Essays in the political economy of international trade and economic growth

This dissertation has four chapters. The first chapter argues that malapportioned legislatures, such as the U.S. Senate, can lead to constituencies with different sizes. As a result, industries that are located in smaller constituencies are likely to receive greater protection from international trade. This chapter combines a model of legislative bargaining and a model of lobbying to study trade protection while allowing for a legislature with multiple legislators and differently sized constituencies. We then test the predictions of this model using tariff votes from the U.S. Senate in the late 19th and early 20th centuries and a panel of tariffs and non-tariff barriers to trade in the U.S. in the 1990s.; The second chapter (coauthored with John Hatfield) studies how trade protection varies with the mechanism used to elect the legislature. We study how trade policy differs between countries with legislatures elected by a plurality election rule m single member constituencies and legislatures elected by a proportional rule. We find that proportional systems have more trade protection than majoritarian systems. We find empirically that parliamentary countries with proportional electoral rules have much more trade protection than those with majoritarian electoral rules.; The third chapter looks at the influence of industry concentration, industry location across U.S. states and the political ideologies of U.S. states and institutions of the U.S. Senate on tariffs and non-tariff barriers to trade. This empirical study gathers a wide-ranging dataset on U.S. industries and governmental actors during the 1990s and compares which of these factors are the most important for understanding international trade policy in the U.S.; Using Monte Carlo simulations, the fourth chapter (coauthored with Romain Wacziarg) evaluates the bias properties of common estimators used in growth regressions derived from the Solow model. We allow for measurement error in the right-hand side variables, as well as country-specific effects that are correlated with the regressors. Our results suggest that an OLS estimator applied to a single cross-section of variables averaged over time performs best in terms of the extent of bias on each of the estimated coefficients when compared to a fixed-effects estimator and an Arellano-Bond estimator.