Description:
This paper proposes a new equilibrium concept organizational equilibrium for models with state variables that have a time inconsistency problem. The key elements of this equilibrium concept are: (1) agents are allowed to ignore the history and restart the equilibrium; (2) agents can wait for future agents to start the equilibrium. We apply this equilibrium concept to a quasigeometric discounting growth model and to a problem of optimal dynamic fiscal policy. We find that the allocation gradually transits from that implied by its Markov perfect equilibrium towards that implied by the solution under commitment, but stopping short of the Ramsey outcome. The feature that the time inconsistency problem is resolved slowly over time rationalizes the notion that good will is valuable but has to be built gradually.