|
on Game Theory |
| By: | Claude, D’ASPREMONT; Rodolphe, DOS SANTOS FERREIRA |
| Abstract: | In applying the common agency framework to the context of an oligopolistic industry, we want to go beyond the classical dichotomy between Cournot and Bertrand competition. We define two games, the oligopolistic game and the corresponding concept of oligopolistic equilibrium, and an associated auxiliary game that can be interpreted as a common agency game and that has the same set of equilibria. The parameterization of the set of (potential) equilibria in terms of competitive thoughness is derived from the first order conditions of the auxiliary game. The enforceability of monopolistic competition, of price and quantity competition, and of collusion is examined in this framework. We then describe the (reduced) set of equilibria one would obtain, first in the non-intrinsic case and then in the case where a global approach would be adopted instead of partial equilibrium approach. Finally, we illustrate the use of the concept of oligopolistic equilibrium and of the corresponding parameterization by referring to the standard case of symmetric quadratic utility. |
| Date: | 2005–02–17 |
| URL: | http://d.repec.org/n?u=RePEc:ctl:louvec:2005018&r=gth |
| By: | R.J., AUMANN; Jacques-Henri, DREZE |
| Abstract: | Modern game theory was born in 1928, when John von Neumann published his Minimax Theorem. This theorem ascribes to all two-person zero-sum games a value - what rational players may expect - and optimal strategies - how they should play to achieve that expectation. Seventy-seven years later, strategic game theory has not gotten beyond that initial point, insofar as the basic questions of value and optimal strategies are concerned. Equilibrium theories do not tell players how to play and what to expect; even when there is a unique Nash equilibrium, it is not at all clear that the players “should” play this equilibrium, nor that they should expect its payoff. Here, we return to square one : abandon all ideas of equilibrium and simply ask, how should rational players play, and what should they expect. We provide answers to both questions, for all n-persons games in strategic form. |
| URL: | http://d.repec.org/n?u=RePEc:ctl:louvec:2005021&r=gth |
| By: | Dequiedt, V. |
| Abstract: | In a first part, we provide a general approach to mechanism design subject to collusion. It is modeled as a Stackelberg game between the designer and a third-party which organizes collusion among the buyers. In this multi-principal context, the standard "Revelation principle" can be replaced by a "Collusion-proofness Principle" if, and only if, the collusion technology satisfies a transitivity condition. In a second part, we apply this approach to collusion in a private value auction where bidders' types are independent and study the optimal response of the seller to different threats of collusion between the buyers. We show that collusion in the optimal auction is efficient when the third-party can implement monetary transfers as well as when it can implement monetary transfers and reallocations of the goods. |
| Keywords: | COLLUSION PROOFNESS; MULTI PRINCIPAL; OPTIMAL CONTRACT |
| JEL: | D44 D82 L41 |
| Date: | 2004 |
| URL: | http://d.repec.org/n?u=RePEc:rea:gaelwp:200402&r=gth |
| By: | Pablo Brañas Garza (Department of Economic Theory and Economic History, University of Granada); Antonio Morales (Universidad de Málaga) |
| Abstract: | Recent papers on double-blind dictator games have obtained significant generous behavior when information regarding recipient is provided. But the lack of information disincentives other-regarding behavior and then, the subject’s behavior closely approximates the game theoretic prediction based on the selfishness assumption. This paper conducted four treatment of dictator games. We used one-room design, between-subjects anonymity and extra-credit point as rewards. Two treatments were used as baseline whereas the other two were aimed at reinforcing the recipient powerlessness and positive reciprocity. To promote these environments we include a “non—neutral” sentence to the instructions. Our baseline and modified DG are statistically di fferent from each other, indicating that the additional sentences promote other—regarding behaviour. In fact, pure-selfish behavior vanishes. |
| Keywords: | dictator game, framing e ffect, social issues, fairness, reciprocity. |
| JEL: | D63 D64 C91 |
| URL: | http://d.repec.org/n?u=RePEc:gra:wpaper:05/06&r=gth |