Time-consistent protection with learning by doing

Can a government induce efficiency gains in his domestic industry by protecting it against foreign competition? Would such trade protection be time consistent? The present paper builds a dynamic equilibrium model that accounts for learning-by-doing effects that link firms’ strategies over time. The model shows that the existence of dynamic economies of scale suffices to overcome the traditional government's lack of commitment of its tariff policy. This paper compares the infinite horizon Markov perfect equilibrium of this game with the dynamic equilibrium under commitment as well as the static Nash equilibrium. Equilibrium strategies are derived in closed form by solving a linear-quadratic differential game. Optimal trade policy involves higher tariff levels than in the static setup in order to account for future gains in efficiency. Under reasonable assumptions, the unique stable MPE is characterized by a domestic price and tariff that decrease as experience accumulates, thus supporting the future liberalization of trade as an equilibrium feature of this dynamic game.