Essays on trade, FDI and development

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Devashish Mitra


Trade, Development, Foreign direct investment, Free trade agreements, Political economy

Subject Categories



The general theme of my dissertation is globalization and economic development. My first essay considers the distributional concerns of free trade agreement (FTA) formation. This can be studied by empirically testing two conflicting political economy models---the median voter model and the lobbying model, which predict pro-labor and pro-capital trade policy, respectively. Applying duration analyses, I find robust empirical support for the median voter theory of FTAs proposed by Levy [AER, 1997]. This paper provides a better understanding of the conflicting distributional forces when an unequal society votes for or against FTAs, and proposes a better empirical approach for bilateral FTA panel data analysis.

The second essay was motivated by Rose [AER, 2005], who finds no impact of the GATT/WTO on the world trade. This essay addresses this puzzle by investigating the sample selection bias and the gravity model specification. First, the GATT/WTO not only promotes trade between existing trading partners (intensive margin), but also creates new trading relationships (extensive margin). By restricting analysis to observations having positive trade flows, existing studies ignore the extensive margin. Second, the traditional specification of the gravity regression is inappropriate and also accounts for part of the failure in the previous literature to uncover the effectiveness of the GATT/WTO. Applying Poisson regressions to the gravity equation, this paper finds that both the intensive and extensive margins are important in evaluating the trade impact of the GATT/WTO.

My third essay was motivated by the hypothesis of race to the bottom in wages. This paper investigates whether foreign firms are attracted to lower wage provinces within China given that the overall wage level is already very low. Existing plant-level location choice studies provide almost no evidence for this proposition. Using a unique data on equity joint venture in China, we find that standard estimation suffers from omitted variable bias. Applying a two-step control function technique to conditional logit analysis, we find strong support for the attractiveness of low wages. This paper also provides new evidence on the attractiveness of low wages for different types of activities based on their skill and capital intensity.


Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.