How Advocates and Arbitrators May Use Psychology to Improve Panel Selection and Decision-Making

Since 2017, when I edited The Roles of Psychology in International Arbitration, articles on psychology have become a standard feature of edited collections on arbitration and in talks on psychology at arbitration conferences. The focus of this attention has concerned understanding the effect of cognitive biases on individual decision-making and how our brains can lead us to conclusions that aren’t fully justified. 

Focusing on the operation of individual decision-making neglects another area of psychological research directly relevant to arbitration, and in particular, construction arbitrations with three arbitrator tribunals: specifically, the ways that the decision-making of each member of a tribunal of arbitrators will be influenced by the fact that they are making a decision with other people rather than alone. This article will provide a brief overview of these group biases and suggest ways that advocates and members of an arbitral panel might use awareness of these biases to improve the quality of tribunal deliberations.

What Advocates Can Learn About Choosing Arbitrators

In 2013, researchers at University College London published the leading study in this area, “Understanding Tribunal Decision-Making,”[1] based on an empirical study of decision-making by 66 three-member tribunals. In this study, each tribunal was assigned the same dispute case file, the facts of which did not favor either party. Each tribunal was asked to issue a decision on the basis of the written documentation. Overall, 52% of the tribunals decided in favor of Party A, while 48% decided for Party B.

The researchers then examined the impact of group deliberations on the final decisions of individual tribunal members, reporting that 22% of panel members changed their view on the correct outcome of the case after group deliberations had occurred. In other words 22% of tribunal members believed after reviewing the written case file that Party A should win but, after discussions with the other tribunal members, concluded that Party B should win. In itself, this result is not particularly notable and may indicate that the tribunal members took group deliberations seriously, as one would hope. More notable, however, is the data the researchers identified regarding when these changes in outcome occurred.

In only one case out of 66 did a tribunal on which all three members believed, prior to group deliberations, that Party A should win ultimately decide that Party B should win. However, in cases when there was a majority but not unanimous position at the beginning of deliberations, in only 19% of cases did one member from the majority change their opinion so the tribunal ultimately reached a different decision. By contrast, in 50% of the cases where the panel disagreed at the outset, the minority member ultimately joined the majority’s position. 

The report made a number of observations about the factors that led members of the panel to change their minds that advocates may find instructive when selecting arbitration panel members.

Arbitrators With More Experience or Relevant Knowledge Tend to Influence Other Panelists

First, the researchers found that comparative “rank” had a clear impact on decision-making, with tribunal members who initially held a minority view significantly more likely to switch to the majority view if both members of the majority were more experienced than the member with the minority view. Second, individuals with specific knowledge relevant to the issues presented by the dispute were less likely to change their views than generalist tribunal members.

Both these results indicate that where the “right” decision is not obvious, as is usually the case in a legal dispute, and when the final outcome necessarily relies on a judgment call, as it often will, individuals without significant experience or subject-specific knowledge of their own will often rely upon the experience or knowledge of other tribunal members on the presumption that their conclusion is likely correct. 

The Existence of Group Cognitive Biases in Arbitral Decision-Making

While the psychological research on group cognitive biases reveals a range of additional impacts on panel decision-making, two biases are particularly worth noting for their relevance to arbitral decision-making and arbitrator selection.

First, researchers have found evidence of social comparison bias, or a tendency in group deliberations to favor individuals who do not compete directly with a panelist’s own strengths. For example, in a tribunal with two construction specialists and a generalist chair, each specialist will be more likely to oppose positions advanced by the other specialist and less likely to oppose the same position if advanced by the generalist chair. Each specialist subconsciously wants their own proficiency recognized, and this is challenged if they simply agree to positions advanced by the other specialist, but not if the same position is advanced by the nonspecialist chair.

Second, the bandwagon effect and conformity bias both relate to the tendency to be more likely to adopt positions that have already been endorsed by others, particularly where they are seen to reflect a majority position. So, in tribunal deliberations, it can matter a great deal who talks first, as that first speaker creates a normative foundation for the subsequent discussion and reduces the likelihood of conflicting views being advanced. It raises the level of confidence subsequent speakers must have in their conflicting views for them to be willing to express them.

How to Address Such Biases

Given the evidence that the group decision-making context influences the conclusions that an arbitral tribunal will reach, the question then becomes what actions might a panel chair consider to improve the quality of the tribunal’s decision-making? The chair is hardly the “boss” of the tribunal, but they usually are given discretion over many of the procedural aspects of the tribunal’s work. Alternatively, the other members of the panel may want to make certain that the chair follow those processes that will improve the tribunal’s decision-making. Additionally, when selecting panel members, the advocates should look for neutrals with sensitivity to these issues.

Arbitrators Should Get to Know One Another

First, an arbitral tribunal often will be composed of three individuals who have not worked together and may know one another only by name or through informal social gatherings. This initial detachment between arbitrators may be exacerbated when much of the tribunal’s work is undertaken virtually, and by email, limiting the ability of tribunal members to gain the familiarity that has been proven to benefit group deliberations. 

Indeed, a German study published in 2022[2] concludes that the extent to which judges on a tribunal knew each other professionally significantly impacted both the likelihood that dissenting views would be expressed in deliberations and the length of the justification included in the final written decision: “Familiarity cultivates a collegial atmosphere that promotes information sharing, more discussion of different opinions, and lowers barriers for individual judges to speak out against a majority opinion within the group.”[3] In short, we are more willing to be different and to stand up for a minority view we hold when we are psychologically comfortable within a group.

While there are benefits to virtual arbitration, there are actions that can be taken that maintain those benefits while increasing the familiarity of tribunal members. The tribunal may hold an initial informal meeting, even online, prior to any substantive engagement with the case to get to know one another and thereby set a foundation for future interaction. For longer cases, further periodic meetings, virtual or in person, could then be arranged, even if exchanges of emails might otherwise be sufficient, to maintain and further develop intra-tribunal familiarity. Finally, where cost allows, the chair might attempt to coordinate all members of the tribunal assembling for both the hearing and deliberations, even if the hearing is online, with the tribunal members gathering in one location to join the online hearing.

Avoiding the Effects of Cognitive Dissonance or Premature Dominance by One Panellist

Research on cognitive dissonance confirms that publicly expressing a view increases the likelihood that we will maintain that view, while the research discussed above states that we are more likely to adopt beliefs and behaviors when they have already been adopted by others in the tribunal. Given this context, a chair—or other members of the panel—may wish to suggest that prior to in-person deliberations, each tribunal member (including the chair) prepares a written statement of opinion on the primary issues to be decided and that deliberation on each issue commences with the chair reading those opinions aloud. Such a procedure would create an incentive for each tribunal member to offer an effective defense of their own initial view and reduce the likelihood of any member simply acquiescing to the majority view.

These are merely two examples of ways that the psychological research on group cognitive biases can be implemented in arbitration to improve the decision-making of tribunals. The goal of this article is to raise awareness of this research, which has the ability to substantially improve the quality of work of any arbitral tribunal.

[1] Cheryl Thomas & Hazel Genn, “Understanding tribunal decision-making: A foundational empirical study”, available at https://www.nuffieldfoundation.org/wp-content/uploads/2019/12/NUFJ0000_Tribunaldecision_making_27_03_14.pdf (last visited 3 January 2025).

[2] Tilko Swalve, “Does Group Familiarity Improve Deliberations in Judicial Teams? Evidence from the German Federal Court of Justice,” 19 Journal of Empirical Legal Studies 223 (2022).

[3] Id. at 240.