Abstract:This paper quantifies the welfare impact of unilateral trade liberalization and computes the optimal tariff structure for Costa Rica in the presence of trade-policy-induced international capital flows and foreign capital taxation. For this, an applied general equilibrium model integrating trade, capital flows and international capital income taxation is used. The model has been calibrated to a 1990-91 data set for the economies of Costa Rica and a group of OECD countries. In the model, foreign capital income is taxed by host countries and the tax-credit system operates in foreign investors home countries. Results for Costa Rica show that complete trade liberalization ends up being welfare-reducing, as it leads to an outflow of capital and loss of tax revenue which more than offset the efficiency gains from an enhanced resource allocation.The optimal tariff structure for the Costa Rican economy turns out to be a mixture of import tariffs and subsidies, though of a relatively small level.
Non-Technical SummaryIt is well-known that, in the absence of foreign capital taxation, international capital inflows Rica's optimal tariff structure under this setting. The model used disaggregates economic activity into 10 sectors, and is calibrated to a 1990-91 data set on production, trade, consumption, factor use, and capital income taxation both for Costa Rica and a group of OECD countries. The OECD countries included in this group are those which, as of 1990-91, were using the tax-credit mechanism described above, and accounted for roughly 80% of Costa Rica's foreign direct investment.Our results show that, with foreign capital taxation by Costa Rica and the tax credit mechanism in force in the OECD countries included in the model, complete elimination of import-tariffs by Costa Rica would reduce its welfare. The optimal tariff structure for this economy turns out to consist of a combination of relatively low import tariffs and subsidies. Though literally interpreted these results seem to provide a case against free trade, we qualify them by noting that the nature of the model used does not capture neither dynamic gains associated with trade nor other static cost usually linked with non-uniform tariff structures (such as rent-seeking-related costs).