Learning by doing

This paper examines the implications for strategic trade policy of different assumptions about precommitment. In a dynamic oligopoly game with learning by doing, the optimal first-period subsidy is lower if firms cannot precommit to future output than if they can; and is lower still if the government cannot precommit to future subsidies. In the linear case the optimal subsidy is increasing in the rate of learning with precommitment but decreasing in it if the government cannot precommit. The infant-industry argument is thus reversed in the absence of precommitment which has important implications for economic policy in dynamic environments.