Are international arbitrators biased in favor of the nominating party? The answer is: probably yes! This is in part the result of what some scholars refer to as selection and affiliation effects.
Selection effect: Because the parties nominate the arbitrators, they can choose party-appointed arbitrators with the maximum predisposition towards their case. The litigant can draw from the pool of arbitrators who have shown reliability and the appropriate attitudes towards the party’s case.
Affiliation effect: Because the parties nominate the arbitrators, party-appointed arbitrators may have some predisposition to favor the nominating party. This predisposition may operate as an implicit bias; as a result, arbitrators would have a tendency to side with the nominating party even if he or she attempts to maintain neutrality and independence.
This distinction between selection and affiliation effect is very important, although the two effects are often confounded. Only the first can be directly addressed with blinding appointments—preventing nominees from knowing which party appointed them. How strong then is the affiliation effect among arbitration professionals?
In a project with Anton Strezhnev, we designed an experiment to measure affiliation effects in international arbitrators by manipulating the source of the appointment to the tribunal. We randomly assigned actual international arbitrators participating in our experimental survey to one of two conditions: a party-appointment or a blind appointment. After presenting the survey respondents with a brief vignette describing a hypothetical arbitration scenario, respondents were asked how they thought the parties’ arbitration expenses should be apportioned. By randomly manipulating the appointment source, any observed variation in response can be attributed to the affiliation of the arbitrator itself as opposed to the selection by the litigant.
Based on the results of 266 participants, we estimate that assignment to being appointed by one of the parties to the dispute made a respondent roughly 20 percentage points more likely to shift costs to the non-appointing party, suggesting a strong affiliation effect when arbitrators exercise discretion. Our results suggest that blinding could be an effective measure to mitigate affiliation effects resulting from the current practice of unilateral party appointments. To be sure, the limitations of our experiment are clear; it is unable to assess how exactly implicit biases of individual arbitrators affect outcomes of the collective body. Nevertheless, if the affiliation effect is as strong as we believe based on the experiment, a blinding mechanism could improve the practice of arbitration significantly.