Collusion in spatial competition markets

Abstract In this paper a Hotelling model with quadratic costs, as modeled by Chang (1991), is altered for different assumptions in order to determine to consequences for the stability of collusion. In particular, the model is altered by introducing endogenous locations and a capacity constraint. For each modification, the new outcome is computed. A critical discount factor for each of the models is determined by computing the profit. In the composed models the outcome is analyzed for two different assumptions of the reactiveness of the rival firm to deviation. The results show that collusion is less stable, for markets where the reactiveness to deviation of the rival firm is low, when locations are endogenous to the model. Furthermore, it is shown that a capacity constraint only diminishes the deviation profit. This implies that cooperative practices are more likely to be found in spatial competition markets where firms are constrained by their capacity. In markets with endogenous capacity choice, the location bundle (1/4,3/4) signals collusion. As the outcome is not robust to changes in assumptions, a case-by-case analysis is suggested.