on procedural rights (the right to be heard, the opportunity to confront the opponent’s case, equality of arms) rather than risking the ineluctable contamination of the idealthat of an arbitrator trusted by both sides – by notions of constituency.

That said, recognizing the likelihood that this genie – the “right” to appoint an arbitrator – cannot easily be put back into the bottle, there are ways of reducing contamination.

One involves the restriction of unilateral nominations by specific contractual limitation, such as a requirement that no arbitrator may have the nationality of any party. In the absence of such a restriction, a party may find it politically impossible not to name one of its nationals as arbitrators. Worse, that nominee may feel subject to political pressures – whether he or she succumbs to them or fights them. Such restrictions, in other words, are capable of reducing the risk of the subversion of arbitral authority.

An even more effective mechanism, provided that it is properly conceived, may be an institutional requirement that unilateral appointments be made from a pre-existing list of qualified arbitrators. The danger here is that an arbitral institution ends up skewing the list to favour an “in-group” operating as an opaque oligopoly. Still, when composed judiciously by a reputable, inclusive, and continually replenished international body, such a restricted list may have undeniable advantages. In fact, it may be seen as a hybrid of institutional and unilateral appointments; a party may indeed select any one of a number of arbitrators, but each of the potential nominees has been vetted by the institution and is less likely to be beholden to the appointing party.

An example is that of the international body created in 1985 as the Court of Arbitration for Sport. Most of its panels are comprised of three members. Some cases are of international notoriety, involving the disqualification of famous champions or the transfer fees of wealthy professional athletes. Most CAS arbitrations, however, involve far more modest disputes and are resolved after hearings that take less than a full day. Of present importance is that (i) the parties come from all parts of the globe, and (ii) that most of them are involved in such proceedings for the first time of their lives. If they were given the unfettered right to make a unilateral nomination, they would in all likelihood – out of ignorance, fear, or calculation – appoint someone unknown, and practically speaking shielded from any meaningful verification by reason of coming from an environment wholly unfamiliar to the other participants in the process. Such an appointee would have no moral standing with the presiding arbitrator beyond what is possible to reveal in a few hours of collaboration – which is not very much, nor very reliable. The result might be that the presiding arbitrator would tend to decide alone, thus defeating the purpose of three-member tribunals.

The CAS solution is to require all nominees to be found on a list of qualified arbitrators. It is of crucial importance that this list is lengthy and inclusive, containing names from all over the world. (In 2008, there were 267 arbitrators on the CAS general list; 61 – with some overlap – on the specific list for football.) Any party is free to choose the arbitrator it considers the best for its case. However selfish its motives, it is restricted to this list of prequalified individuals – and it knows that the same is true for its opponent.

One may well wonder if parties, when drafting arbitration clauses, have thought through the dynamics of three-member tribunals and their practical effect. The right to name “my” arbitrator is counterbalanced by the opponent’s right to do the same thing. Why should I assume that I will play this game better than my opponent? The two evident solutions (apart from restricting the pool of eligible nominees by nationality, professional status, or specific pre-qualifications such as that of the CAS) are (i) to opt for a sole arbitrator to be chosen, failing agreement, by a highly reputed institution or, (ii) if the true concern is that the case is too important to risk the lapse of even the most outstanding individual person (Homer’s nod), three arbitrators appointed in the same way, i.e., eschewing any unilateral nomination. Institutions may experiment with a variety of solutions, such as “blind appointments” (i.e. seeking to ensure that nominees do not know who appointed them) or simply a variety of list procedures which have in common the feature that the initial identification of the candidate comes from the institution rather than from one party.

The two main purported justifications for unilateral appointments (see above) are perhaps hypocritical. Parties in a given case care less about a tribunal’s getting the true answer than about winning the case. So the real reason for clinging to the opportunity of unilateral appointments is very likely reluctance to trust the relevant appointing authority. There is of course a flaw in this self-interested analysis as well. For surely it is as easy to imagine that one’s adversary’s nominee will be able to tilt the third arbitrator as to imagine that one’s own nominee will keep him straight. Instead of asking one’s client “are you ready to give up your right to name an arbitrator?” should the advocate not rather ask “why don’t we try to convince the other side to waive its right?” Yet in the end many parties instinctively feel that giving up this “right” is to relinquish a measure of control – no matter how illusory. This instinct is unlikely to be muted unless and until relevant appointing authorities are accorded full confidence.

These reflections on one sensitive subject thus lead ineluctably to consideration of a second, equally important and delicate matter: the standards and practices of arbitral institutions. None of the solutions imagined above could lead to significant progress if appointing authorities cannot convince the participants in the arbitral process that their selection process is absolutely untainted by undue influence, and to the contrary is dependably and uncompromisingly focused on selecting the best arbitrators for the job. But that is another story which deserves treatment on its own.



  1. Jan is actually advocating the Dutch legislator’s view. While the legislator does not prohibit party-appointed arbitrators, it is not its preferred method either. For the very reasons set out in Jan’s contribution, the Netherlands Arbitration Act of 1986 provides that when parties have not agreed on a method for appointing arbitrators, all arbitrators must be appointed by consent of all parties (Art. 1027(1) CCP). This is in contrast to most arbitration laws, including notably the Uncitral Model Law, which have as default method the system of party-appointed arbitrators. Following Jan’s comments, the Model Law system of appointment may be a candidate for future amendment.

    I would like to add a comment on the list procedure. Jan mentions the CAS. It was actually the AAA that, to my knowledge, has invented the list procedure to avoid party-appointed arbitrators. I have some 8 year of experience myself with the list procedure as SG for the NAI (which had copied it from the AAA). It involves the drawing up of a list of candidate arbitrators (minimal 3 for a sole, and minimal 9 for a 3 member panel), taken from a large General List of Arbitrators (pre-screened and now on a data base). Each party receives an identical list. Each party may delete the names of candidates against whom that party has a major objection and number the remaining names in the sequence of its preference. The list must be returned to the Registrar within 2 weeks. Key in the process is that neither party knows how the other fills out the list. It is interesting to note that the preferences coincide in 70-80% of the cases. The advantages of the list procedure are that the parties have a certain influence regarding who will be nominated and that they are not tainted by the factor that a candidate is suspect for the mere reason that he or she is proposed by the other side.

  2. Since last August I have watched international arbitration cases from my arm chair and I marvel at how some issues, such as selecting an arbitrator, continue to spark so much comment and emotion. In 1988, when I started working with Mike Mealey the selection of arbitrators was an issue that could delay or derail a multi-million dollar case for many months leaving parties and expensive projects in limbo.

    With so many issues swirling around a party as it proceeds to arbitration, selecting an arbitrator from a list, would offer order to both sides. A list of qualified candidates would extinguish some of the alleged politics and lobbying that arbitral candidates face.

    In 1996, while attending an ICC international arbitration conference in Boston, I witnessed several heated arguments in an audience when the issue of partiality of arbitrators was discussed. A list of arbitral candidates could address such an issue and perhaps bring a more temperate climate to the arbitration hearing room.

Leave a Reply

Your email address will not be published. Required fields are marked *