清水 大昌

We investigate the flexible manufacturing system (FMS) by using the spatial price discrimination framework of Pal. The degree of flexibility when producing variants can be increased by increasing the investment or production cost and is endogenously determined. Our result is that private incentives for increasing flexibility are excessive from a social welfare perspective.

We investigate in a spatial Cournot competition setting how much product research and development (R&D) investment firms make in order to differentiate their goods. The model features a location competition in the first stage, R&D competition in the second stage, and Cournot competition in the third stage. There are two differentiation methods in the model-spatial differentiation affecting the firm costs, and differentiation in the demand function affecting the demand levels. We find that comparing to the setting with no spatial connotation, the firms tend to invest less in product R&D. We also consider welfare implications. Although the firms employ maximum differentiation in equilibrium, there are cases in which welfare can be improved if the firms are to locate closer to each other.

We investigate equilibrium locations of a spatial Cournot competition model that includes both circular city and linear city models as special cases. We find that the equilibrium location is discontinuous with respect to parameter beta (beta = 0 implies circular and beta = 1 implies linear), and multiple equilibria appear for some range of beta. The location pattern becomes exactly the same as that of the linear city for a wide range of beta, while it exactly matches that of circular city only when beta = 0. This implies the property of the linear city holds for non-measure zero case, while circular city does not.

We investigate a mixed economy where state-owned public enterprises compete against private firms. We examine sequential privatization of public enterprises, and find that under plausible assumptions one privatization increases the welfare gains of the subsequent privatizations. Thus, even if privatization does not improve welfare at the early stages, it can eventually lead to a point such that privatizations after that point are beneficial to the society and the privatization program ends up with a success.

We investigate how a leniency program, which is a policy that gives reduced penalties to colluding firms that cooperate with investigations of the Antitrust Authority, affects firms' collusive behavior. Using a model of quantity competition, we show that an amnesty to the second or later candidates of a leniency program is of no use if colluding firms can choose the most profitable collusion. Because a leniency program is a general rule that is applied to various markets in a country, our result implies that the design of a leniency program depends on which kinds of market structure are prevalent in the country.

We study sequential merger incentives under presence of product differentiation. Two sets of firms produce closely related goods, whereas each set produces more differentiated goods. Merger incentives under product differentiation are found to be stronger for two firms producing closely related goods than more differentiated goods. Also, after one merger, other firms are willing to follow with their own merger, resulting in sequential mergers. This result is consistent with the recent mergers in the video game software industry in Japan.

The mixed strategy equilibria in a location quantity duopoly model with circular markets are investigated. We find that a continuum of equilibria exists when the transport cost function is linear. However, if costs are strictly concave or convex, most strategies fail to qualify as equilibria. For any integer n, there are nondegenerate mixed strategy equilibria in which each firm locates at n possible locations with equal probability. This result explains possible outcomes in which firms' on-path locations are not minimally or maximally differentiated.

We investigate two shipping (delivered pricing) duopoly models with circular markets, Cournot and Bertrand. In the Cournot model, maximal distance between two firms becomes the unique equilibrium outcome if transport cost is increasing with distance, irrespective of whether the transport cost function is convex or concave. This result is robust in the sense that cost heterogeneity between two firms does not change the result. In contrast to the Cournot model, this location pattern can fail to be an equilibrium in the Bertrand model.

We investigated welfare implications in location-quantity models in a symmetric linear city. We found that when firms are not agglomerated in equilibrium, increasing the distance between firms raises (reduces) producer surplus and social welfare (consumer surplus). Moreover, central agglomeration is always optimal for consumers among symmetric locations, but not necessarily for producers. Central agglomeration can be inefficient even if it is the unique equilibrium outcome. In short, the firms are more likely to agglomerate or locate closer than what welfare maximizers would dictate, whereas they locate farther apart than what consumer surplus maximizers would recommend. (c) 2004 Elsevier B.V. All rights reserved.