Trade and Integration Policy Analyses for Mercosur: Methodologies of the CGE Modeling with Economies of Scale and Imperfect Competition

This paper addresses the approaches and methodologies used to construct the CGE model to evaluate trade and integration options for Mercosur. The model is a multi-region, multi-sector, comparative static CGE model with 25 sectors and 10 regions, benchmarked in 2001. We introduced several features in both database and modeling, which incorporates some elements of new trade theory beyond the standard neoclassical paradigm. First, in modeling scale economies, we follow the tradition of Smith and Venables (1988), Gasiorek, Smith and Venables (1990, 1992) and particularly Flores (1997). Instead of employing variable and fixed cost structures as in the majority of the models with scale economies, our model applies polynomial cost functional form, as presented in Flores. Second, the model assumes that firms in non-competitive sectors compete under the Cournot-Nash oligopolistic behavior. This enables us to analyze strategic interaction among firms at home and with foreign competitors. Third, the model is built on the new trade database, and the comprehensive hemispheric tariff database constructed from the FTAA database and the inclusion of numerous ALADI agreements. It incorporates all relevant trade agreements and preferential treatments in place in the Western Hemisphere with estimations of ad valorem equivalents of specific, compound tariffs plus TRQs. Finally, the market concentration for noncompetitive industries in Mercosur is taken from the recent study, examining manufacturing competitiveness for large countries in Latin America, using the Herfindahl index of concentration in Brazil.