NOT ALL RIVALS LOOK ALIKE: ESTIMATING AN EQUILIBRIUM MODEL OF THE RELEASE DATE TIMING GAME

I develop a new empirical model for discrete games and apply it to study the release date timing game played by distributors of movies. The results suggest that release dates of movies are too clustered around big holiday weekends and that box office revenues would increase if distributors shifted some holiday releases by one or two weeks. The proposed game structure could be applied more broadly to situations where competition is on dimensions other than price. It relies on sequential moves with asymmetric information, making the model particularly attractive for studying (common) situations where player asymmetries are important. (JEL C13, C51, L13, L15, L82).