Trade, Technology, and Capital Flows
This dissertation consists of three essays about two interconnected aspects of globalization: global trade integration and global imbalances. The first two essays focus on containerization, a technology that played a pivotal role in twentieth century global trade integration. Containerization, an intermodal system of door-to-door shipping, replaced traditional breakbulk methods of shipping and enabled the swift movement of cargo across the world. In the first essay, I present evidence on the historical evolution of containerization. For that purpose, I constructed a comprehensive dataset with information on the timing and intensity of adoption of containerization for 147 countries. I find that adoption moves in an S-shaped pattern, while usage moves more slowly and linearly. Additionally, I identify four stages in the historical diffusion of containerization, based on my data analysis and on historical and anecdotal sources: innovation and early adoption (1956--1965), internationalization (1965--1974), worldwide adoption and intensification of use (1975--1983), and late adoption and growth of usage (1984--2008). These findings guide the construction of a theoretical framework for understanding the determinants of adoption and usage of containerization. The theoretical model focuses on firms' choices between two transportation technologies (containerization and breakbulk) and on the decision by the country's transportation sector to construct, maintain and operate a container port (adoption). Changes in fixed costs and network effects generate the patterns observed in the data.
In the second essay, I investigate the determinants of adoption and usage, using a two-step procedure which is derived from the theoretical model presented in the previous essay. The empirical results, which are consistent with the theoretical predictions, show that fixed costs and network effects are the main determinants of usage of containerization. Fixed costs affect containerized trade as a result of the spread of leasing companies and changes in the domestic transportation network. Network effects operate through network size, network usage, and network income. With regards to adoption, my results show that expected future usage of containerization, institutions, a country's size in terms of trade and geographical area, and trade with Australia and the United Kingdom are the main determinants. Trade with the United States, surprisingly, has no effect.
In the third essay, co-authored with Barry Eichengreen, I turn to the other aspect of globalization, global imbalances. First, we extend the two-country model presented in Obstfeld and Rogoff (2007). and investigate the exchange-rate implications of several rebalancing scenarios. We find that it matters tremendously how many countries are on the other side of the US current account adjustment; whether surplus countries like China, as they continue to grow, concentrate their investment in productive capacity in traded or nontraded goods; and the nature of structural reforms that change spending patterns in the US and abroad. Second, we investigate the likelihood of a sustained reduction in global imbalances. While previous literature has provided evidence on the circumstances under which large deficits come to an end, we add evidence on how chronic current account surpluses are also eliminated. We find that large current account surpluses tend to be eliminated when they have been allowed to rise previously to exceptionally high levels, when the economy doing the reducing is less open (smaller political resistance to resource allocation), when an earlier period of rapid growth comes to an end (presumably both moderating the rate of growth of the capacity to produce tradable goods and rebalancing demand toward domestic goods), after reductions in budget surpluses, and in the case of oil-exporting economies, when oil prices are unusually low.