The Asymmetric Effects of Reversible Tariff Changes under the United States GSP

Abstract

In a partial-equilibrium textbook model of competitive international trade, the volume and net value of imports always vary inversely with the tariff rate. This result depends on assumptions that the tariff changes are substantial, exogenous, permanent, and applied on a most-favorednation (MFN) basis. Empirical applications of this theory however, encounter different conditions. Generalequilibrium effects, such as changes in prices, exchange rates, and income, tend to be offsetting. Imperfect competition can make the incidence of the tariff indeterminate. Many tariff changes are small, endogenous, temporary, and/or discriminatory. To illustrate how these differences in tariffs can alter their observed relation to imports, we offer the following examples. (a) Tokyo Round cuts in U.S. tariffs were staged generally over eight years and averaged only 0.3 percent per annum [6] as compared to 6.4 percent average annual changes in American prices over the same 1980-87 period [18]. If search and recontracting costs caused buyers to ignore relatively small price variations, then many observed tariff changes would produce negligible effects. (b) Governments tend to lower tariffs when imports are low enough to obviate political problems, and they raise them as a response to surges. Observers who ignore this endogeneity might conclude that imports are positively related to tariff rates. (c) While many tariff concessions are bound under the GATT, some duties are temporarily altered to provide relief for domestic producers or preferential access for certain trade partners. Importers may not respond if the temporary benefits of doing so are exceeded by the costs of finding substitutes. (d) The importers' response to a tariff change will also differ between cases where some or all imports of a given product are affected. A discriminatory tariff change might shift the shares of suppliers while leaving total imports virtually unaltered. Empirical studies seldom test the relation between imports and tariffs directly. Time-series estimates of import demand equations usually subsume the tariff in a gross price variable.'