Arbitrators are inevitably bound by a duty of impartiality, whether that arises from of article 12 of the UNCITRAL Model Law, article 11 of the ICC Rules 2021, rule 11 of the CIArb Code of Professional and Ethical Conduct, the IBA Guidelines of Conflicts of Interest 2014 or any other applicable rule. What is sure is that some incarnation of this duty will apply.
Impartiality refers to an arbitrator’s intellectual predisposition towards the parties or the subject matter or controversy at hand. It refers to the arbitrator’s frame of mind. Parties need confidence in those who adjudicate over their disputes, to prevent challenges to unfavourable awards. Duties enshrined in codes to act fairly, treat parties equally and avoid predisposition to matters other than the legitimate merit of the legal arguments and evidence arise from this need.
The subject of unconscious bias, sometimes termed psychological or cognitive bias has been laid bare in international arbitration in recent times. In the case of Halliburton it was argued that Ken Rokinson QC, was susceptible to unconscious bias, displacing the necessary, or at least perceived, equality between the parties, due to his participation in another related arbitration with one common party.
The ultimate goal must be to maintain the integrity of international arbitration as a form of justice. It is firmly rooted in the notion that justice must not just be done but must be seen to be done.
Broadly speaking, unconscious biases can be split into two types: information processing and emotional biases. Information processing biases are statistical, quantitative errors of judgment that may be mitigated if not resolved with new information. Emotional biases are more difficult to address as they are based on attitudes and feelings both conscious and subconscious.
Biases are experienced in different ways and to a differing extent by arbitrators, but even one or, as tends to be more common, a combination of some can change critical thinking and affect decision making. This can lead to a breach of the arbitrators’ duties and ill serve the interests of the parties, satisfying personal biases, sub-conscious egos and agendas instead. I will examine a non-exhaustive list of unconscious biases which are likely to affect decision making by arbitrators.
Confirmation bias is where arbitrators believe what they want to believe, favouring information that confirms pre-existing beliefs or preconceptions. This results in looking for solutions that confirm their beliefs rather than challenge them, making them closed to alternative outcomes. It is not uncommon for arbitrators to form a preliminary view and then interpret the evidence and arguments in accordance with that initial view. This is certainly the exercise counsel will engage in, arguably, necessarily to represent the best interests of their client.
Conformity bias can occur where choices of two other arbitrators influence how the third thinks, even if contrary to their own, independent judgment. This can result in poor decision making and although dissenting opinions come with their own difficulties, a lack of ventilation and discussion of diverging views and opinions within the confines of the tribunal can be very damaging to the outcome and the parties.
Authority bias can arise where an inexperienced or relatively junior arbitrator may favour an authority figure’s opinion or ideas within the tribunal. It is effectively a deferential bias. Ideas coming from arbitrators who are perceived to be more senior or superior in some way, may gain more traction even where they may be misconceived or irrelevant to the issue.
Loss-aversion bias is where once a decision has been made, an arbitrator or tribunal sticks to it due to the fear of losing what they gained in starting something and wishing to see it finished. This can occur where a tribunal may discuss their findings after the hearing and come to one decision, but in the exercise of drafting the award and setting down all the evidence and arguments realise that a different decision should be reached. As a consequence of the effort, time and energy put into reaching the initial decision, arbitrators can become biased and become emotionally attached to their original conclusions.
Self-serving bias, where arbitrators may favour decisions that enhance self-esteem. This results in attributing positive events to oneself, and conversely, negative events as blame on oneself. This can mean that decisions made can be loaded with personal agendas rather than the parties’ interests. It may be that an arbitrator had a decision publicly reviewed or expressed a view in an article and wishes to assert their legal theory rather than decide on the merits.
Framing bias is where an arbitrator may be influenced by the way in which information is presented rather than the information itself. This is particularly common where experienced and effective counsel present arguments or submissions, and in the cross-examination of witnesses, especially expert witnesses where accuracy may give way to coherence. This can work both ways, leading to an arbitrator finding nervousness suspicious as well as finding confidence persuasive. Arbitrators must take care to value substance over form.
Bandwagon bias is a commonly known bias favouring ideas already adopted by others. This is especially influential when linked to authority bias. The bandwagon effect can be a common occurrence in tribunal panels. The rate and speed at which ideas are adopted by others can significantly influence the likelihood of those ideas and concepts being selected by the group and taken forward. This bias can affect arbitrators where a case they are adjudicating over gets press coverage and is subject to media spin or public opinion.
Anchoring bias is where an arbitrator may be influenced by information that is already known or that is first shown. This causes pre-loaded and determined tunnel vision and influences final decision making. Anchoring may happen if an arbitrator feels under pressure to make a quick decision or act hastily. It can also occur where there is substantial repetition of points in a mantra like fashion.
Status-quo bias, where an arbitrator may favour the current situation or status quo and maintaining it due to loss aversion (or fear of losing it). This is a subtle bias on an emotional level that makes an arbitrator prefer what is familiar. This can affect case management decisions, especially where there are procedural or cultural differences from those with which the arbitrator is accustomed to. An example is the civil law chairman who does not wish to entertain disclosure of documents despite the parties and co-arbitrators being from common law jurisdictions, or vice versa.
Feature positive effect is where, due to limited time or resources, an arbitrator will focus on the ‘good’ benefits whilst ignoring negative effects even when the negative effects are significant. It is arguable that this has happened, for example, with remote hearings due to the pandemic, where arbitrators have been forced to undertake hearings by means of technology and may feel that either in person or remote hearings are always good, while failing to consider any negative effects.
If all these biases affect the decision-making process of arbitrators, how can parties and institutions appointing them rely on the awards they render?
The starting point must be to be aware that arbitration, like all forms of administration of justice is not flawless. It remains, however, the preferred current solution to the problem of resolving international disputes, in the fairest possible way.
When it comes to mitigating the effect of unconscious bias, the answer must be first for the arbitration community to become aware of the issues and seek to address them, individually, where possible. Ultimately, an unconscious bias is precisely that, unconscious. I therefore struggle to accept that education is the full answer. After all the most diligent of arbitrators will still be susceptible to their own biases. The full answer, in my mind, must incorporate an increase in diversity in arbitration tribunals so that one person’s bias is neutralised by another.
When I say diversity, I do not just mean more women and ethnic minorities. I always find the term ethnic minority ironically amusing in this context. However it is defined, it tends to include a demographic which makes up the majority of the global population. Anyway, I digress, what I mean by diversity includes sex, gender, race, nationality, religion, age, education, class, neurodiversity, legal background and a whole host of other factors that affect what determines our individual confirmation or authority bias or what influences our own framing or anchoring bias. Arbitral institutions and organisations are moving in that direction, but real change will only come when counsel begin to really engage with parties in this respect. In my view, the issue of unconscious bias in international arbitration must be addressed not just by individual arbitrators, but by the system as a whole.