Noncooperative selection of the core

Abstract

For bargaining environments given by transferable utility characteristic functions that are zero-normalized and admit a nonempty core, we find a class of random-proposer bargaining games, generalized from Okada (1993), such that there is a one-to-one mapping from these games to the core, each game realizes the corresponding core allocation as its unique (ex ante) Stationary Subgame Perfect Equilibrium (SSPE) payoff profile, and every ex post SSPE payoff profile converges to the core allocation as the discount factor goes to one. The result has a natural interpretation in terms of bargaining power.