Regulating a monopolist with limited funds

Summary. We consider the problem of regulating a monopolist with unknown costs when the regulator has limited funds. The optimal regulatory mechanism satisfies four properties. The first property is bunching at the top, that is the more efficient types produce the same quantity irrespective of their costs. The second property is separability of less efficient types. The third property is full bunching of types when the available fund is small enough. The fourth property of the mechanism is that it is a third best one, that is, the output under this regulatory mechanism is strictly lower than the second best output for any given type.Received: 14 June 2004, Revised: 31 August 2004, JEL Classification Numbers: D82, H42, L51. Correspondence to: Manipushpak MitraThe authors would like to thank Georg Nöldeke for encouraging this work. The authors are also thankful to Debajyoti Chakrabarty, Sougata Poddar and one anonymous referee for helpful comments and important suggestions. The authors gratefully acknowledge the financial support from the Deutsche Forschungsgemeinschaft Graduiertenkolleg (DFG) 629 at the University of Bonn and from the Brussels Capital Region. The authors would like to thank the seminar participants at the EEA Conference 2003 (Stockholm), at the Indian Statistical Institute (New Delhi and Kolkata), at the Jadavpur University (Kolkata) and at the Centre for Studies in Social Sciences (Kolkata). The authors are solely responsible for the errors that may remain. An earlier version of this paper was called Financing infrastructure under budget constraints. Since the applicability of the problem developed in the earlier version is not limited to infrastucture, in this vers ion we have changed the name and content to incorporate this generality.     

Some rationalizability results for dynamic games

We study the relation between dynamical systems describing the equilibrium behavior in dynamic games and those resulting from (single‐player) dynamic optimization problems. More specifically, we derive conditions under which the dynamics generated by a model in one of these two classes can be rationalized by a model from the other class. We study this question under different assumptions about which fundamentals (e.g. technology, utility functions and time‐preference) should be preserved by the rationalization. One interesting result is that rationalizing the equilibrium dynamics of a symmetric dynamic game by a dynamic optimization problem that preserves the technology and the utility function requires a higher degree of impatience compared to that of the players in the game.