Rules and discretion in trade policy

We argue in this paper that the second-best nature of trade-policy intervention makes it likely that the issue of time consistency viii be an important consideration in determining both the extent and the efficacy of such intervention in most environments. The point is seen most directly by noting that a tariff is both a tax on consumers and a subsidy to producers of the import-competing good. Since first-best intervention typically calls for targeting each distortion with a separate tax/subsidy, the tariff will be a more effective policy tool if its consumption tax aspect can be separated from its production subsidy dimension. Consequently, if production decisions are made prior to consumption decisions, a government with sufficient policy flexibility will be tempted to surprise producers with policies other than those announced in an effort to make this separation. This leads optimal trade policy intervention to be time-inconsistent in a wide range of environments. We explore this idea in general terms and illustrate the results with specific examples.(This abstract was borrowed from another version of this item.)