Gambling for the upper hand: Settlement negotiations in the lab

We exploit a controlled frameless laboratory experiment to study settlement negotiations and the plaintiff's decision to raise a lawsuit in case of an impasse. We find that greater variance in court outcomes increases the litigation rate and lowers the settlement rate. This latter finding goes against the received wisdom and earlier experimental evidence (Ashenfelter et al. 1992) that greater risk in arbitration outcomes increases the settlement rate. We find that self-serving biases about the protagonist' course of action are accountable for the lower settlement rate, while an impasse payoff inferior to that of the defendant induces the plaintiffs to excessive risk-taking in an attempt to narrow the gap.