Effects of Trade Costs and Capital Controls on Trade Imbalances

This dissertation consists of two essays on the determinants of global trade imbalances. In the first essay, I evaluate the effects of declining trade costs and capital controls on global imbalances using a model-based quantitative analysis. I develop a multi-country general equilibrium trade model in which trade imbalances are endogenously determined. Declines in trade costs and capital controls imply that fundamental shocks, such as productivity shocks, propagate more strongly to trade imbalances. I calibrate the model to 25 countries by exploiting data on bilateral trade flows, aggregate prices, net exports and measures of capital controls. I conduct counterfactual exercises where I fix trade costs or capital controls at the 1970's level. The results show that the decline in trade costs accounts for 42 percent of the trade imbalances that occurred between 1970 and 2007, while the decline in capital controls explains 22 percent of the imbalances. I also find the effects are heterogeneous across countries. Finally, my model suggests that welfare implications from lowering trade costs and capital controls are quite different. A reduction in trade costs leads to positive welfare gains for all countries, but a decrease in capital controls does not necessarily bring positive welfare gains. In the second essay, I address the empirical relationship between trade imbalances, trade costs and capital controls. In particular, the model suggested in the first essay predicts that lower trade costs and capital controls amplify the effects of productivity shocks on trade imbalances. I test this propagation mechanism by taking three empirical approaches; a fixed effects regression with panel data, a 2-country dynamic regression, and a 2-country vector autoregression (VAR). The results of the fixed effects regression show that trade imbalances respond negatively to productivity growth, and a decrease in capital controls makes this effect more negative. The model's implication for the propagation role of trade costs, however, is not supported by this approach. In the 2-country dynamic regression, trade costs and capital controls amplify the effects of productivity growth on trade imbalances in some countries, but not in others. Finally, the 2-country VAR(1) does not provide any evidence that is consistent with the model's prediction. In sum, there is mixed evidence on the propagation role of trade costs and capital controls.