Sunday, September 2, 2007
Posted by D. Daniel Sokol
Keeping on the theme of collusion from previous posts, an interesting new working paper Reference Points and Self-Enforcing Collusion in Oligopoly is available from Felix Monuz-Garcia of the University of Pittsburgh Department of Economics.
ABSTRACT: This paper examines incentives for total and partial collusion in one-shot oligopoly games. Specifically, firm managers are considered to assign a positive importance to the difference between other firms' actual production choices and a particular reference level of output that, in any collusive agreement, would have restricted the market's total product. These considerations might be important, for instance, when the firm manager operates under weak pressures for short-run profits from the shareholders, whereas when this pressure is relatively high the above considerations are negligible. First, I show that by considering the importance that firm managers assign to this difference - reflecting the firm manager's preference for collusive partners - higher levels of collusion can be supported than in standard oligopoly games. In addition, if firm mangers' preference for collusive partners is high enough, the collusive agreement (e.g. cartel) becomes self-enforced in the Nash equilibrium of the static game. Furthermore, for moderate firms' preferences, this paper shows that the fully cooperative agreement can be sustained as the Nash equilibrium of the infinitely repeated game for less restrictive discount factors than in standard game-theoretic models of oligopoly.