Modeling, empirics and policy implications of firm heterogeneity in international trade

Zeynep Akgul, Purdue University

Abstract

Computable General Equilibrium (CGE) models are essential computational tools for trade policy analysis. While traditional CGE models based on the Armington assumption of national product differentiation have been successfully applied to various policy scenarios, they also have significant limitations in explaining the firm-level information prevalent in the recent international trade literature. The pioneering work of Melitz (2003) has provided a firm heterogeneity theory that can help address the shortcomings of Armington-based CGE models by introducing additional productivity mechanisms and extensive margin effects. Incorporation of firm heterogeneity in mainstream CGE models offers great potential to improve computational policy analysis. Even though there have been some efforts to incorporate firm heterogeneity into CGE modeling, a readily accessible Global Trade Analysis Project (GTAP) implementation is currently not available. This dissertation addresses this gap by a combination of theory, calibration, estimation and simulation to develop and implement a firm heterogeneity module executed within the GTAP environment. Chapter 2 presents the newly developed firm heterogeneity module with a stylized tariff removal scenario and compares the model predictions with those of monopolistic competition and perfect competition frameworks previously established in the standard GTAP model. Chapter 3 proposes a theoretically-consistent way to parameterize the firm heterogeneity module with a focus on the elasticity of substitution across varieties. Results show that the elasticity values that are consistent with the firm heterogeneity theory are considerably lower than Armington elasticities used in the standard GTAP model. Finally, Chapter 4 applies this newly developed module and parameterization to policy analysis in order to investigate the implications of reducing non-tariff measures associated with the beef hormone ban imposed by the European Union on imports from the United States based on the negotiations taking place for the Transatlantic Trade and Investment Partnership (TTIP) Agreement. The firm heterogeneity module predictions of welfare changes in the United States are distinctly different from those predicted by the standard GTAP model. This is explained by the endogenous productivity and variety effects implied by the firm heterogeneity theory. Results also suggest that the choice of policy instrument is an important factor in determining which one of these effects dominates in the final welfare outcome. This dissertation introduces the first implementation of firm heterogeneity into the standard GTAP model which I hope will serve as a powerful tool for policy analysis with improved abilities in tracing out the productivity changes and entry/exit of firms following trade liberalization episodes.

Degree

Ph.D.

Advisors

Hertel, Purdue University.

Subject Area

Economics|Agricultural economics

COinS