Stackelberg beats Cournot: On collusion and efficiency in experimental markets

We report on an experiment designed to compare Stackelberg and Cournot duopoly markets with quantity competition. For each market we implement both a random matching and fixed-pairs version. Stackelberg markets yield, regardless of the matching scheme, higher outputs than Cournot markets. Under random matching total quantities in Stackelberg markets are higher than predicted by the subgame perfect equilibrium while total quantities in the Cournot markets match the equilibrium prediction. Under fixed pairs aggregate output for both, Cournot and Stackelberg, is lower than under random matching, but there is less collusion on Stackelberg markets and, hence, higher efficiency.