Dissertation

Global Value Chains and Trade Policy Making in the 21st Century

Paper 1. Global production sharing and the new demands for deep preferential trade agreements [pdf][supplementary information]

Summary: Modern trade agreements include a diverse and ever-growing set of policy commitments that are seemingly unrelated to trade. Why do contemporary trade agreements invest so heavily in managing such a complex and disparate set of non-trade policy areas? I focus on the rise of global production networks where firms expanding their productions globally face inflated costs of disruption, which invites governments’ opportunistic behaviors with diverse regulatory policies. Therefore, I argue that firms integrated to GVCs will demand greater institutional governance over consistent use of regulatory policies in trade agreements. This, however, constrains governments’ use of domestic policies in aid for firms operating independently from GVCs, pitting them against GVC-integrated firms over the formation and the design of trade agreements. In this paper, I propose a comprehensive theory that unifies the literature on PTA design, which has offered piecemeal explanations for individual trade agreement chapters. To test the theory, I employ a multi-regional input-output table in Eora project to compute value-added contributions for each type of firm. The paper makes three findings. First, countries with a greater share of domestic-oriented value-added are less likely to enter any type of PTAs. Second, countries with a greater share of value-added in final goods trade favor market-access provisions in PTAs but oppose regulatory provisions in PTAs. Finally, I find that countries with a greater share of GVC-integrated value-added have a greater number of regulatory policy provisions in their PTAs. These findings highlight the need to reexamine the conventional cleavages between import-competing and export-oriented in the face of deepening global production networks and the growing complexity of PTA designs.

Paper 2. Mapping the global division of labor: dynamic clustering analysis on task specialization in global value chains

Summary: Understanding the structure of the global economy is important as it shapes and constrains countries’ political and economic decisions. Existing studies have utilized varying levels of trade data to offer a systematic summary of the global economy and where countries are positioned in it. However, a growing number of studies point out that the deepening global production networks and internationalization of production processes mandate a new perspective on the collection and analysis of trade data. Specifically, they emphasize the need to account for global value chains, the full range of production activities by global economic actors to bring a product from its conception to its end use and beyond. In this paper, I employ global value chains as the main unit of analysis to advance our understanding of the structure of the global economy. Specifically, I analyze the “role” that countries play in global production networks. Employing MRIO table in Eora project and Wang et al. (2017)’s measures on GVC, I construct a country-level vectors of GVC position and GVC participation. I then fit a Gaussian mixture Hidden Markov model to cluster countries’ roles in GVC. The penalized log-likelihood suggests that 6 clusters offer a succinct explanation of the different roles that countries play in the global economy. Specifically, countries in the global production network specialize in the production of primary inputs, dissemination of high-tech inputs and know-how, final assembly, financial services, and supply of energy supplies such as petroleum. The paper then explores how such specializations shape countries’ political behavior, with a focus on international economic sanctions and trade disputes.

Paper 3. Spiders and snakes: the organizational structure of global production

Summary: How do we explain a country’s preference for different modes of trade liberalization? Despite the proliferation of preferential trade agreements (PTAs), some countries opt for multilateral channels of tariff reductions while others competitively enter into minilateral or bilateral agreements with deeper policy commitments. Existing studies have primarily focused on a country’s domestic interest groups and political institutions to explain preferences for trade liberalization. This paper highlights the importance of the topological structure of a country’s production network as the new factor of preference for trade liberalization. Specifically, I highlight the new typology for the structure of a country’s production network - spider and snake. Spider refers to countries whose value chains end in one-step transactions with other countries. Snake structure denotes a multi-step production sequence where intermediate inputs travel sequentially through multiple other countries. I theorize that spider countries put more weight on extensive margins or horizontal expansion of multiple trade relations while snake countries focus more on intensive margins or vertical deepening of trade relations with a few trade partners. These differences incentivize spider countries to prefer multilateral engagement and snake countries to opt more for bilateral forms of trade liberalization. To test the theory, I propose a Bayesian latent variable model to uncover core value chain structures at the country level. Then I use allometric scaling to obtain a one-dimensional summary score that gauges the extent to which a country’s production network is spider or snake.