Dynamic Coalition Formation from a Bargaining Perspective

Abstract:     
Extending Konishi and Ray (2003), we study coalitional bargaining in "real time", a situation in which coalition formation is interwined with the ongoing receipt of payoffs. A particular focus is placed on the case in which agreements are permanently binding, so that they can be changed only with the consent of existing signatories. For characteristic function games our results may be viewed as extensions of Seidmann and Winter (1998) and Okada (2000): we prove that equilibrium processes must converge to efficient absorbing states and that dynamic efficiency must obtain from at least one initial condition. In contrast, none of these properties continue to hold when we move to games with widespread externalities. The most striking illustration of this is the existence of games in which all equilibria are inefficient from every initial state, despite the ability to write binding agreements. The results for games with externalities complement those obtained by Gomes and Jehiel (2003), who suppose that individuals are able to make upfront one-time transfers, a feature that is ruled out by assumption here.

Number of Pages in PDF File: 35

Keywords: Coalitional bargaining, inefficiency

JEL Classification: C72, C73

working papers series