This dissertation consists of three essays. The first essay examines the effects of tax differentials on the trade balance across countries. Given that intrafirm trade accounts for the sizable share of the world's international trade, it is expected that income-shifting activities of multinational firms can bias the trade balance in many countries. Specifically, an increase in the relative tax liability in one country is expected to decrease the trade balance of that country. Using proxies to the effective tax liability of 19 OECD countries, the cointegrating regressions show significantly negative relationships between tax differentials and the trade balance among relatively small industrial countries.
The second essay asks whether the empirically observed home biases in international trade are accounted for by a theoretical model. It has been pointed out that trade among individual Canadian provinces is much larger than the trade between individual Canadian provinces and individual U.S. states. There is a similar tendency in the trade among the OECD member countries. Obstfeld and Rogoff (2000) claim that such a bias can be explained if one takes into account the interaction between transaction costs and the elasticity of substitution. This study tests their claim using a dynamic general equilibrium model where agents pay proportional transaction costs. The simulation results show that the bias levels generated by the plausible values for transaction cost and elasticity are not particularly inconsistent with the observed levels in the US -- Canada relationship.
The third essay tests a version of international real business cycle model aimed at examining the effect on the exchange-rate volatility of market segmentation generated by a trade friction across countries. Obstfeld and Rogoff (2000) argue that segmentation in international goods market can explain the empirically observed real exchange-rate volatility. In this study, a trade cost in goods market combined with income heterogeneity of consumers endogenously generates market segmentation by preventing a fraction of consumers from participating in international trade. Under such a circumstance, the volatility of exchange rate actually rises, but the volatility is still below the observed reality, suggesting that trade cost alone cannot explain the anomalous exchange-rate behaviors.
