Essays on international trade and econometrics.

Abstract

The first chapter studies the GATT/WTO dispute settlement mechanism and develops a unified theoretical model to explain the stylized facts observed across different decades of the GATT/WTO regimes. This study first explores the role of the degree of legal controversy over a panel ruling in determining countries' incentives to block/appeal a panel report under the GATT/WTO regime. The model is able to explain the surge in blocking incidence during the 1980s over the preceding GATT years and the immense frequency at which the new appellate procedure under the WTO is invoked. Furthermore, a two-sided asymmetric information framework is used to study the effects of political power on countries' incentives to use, and interactions in using, the GATT/WTO dispute settlement mechanism. The result, when confronted with the statistics on disputes in different decades of the GATT regime, provides us an indicator of how well the dispute procedure has worked during various decades, in terms of how much this procedure has been subject to potential power politics. The second chapter investigates the political equilibrium of trade policy when economic structure is characterized by differentiated products and increasing returns to scale and there exists intra-industry trade. This is accomplished by embedding the Krugman (1980) model into the Ricardo-Viper specific-factors model and employing the political contribution framework of Grossman and Helpman (1994) to derive the endogenous tariff equilibrium. The result shows that endogenous tariffs are positive for all industries with non-negligible shares of world production. However, the level of protection is less than the optimal tariff that would otherwise be imposed by a benevolent government in an unorganized industry, and higher in an organized industry. The protection provided to all unorganized (organized) industries increases (falls) with the relative weight the government attaches to aggregate welfare vis-a-vis campaign contributions and falls with the fraction of the population that belongs to a lobby group. The model also indicates that the endogenous tariff level in an organized industry might be explosive. The higher is the fraction of the population represented by a lobby and the higher is the weight on aggregate welfare in the government's objective function, the smaller is the possibility for such an explosive tariff. The third chapter proposes an robust impulse response estimator. The impulse responses implied by an estimated autoregressive (AR) model are often used to estimate the true impulse responses. This conventional method could mislead us, however, if the AR model is misspecified. The proposed estimator is less affected by such misspecification. It has the regular asymptotic properties of parametric estimators. The Monte Carlo simulations demonstrate that (i) it is capable of capturing a sudden drop of the true impulse response, unlike the conventional estimator, and that (ii) its performance is comparable to that of the conventional estimator even under correct specification.