Most products traded internationally are intermediate goods, which are often produced from natural resources. This dissertation analyzes how the simultaneous selection of optimal tariffs on intermediate and final goods is affected by two features of natural resources: their dynamic nature, which causes tariffs to change over time, and their uneven distribution across nations, which leads to tariff retaliation between importing and exporting nations.

Effects were first considered in a theoretical, general equilibrium model in which both an intermediate good and a final good made from it were imported by one region and exported by another. In the absence of retaliation, static optimal tariffs were positive on both goods in the importing region but ambiguously signed in the exporting region, and dynamic optimal tariffs on the two goods tended to move in opposite directions over time in both regions. In the static context, anticipated retaliation tended to decrease the size of the tariff on the intermediate good in both regions.

Next, a three-region simulation model of trade in Southeast Asian forest products was estimated econometrically and solved via nonlinear programming for optimal tariffs on logs and processed products. Tariffs were large and positive in nearly all instances and tended to reduce forest depletion, especially in the dynamic case. Welfare gains compared to free trade were moderate and were contingent on the redistribution of tariff revenue. Losses associated with pursuing static instead of dynamic optimization were relatively small. In the static case, tariff retaliation could lead to welfare losses for all regions.

By making trade policy endogenous, the simulation methodology offers a promising complement to traditional scenario analysis. Future research should concentrate on refining the methodology for application to trade models and on analyzing tariff retaliation in a dynamic context.