Reinterpreting Arbitration's Narcotic Effect: An Experimental Study of Learning in Repeated Bargaining

Abstract

Field evidence suggests that arbitration increases negotiation dispute rates. We study repeated bargaining in a laboratory to understand the reasons why. Our results represent a reinterpretation of an explanation known as the narcotic effect. The standard interpretation assumes that the probability of dispute without arbitration is constant across negotiations, but field evidence suggests that experienced bargainers have fewer disputes. To properly assess arbitration's impact, we compare bargainer learning with and without arbitration, under otherwise comparable laboratory conditions, and develop a model to measure learning. We find strong evidence that learning occurs in both cases, but is slower with arbitration.

JEL Classification: C78, C92, D77, J52