Gravity in International Trade : Econometric Challenges and Environmental Extensions

Abstract

The dissertation contributes to the literature on the gravity model of international trade. The first part of the thesis deals with econometric challenges that arise in the estimation of gravity models. Specifically, the first chapter considers Poisson Pseudo-Maximum Likelihood (PPML) estimation of the intensive margin of trade, i.e. of how much two countries trade with one another. It offers a solution to the computational challenge posed by several sets of high-dimensional fixed effects in a non-linear panel data setting. It then applies the proposed iterative estimation procedure to re-assess the effect of a common currency on bilateral trade flows and does not find a significant effect. The second chapter considers the extensive margin of international trade, i.e. the question whether two countries trade with each other at all. It provides a model and econometric specification that allow for persistence due to both unobserved bilateral heterogeneity and true state dependence. An incidental parameter bias correction for the proposed dynamic high-dimensional fixed effects probit estimator is developed and the desirable properties of the bias-corrected estimator are demonstrated in Monte Carlo simulations. The estimator is then applied to the estimation of the extensive margin determinants. Taking into account the two sources of persistence is shown to matter considerably for the empirical results. Specifically, the chapter presents evidence that the extensive margin is characterized by a high level of persistence. The second part of the dissertation deals with the interplay of climate policy and international trade. In the third chapter, the trade, welfare, and emission effects of carbon tariffs are analysed. Carbon tariffs equalize carbon price differentials in order to reduce carbon leakage. For the analysis, an extended structural gravity model including a sectoral structure and a multi-factor production structure with emissions is developed that allows a decomposition of emission changes into scale, composition, and technique effects. The counterfactual analysis shows that carbon tariffs can reduce global emissions by mainly altering the sectoral composition, however at the cost of reduced trade flows and welfare losses for most, especially developing, countries. The last chapter of the dissertation deals with the consequences of unilateral withdrawals from the Paris Agreement. It considers not only the direct effect of a lower global sum of reduction targets, but also the indirect effect of the withdrawing country increasing its emissions beyond the business-as-usual emission path in response to the rest of the world’s climate policies. An extension of the model developed in chapter three is presented that incorporates both carbon leakage due to comparative advantage changes in international trade and via the international fossil fuel market. The counterfactual analysis reveals that a US withdrawal would harm the effectiveness of the Paris Agreement most strongly and eliminate one third of the global emission…