Learning from Strikes

This paper reports on an experimental study of the influence of asymmetric information and information spillovers on bargaining outcomes. It develops and tests Kuhn and Gu's model (1999) of learning in sequential wage negotiations, by means of two Ultimatum Bargaining Games with uncertainty on the proposer's side. Evidence shows that Dunlop's assertion of inflationary wage demands does not systematically hold and strike incidence is lowered by information spillovers, since demands are revised according to previous bargaining outcomes. However, in the presence of fairness concerns, the ability to observe outcomes but not the bargaining process does not entail a sufficient reduction in information asymmetry to a point of guaranteeing Pareto-improved bargaining outcomes.