The Optimal Structure of Spectrum Charges Under Incentive Contract

This paper studies the optimal structure of spectrum charges under the information asymmetry between the government and firms, extending the model of Yoon and Lee (2005). In our model, the government assigns the spectrum to the firms in the first stage during which the demand is not realized yet and there is no information asymmetry between the government and the firms. The demand is realized in the second stage and it is known only to the firms - the firms have private information about the realized demand. The government designs the composition of spectrum charges to maximize its revenue from spectrum charges and offers contract menus to the firms in the first stage. The firms choose one of the contract menus after observing the realized demand at the second stage. First we consider the case where the spectrum is assigned to one monopoly firm. It turns out that the optimal structure of spectrum charges depends on the realized demand. The government designs the contract menus such that the monopoly firm chooses the menu under which the monopoly firm pays only fixed charges when high demand is realized, while the monopoly firm chooses the menu under which the monopoly firm pays both of royalty and fixed charges when low demand is realized. Royalty in our model is per-unit charges and total amount of royalty charges increases with output while fixed charges do not depend on output. The government's revenue under this type of incentive contract is larger than that under the regime where the government imposes the same structure of spectrum chargers regardless of realized demand. An additional theoretical model is provided in this paper for the purpose of analyzing the episode of the spectrum assignment of IMT-2000, where the spectrum was assigned to three firms, extending the model of Kang and Yoon (2006). In our model, three firms engage in Cournot competition after observing the realized demand at the second stage. The realized demand is common knowledge among the firms in the second stage. The government, however, does not observe it. We show that it is optimal to impose royalty with fixed charges not only when low demand is realized but also when high demand is realized, which is different from the case where spectrum is assigned to one monopoly firm. It is because output distortion resulting from royalty is mitigated when three firms compete