Does the neighborhood matter? Three essays in international economics (Singapore, Japan)

Chapter one investigates whether available evidence supports the common hypothesis that the influence of world markets trumps the importance of domestic markets. Using input-output tables, I empirically assess the importance of domestic markets in Singapore—a small economy well integrated with the world economy and an economy generally not expected to exhibit strong domestic linkages. I show that domestic markets do substantially influence local specialization patterns in Singapore: approximately three-quarters of the change in downstream demand is directed toward purchases from domestic suppliers. The evidence also suggests that trade costs are more important in the market for differentiated intermediate inputs, while the persistent correlation in the market for a small sample of homogeneous final goods is partially explained by a high degree of substitutability between foreign and domestic varieties.; Chapter two investigates a related hypothesis that imports are goods substitutes for domestic goods. Using panel data and an earthquake to predict exogenous changes in output, I explore this hypothesis for Japan. The estimated equation for the import response to exogenous changes in output shows that in the short-run there is only a ¥2,000 increase in imports for every ¥100,000 exogenous decrease in output. In other words, imports are imperfect, if not poor substitutes for domestic production in Japan. This is consistent with results from the first chapter and confirms the notion that the world economy is far from integrated.; Chapter three enters the debate over sources of relative advantage. Specifically, this chapter empirically investigates the claims of economic geography. This theory states that access to markets and access to suppliers creates benefits to firms: location and, therefore, local markets are a source of relative advantage. Yet while an abundance of evidence confirms the existence of a strong correlation between location parameters and production patterns, these empirical results often fail to account for the fundamental endogeneity of the explanatory regressors. In this paper, I again use the exogenous variation generated by an earthquake in Japan to solve this problem. I show that, yes, economic geography is a source of relative advantage, but only in markets and across regions where trade costs are relatively high. Specifically, because the relative costs of trade in intermediates is larger than trade in final goods, firms benefit from proximity to one another, but do not gain from relative proximity to consumers; and because trade costs are relatively larger when goods travel across borders than when goods travel within borders, economic geography can explain international, but not intranational patterns of production.