Using empirical Armington and demand elasticities in computable equilibrium models: An illustration with the CAPRI model

Computable equilibrium models play a significant role in ex-ante policy evaluations of free trade agreements. Previously, such models have generally used elasticities from the literature, although those estimates are mostly obsolete or else estimated using different aggregation levels to those used in the models. Furthermore, the same elasticities are used for all countries. These drawbacks highlight the importance of using empirically validated elasticities. In order to demonstrate the extent of the identified drawbacks, we use the Common Agricultural Policy Regionalized Impact (CAPRI) model and empirically assess the Armington and demand elasticities for Switzerland. We then compare the results of a free trade agreement between Switzerland and the European Union with the old and the estimated elasticities derived from the CAPRI model. The results differ remarkably from each other, which supports the importance of the estimated elasticities used in computable equilibrium models in terms of making policy assessments more reliable.