Trade Potential Revisited: A Panel Data Analysis for Zimbabwe

Abstract

This paper notes that previous results on trade potential based on a panel data set may be biased and proposes the adequate Poisson Pseudo Maximum Likelihood method to estimate trade potential based on the elasticity estimates generated by a gravity model, with conclusions on trade potential based on confidence intervals estimated with the Delta method. This approach had not been yet considered in the literature for panel data. This methodology is used to evaluate Zimbabwe export potential in a period characterized by strong restrictions on trade, based on the elasticity estimates generated by an augmented gravity model for six Southern African Development Community member countries and their exports to the rest of the world. Results show that Zimbabwe has a large unexploited trade potential which will not be realized without political stability and structural reforms. For comparison purposes, we also present the gravity coefficients calculated with other estimation methods.

Keywords: gravity model, export potential, Poisson Pseudo-Maximum Likelihood estimator, panel data, confidence intervals, Delta method, Zimbabwe, Southern Africa Development Community

JEL Classification: F14, F15, F16