David Horton


For decades, the U.S. Supreme Court has insisted that forcing a plaintiff to arbitrate—rather than allowing her to litigate—does not affect the outcome of a dispute. Recently, the Court has invoked this “parity assumption” to expand arbitral jurisdiction. Reasoning that it does not matter whether an arbitrator or a judge resolves a particular issue, the Justices have allowed arbitrators to decide important questions about the arbitral proceeding itself.

The parity assumption has proven impossible to test. First, cases that are arbitrated differ from those that end up in the judicial system, complicating efforts to compare outcomes from each sphere. Second, arbitral awards are rarely published and thus remain shrouded in mystery.

However, one important topic defies these limitations. Jurisdictions are divided over whether courts or arbitrators should perform a task known as “clause construction”—determining whether an arbitration clause that does not mention class actions permits such procedures. As a result, both judges and arbitrators have been weighing in on the same question. Moreover, because class members are entitled to notice of rulings that impact their rights, the American Arbitration Association requires arbitral clause-construction awards to be available to the public. For once, then, it is possible to assess how the two kinds of decisionmakers resolve the same issue.

This Article capitalizes on this opportunity by analyzing a dataset of 150 recent judicial and arbitral clause-construction decisions. Its logit regression analysis concludes that arbitrators are nearly 64 times more likely than judges to allow class actions. This Article then uses its findings to propose a solution to the circuit split over clause construction and to inform the broader debate over the boundaries between judicial and arbitral power.

Included in

Law Commons