The Trade Effects of Tariff Preferences

Abstract

In recent years the international community has heard a great deal of discussion of proposals for tariff preferences, to be extended by all developed countries to all less-developed countries (LDC's). It is obviously very important for policy purposes in both sets of countries to have some idea what the trade effects of such a preference scheme would be. The problem, however, poses a very difficult task for empirical economics, for the following reasons: There are, first, all the problems encountered in estimating the static effects of most-favorednation (MFN) tariff reductions [4; 5]. Second, since the tariff cut is preferential, we have to be concerned not merely with the displacement of domestic production by imports, but also with the substitution of one group of imports for another. Even less is known about elasticities of substitution between different classes of imports than about import demand elasticities. Third, export supply elasticities are probably lower for less developed countries than for developed countries. The trade effects of tariff cuts become more sensitive to the precise value of supply elasticities as these elasticities become smaller. Investigators studying the effects of MFN tariff cuts on trade in manufactured goods have frequently assumed (long-run) supply curves to be perfectly elastic, an assumption which causes little error if supply elasticities are high relative to demand elasticities. Such an assumption would be dangerous in the case of LDC exports. Finally, there are all the limitations of a static model, with its