| We study a model in which heterogenous boundedly rational agents interact locally in order to play a coordination game. Agents differ in their mobility with mobile agents being able to relocate within a country. The model yields the following predictions: (1) mobile agents always benefit from increased mobility, (2) immobile agents benefit from increased mobility at low levels of mobility, (3) immobile agents lose from increased mobility at high levels of mobility, (4) there is an optimal ``country size,'' (5) ``income inequality'' is weakly increasing in mobility, (6) if there are arbitrarily small payoff differences between two countries, opening borders causes a ``brain drain'' effect; in the long run, all mobile agents reside in the favored (former) country and efficiency is attained only in that country. |