Hicks (1932) introduced the idea that the outcome of successful wage negotiations depends on the costs that would have been incurred had a strike taken place. Hick’s profound insight attracted attempts by many authors to construct formal bargaining models of this type. However these attempts, the most notable of which were by Bishop (1964) and Foldes (1964), must be considered to have failed, primarily because no adequate general bargaining theory was then available. The model pioneered by Ståhl (1972) and developed by Rubinstein (1982) using Selten’s (1975) concept of perfect equilibrium has now filled this lacuna. The present paper conbines this Ståhl-Rubinstein approach with the special characteristics of union-management negotiations to provide a simple yet far reaching account of how wage settlements are determined.

Two aspects of union-management negotiations distinguish them from general bargaining. In the latter it is usually assumed that the costs of delay are constant and that failure to agree causes the parties to forgo what would have been their share in the surplus. However in the former, the costs borne by a party rise steeply as it runs out of resources and, when agreement is not reached, the winner of the strike receives the entire surplus. The model developed below shows how, when there is no strike, costs and ability to win jointly determine the wage settlement. The paper closes with a discussion of the broader implications of the cost rise phenomenon.