Demystifying modelling methods for trade policy

In recent years, quantitative analysis of the effects of policies on economic outcomes has grown sharply. These exercises in quantification have been made possible by advances in theory and analytical techniques, and no less importantly, by the dramatically increased computational and data processing power of computers. This paper focuses on two classes of quantitative tools - computable general equilibrium (CGE) models and gravity models. These are perhaps the most commonly encountered quantitative analytical techniques in the area of trade. The primary purpose of this paper is to offer a non-technical explanation of CGE and gravity models to trade policymakers. We try to capture the essence of the analytical techniques, explaining the requirements of the models and computational procedures. We also seek to identify as clearly as possible the strengths and limitations of these analytical techniques. A second objective of the paper is to survey a range of studies based on CGE, particularly simulations of multilateral trade negotiations, and gravity models. The survey is useful in conveying a sense of how results can vary depending on what goes into the models by way of their structure and data, emphasizing the importance of judicious, critical interpretation. The main benefit of CGE models is that they offer a rigorous and theoretically consistent framework for analysing trade policy questions. The numbers that come out of the simulations should only be used to give a sense of the order of magnitude that a change in policy can mean for economic welfare or trade. Much more can be done to create confidence in the results. The simulations should benefit from more systematic and informative employment of sensitivity analysis. Ex-post validation of CGE models is needed to increase confidence in the numerical results. Correctly specified gravity models can illuminate questions that are important for trade policymakers. For example, what are the trade effects of WTO membership? How does entering a proposed preferential trade arrangement (PTA) affect a country's trade? How is non-members' trade affected? Does more trade lead to faster growth? Does trade improve the environment? Three important theoretical requirements that need to be taken into account in gravity models are highlighted in this study. First is the importance of relative distance and trade costs. Second is that liberalization, whether multilateral or regional, creates new trading relationships and not just increases the volume of existing trade. Third, trade is dynamic and this shows itself in new products and new firms that enter international commerce.