The Extent to Which Learning Rules Impact the Level of Cooperation in Spatial Price Competition

This paper investigates the extent to which the level of cooperation between spatially distributed firms differs with the use of alternative learning rules. The results are derived based on simulations of a structure conform to that presented by Waltman et al. (2013). Their model has been converted to Java in order to improve the efficiency of the model and simultaneously facilitate the comparison of various learning rules. Testing these rules in a homogeneous setting allows for comparability and this paper finds that imitation and reciprocity both lead to forms of cooperation, one on a global scale and the other on a local scale respectively. Furthermore, the results have proven to be robust to the influence of noise, experimentation and the presence of explicit cooperation in the form of a price agreement between two neighbouring firms. The level of information does however have an impact, often decreasing the level of cooperation between firms.