Divide and conquer: noisy communication in networks, power, and wealth distribution

Description: The political and economic evolution of countries such as Nigeria and Congo challenge the conclusions of Mancur Olson’s (1993, 2000) “stationary bandit” model. In this paper I argue for the necessity of an alternative model where the balance of political power and the distribution of income are endogenous. Such a model is developed in this paper as a game played between a ruler who has to decide the distribution of income and a group of agents/citizens who can communicate through a network. The citizens also have the opportunity to revolt if they are not happy with the distribution, but if too few of them revolt, they are defeated and get zero consumption. On the other hand, a successful revolt increases the consumption level of the rebels while the ruler gets nothing. Communication through the network is noisy, however, which could preclude the emergence of common knowledge and collective action among the citizens. The ruler could take advantage of this to make an unfair income distribution. The formalization of the game is accomplished using such concepts as p-beliefs and p-dominant strategy (Monderer and Samet ,1989, and Morris and Shin, 2002). We use this model to offer some reflections about how rent-seeking governments manage to survive.