Georgia Law Review, Vol. 39, No. 1 (Fall 2004), pp. 151-214


This Article breaks from conventional wisdom in both case law and scholarship. It proposes a simple but novel thesis: Arbitrators and arbitral institutions, in cases of voluntary submission of disputes, should not be entitled to any form of legal immunity. Instead, any limit on or waiver of the arbitrator's or institution's liability should come in the form of a contractual release-either adopted in the parties' arbitration agreement or negotiated between the parties and the arbitrator.

Central to this thesis is a distinction between two types of immunity. The first form of immunity is “contractual immunity.” The hallmark of contractual immunity is an agreement to release the arbitrator and, where relevant, the arbitral institution from liability.

The second form of immunity is “legal immunity.” Unlike contractual immunity, legal immunity does not depend on a term of the parties' agreement or the arbitrator's mandate. Instead, the legal immunity applies by operation of law (typically the law of the arbitral forum) regardless of the parties' intentions.

The distinction between contractual and legal immunity is central to the market-based model of dispute resolution developed in this Article. Under this model, arbitration is one type of “product” in a market for dispute resolution services. A distinguishing feature about this product is that, with few exceptions, it is a voluntary undertaking, not simply among the litigants, but between them, the arbitrator, and, in many cases, an arbitral institution. The parties, the arbitrator, and the arbitral institution can dicker over the terms of their relationship just like professionals in any other setting.

A rule of legal immunity is inconsistent with the market-based model. It discourages the use of arbitration by forcing the parties involuntarily to surrender a remedy when there is misconduct by the decisionmaker. It drives up the price of arbitration by allowing the arbitrators to reap the benefits of immunity without having to make a transfer payment in return for the parties' foregoing the right to sue. Finally, it undermines competition among providers of dispute resolution services because arbitrators cannot differentiate themselves along this axis.

By contrast, limiting immunity to its first form (contractual immunity) comports with the market-based model. It preserves the parties' freedom to structure the arbitration along their preferred lines. It ensures that any waiver of remedies by the parties will be effectively compensated in the form of a transfer payment from the arbitrator. Finally, it enables competitors in the dispute resolution marketplace to differentiate themselves by offering the parties partial or complete remedies as a guarantee of their performance (as exemplified by the ICC and AAA rules).

This Article develops the foregoing argument in five parts. Part I develops a market-based theory of dispute resolution. It shows how the market for dispute resolution services has become a robust, competitive one, marked by competition among various types of services (mediation, arbitration, judicial resolution) and among providers of a particular service (arbitrators). Part II uses this market-based approach to criticize the traditional explanations offered for the immunity doctrine—arguments based on the arbitrator's function and the purposes served by immunity. Neither the functional argument nor the policy arguments fare well when subjected to the market-based paradigm developed in Part I. Part III uses this approach to analyze proposals in the scholarship on arbitral immunity—qualified immunity, limited liability, and institutional liability. Each of them fails to correct the price distortions, the barriers to entry, and the underutilization of arbitration caused by immunity. Part IV applies this approach to defend a model replacing the current system of legal immunity with a system of liability subject to contractual waivers. It answers anticipated criticisms based on the unequal bargaining power of the parties and the potential litigation prompted by a liability rule. It also explores the limitations of the thesis when one relaxes certain assumptions about the voluntary nature of the transaction, the availability of information, and the viability of insurance markets.