Collusion, Discounting and Dynamic Games
研究动态博弈中帕累托最优结果是否只由耐心玩家维持，以及均衡集是否随贴现因子扩大。结果表明，在动态博弈中耐心与合谋可持续性无直接关联，甚至存在接近重复博弈的动态博弈中，不耐心反而促进合谋。
This paper investigates the following claims: in a dynamic game, (i) Pareto-optimal outcomes are sustainable only by patient players and (ii) the set of equilibrium outcomes expands in the discount factor. The intuition for these results, although immediate for purely repeated games, is shown to be incomplete for dynamic games. Consequently, there is no direct link between patience and the sustainability of Pareto-optimal outcomes. In fact, there are dynamic games arbitrarily "close" to a repeated game in which such collusion is facilitated by impatience. However, for a class of dynamic games, any equilibrium outcome that is sustainable by less patient players is also an equilibrium outcome when players are more patient. Journal of Economic Literature Classification Numbers: C73, D90.