This paper reports one-period and two-period bargaining experiments involving asymmetric information. There is a single unit of a good to trade. In one-period experiments the seller makes one take-it-or-leave-it offer to a single buyer. In two-period experiments, with a publicly known positive probability the seller can make a second price offer if his first offer is rejected. The seller’s cost is public information. The buyer’s value is private; the seller knows only the distribution from which it is drawn. This multi-period setting is sometimes referred to as the bargaining version of the durable goods monopoly problem. In the marketing literature, this setting has been referred to as the Tunisian Bazaar.

The key treatment variable is the probability that the game will have a second period. The perfect Bayesian equilibrium predicts that the initial price offer declines as the probability of a second period increases. The experiment includes four different continuation probabilities to test this equilibrium prediction. Prior experiments by Reynolds and by Cason and Sharma suggest that the equilibrium predictions may fail in some respects. We also consider quantal response equilibrium (QRE) to capture bounded rationality of subjects. We estimate the decision error parameter that provides the best fit to the experimental data.

Co-author Tim Cason