Evaluating Employment Arbitration: A Call for Better Empirical Research

Abstract

Since at least 1991, issues surrounding mandatory arbitration of employment and other disputes have intrigued, perplexed, angered, gratified, and confounded academics, politicians, lawyers, and others.As with many legal issues, the first wave of scholarly work centered on the law. As the law has pretty much settled, academics have turned to empirical work, focusing on how employment arbitration works, and how it compares to employment litigation. In part due to pressure from California legislation, the American Arbitration Association (“AAA”), the nation’s leading provider of arbitration services, opened access to its data base. Owing to inevitable data limitations, most analyses have focused on outcomes—comparisons between litigation verdicts/judgments and arbitration awards. This work suffers from a serious selection bias, for there is no reason to believe that cases that result in arbitration awards are otherwise comparable to cases that have survived the several serious gauntlets that lie in the path of a case before it goes to trial. Because mediation is often an initial step in most employment arbitration system and arbitrators are not likely to consider dispositive motions, weaker cases are likely to get to a hearing in arbitration than in court.

We wholeheartedly endorse good empirical work as an important means of understanding and addressing controversial policy issues, especially in the arbitration arena, and tried our hand at such work a decade ago. We have written this paper to encourage research that goes beyond evaluating awards within the AAA data set and to engage in a longitudinal study of the history of claims—from when they are initially filed with administrative agencies or arbitration organizations to when they are settled or adjudicated.