Kecinski, Maik2015-09-302015-09-302015-05http://udspace.udel.edu/handle/19716/17055This paper reports on experimental results from a 2-period sequential Stackelberg game where players have a one time opportunity to invest positive relative pro ts in order to lower marginal cost and gain competitive advantage. Theory predicts one sub-game perfect Nash equilibrium in pure strategies with both Stackelberg leaders and Stackelberg followers playing absolute pro t maximizing strategies and leaders earning much larger pay than followers. Experimental results, however, show that Cournot play is modal. Stackelberg leaders learn to play fair through punishment. Investment to lower marginal cost was more frequent among Stackelberg followers, who used relative pro t maximizing strategies to punish rather than to gain competitive advantage.Repeated gamesDynamic gamesResearch Report