Exclusionary Bundling and the Effects of a Competitive Fringe

Abstract

In its original version, the paper by Caliskan et al [2007] did not offer rigorous hypotheses. My comment aimed at filling the lacuna with respect to the most innovative part of the paper, the introduction of a fringe competitor in the tying market. The published version now has an equilibrium in mixed strategies. This comment offers an alternative view. It assumes a sequential structure of the game: First the fringe competitor chooses a price. The dominant provider responds to this move. Given this assumption, there is an equilibrium in pure strategies. The assumption makes it easier to see the implications of fringe competition in the tying market on competition in the tied market. In a duopoly, the capacity of one provider is in excess. The fringe competitor / is unable to extend it. The dominant provider d is able to serve the entire market. The incentive effects are first shown graphically, then by calculus. Marginal cost c, and hence supply s, are constant. Given the assumption about the sequence of moves, the large provider just tolerates that demand F is lost to the fringe competitor. F measures this competitor's capacity. Under this assumption, the total demand d is no longer relevant for the dominant provider. For it, the firmspecific demand dd matters. Due to the limit on the fringe competitor's capacity, this demand is safe. The dominant firm behaves as if it held a monopoly in this smaller market. It sets a price pd. The resulting quantity is qd. If the fringe competitor starts from correct beliefs, it charges the same price pd, and sells quantity qd+f qd. Note that the community of providers can do better. If they collude, they both set price p*, and they jointly sell quantity g*. As the graph shows, this leads to a higher total profit. Formally, one has the following: In the absence of the fringe competitor, the demand is given by