ARBITRATION AND RULE PRODUCTION.

• Author: Drahozal, Christopher R.

CONTENTS INTRODUCTION I. THE USE OF ARBITRATION IN RELEVANT CONTRACTING MARKETS A. Defining the Relevant Contracting Market B. Classifying Contracting Markets C. Changes in the Use of Arbitration Clauses Since Conception II. RULE PRODUCTION IN A PARTIAL-ARBITRATION MARKET AND THE DISPLACEMENT HYPOTHESIS A. Models of Court Precedent Production B. Strategic Precedent Production and Arbitration III. RULE PRODUCTION IN AN ARBITRATION-ONLY MARKET AND THE POSITIVE EXTERNALITIES HYPOTHESIS A. Rule Production as a Private Good B. Arbitral Practice and Rule Production CONCLUSION APPENDIX INTRODUCTION

Like settlement, (1) arbitration has been criticized as displacing cases from the public courts and thereby reducing the production of court precedent. For example, Richard Alderman has contended that "[arbitration eliminates litigation in a public forum, precedent-establishing decisions, and stare decisis." (2) Myriam Gilles has argued that "[f]or the entire categories of cases that are ushered into this [arbitration] vault--from consumer law, to employment law, to much of antitrust law--common law doctrinal development will cease." (3) Charles Knapp has stated more broadly--and more colorfully--that "[i]f all contract disputes which the parties could not settle between themselves had to be submitted to arbitration for resolution, rather than to a court of law, the common law of contract ... would become merely an historical relic, a legal King Tut in its elaborately detailed Restatement (Second) sarcophagus." (4) I call this the "displacement hypothesis." (5)

Concerns about the displacement of court precedent by arbitration are not limited to academics. Judge Jennifer Walker Elrod of the United States Court of Appeals for the Fifth Circuit, for example, has asserted: "We have a common law system that is enriched by progression and development of law through cases later resolved on appeal. Such development only happens if cases are tried in public in courts of law [rather than in arbitration].... Without cases, our common law will stagnate...." (6) Nor are the concerns limited to the United States. Lord Neuberger, the now-retired President of the Supreme Court of the United Kingdom, has stated: "One of the disadvantages of an increase in [arbitration] awards and a concomitant decrease in judgments, particularly in the common law world, is that the law does not develop, that it becomes ossified." (7)

Arbitration differs from settlement in that the result of the arbitration proceeding--an arbitral award--might itself produce legal rules and substitute for court precedent. But the standard view is that arbitrators have little incentive to issue awards that produce legal rules. In their classic article, Adjudication as a Private Good, William Landes and Richard A. Posner argue:

[B]ecause of the difficulty of establishing property rights in a precedent, private judges have little incentive to produce precedents. They will strive for a fair result between the parties in order to preserve a reputation for impartiality, but why should they make any effort to explain the result in a way that would provide guidance for future parties? To do so would be to confer an external, an uncompensated, benefit not only on future parties but also on competing judges. (8)

In other words, rule production (as opposed to dispute resolution) results in positive externalities--benefits conferred on nonparties to the dispute. Because the parties to the dispute do not receive the full benefit of a rule-producing award, they will not pay the arbitrator enough to issue such awards. As a result, the argument goes, a system of arbitration will produce too little precedent. (9) Other commentators, such as Professors Gilles (10) and Alderman, (11) have echoed this analysis. (12) 1 call this the "positive externalities hypothesis."

This Article critically examines both the displacement hypothesis and the positive externalities hypothesis; it analyzes both the likelihood that arbitration will displace court precedent and the incentive of arbitrators to produce awards that can substitute for court precedent. (13) In so doing, the Article offers new theoretical and empirical insights that provide the groundwork for a comprehensive account of arbitration and rule production. Its central insights are the following:

First, the Article considers the underlying factual predicate for the displacement hypothesis: the extent to which arbitration is used to resolve disputes in the relevant contracting market. I define the "relevant contracting market" as the type or types of contracts likely to give rise to disputes raising legal issues in a particular field of law and in a particular jurisdiction. Determining the relevant contracting market is analogous to determining the relevant market for antitrust merger analysis: it requires consideration of both the relevant product market (the field of law at issue) and the relevant geographic market (the jurisdiction in which the law is applicable). (14) This analysis has two implications for understanding arbitration and rule production: (1) the relevant field of law typically is not "contract law" or the "common law," but instead "employment law" or "franchise law" or even narrower fields of law; and (2) jurisdictions with less use of arbitration might produce court decisions that can serve as (at least persuasive) precedent for jurisdictions with more use of arbitration.

Second, once the relevant contracting market is defined, the next question is what proportion of disputes in that contracting market are resolved in court and what proportion are resolved in arbitration. Some commentators have argued that sufficient cases will continue to be resolved in court so that fears about the displacement hypothesis are unwarranted. (15) According to Rick Bales, for example, "there will always be some employees ... with the bargaining power to refuse [arbitration] agreements, and there will be some jobs ... for which employers will prefer the barriers to access of litigation over arbitration. Thus, there will still be plenty of employment cases and judicial opinions." (16) Other commentators, however, have predicted that all businesses are likely to switch to arbitration clauses in their standard form contracts with employees and consumers. As Professor Brian Fitzpatrick has stated: "[Arbitration clauses with class-action] waivers are tantamount to insulating businesses altogether from liability for the small-stakes injuries they cause. Why wouldn't every business want such insulation? I think every business would." (17) In response to Professor Fitzpatrick's question, this Article offers several possible reasons businesses might not use arbitration clauses and presents anecdotal and empirical evidence consistent with those reasons. Moreover, widespread use of arbitration clauses alone is not enough to make a contracting market "arbitration-only"--i.e., one in which all disputes are resolved in arbitration. In addition, the arbitration clauses must not carve out relevant disputes from arbitration, parties must not opt out of the obligation to arbitrate, and parties must invoke arbitration clauses in the relevant disputes.

Third, the Article examines the effect of arbitration on rule production by courts in a partial-arbitration market--that is, a contracting market with some, but not all, disputes resolved in arbitration. Some commentators, accepting a simple version of the displacement hypothesis, assume an inverse, perhaps even linear, relationship between the use of arbitration and production of court precedent. (18) On this view, more arbitration means less court precedent. Others recognize that the relationship may not be so simple. Professor Knapp, for example, acknowledges that "the widespread use of arbitration by willing parties" has not been and is not "likely to be fatal to the common law of contract." (19) "After all," according to Knapp, "disputants who choose to arbitrate might instead have simply chosen to settle their disputes on their own, without resort to any external decision-maker." (20) This Article extends the analysis in Marc Galanter's Why the "Haves " Come out Ahead to explain how businesses might use arbitration clauses (instead of settlements) to avoid the creation of unfavorable court precedent. (21) One implication of this analysis is that, as suggested by Knapp, increased use of arbitration might displace settlements (of cases otherwise likely to create unfavorable law) rather than reducing the production of court precedent. (22)

Finally, the Article reevaluates the positive externalities hypothesis by considering rule production in an arbitration-only market. While the positive externalities hypothesis is widely accepted, Mark Weidemaier and others have demonstrated that arbitrators do, in fact, rely on rules created by arbitral precedent in a variety of settings. (23) This Article builds on the work of these commentators by squaring this empirical reality with the theory underlying the positive externalities hypothesis. (24) The positive externalities hypothesis incorrectly focuses on the incentives of arbitrators, rather than the incentives of the parties to the dispute, and overlooks the role of arbitral institutions in the arbitration process. (25) On this expanded view, there is good reason to believe that arbitrators will often issue reasoned awards and that at least some portion of those awards will be made public. (26) Indeed, current practice in consumer, employment, and international arbitration belies claims to the contrary. (27)

Part I of this Article addresses the need to define the relevant contracting market, explains why the widespread use of arbitration clauses is a necessary but not sufficient condition for a contracting market to be arbitration-only, and presents empirical evidence on the use of arbitration clauses in various contracting markets. Part II examines the...