| The Canada-US border has a trade reducing effect. Several studies estimate that this border is surprisingly large, in spite of Canada and the US being similar in language, culture and legal system. My thesis offers a simple, yet appealing, explanation.; The reasoning is as follows: Canada and the US are similarly endowed with resources in the West and manufactures in the East. Comparative advantage yields a predominantly westward flow of intermediate goods, and eastward flow of manufactured goods. Since goods produced in both countries are arguably close substitutes, even small border costs can create large border effects by keeping this trade on its own side of the border.; To support this hypothesis, I present a model in Chapter 1 that has a role for intermediate goods, substitutability of goods and small border costs. Using numerical methods I show how the gravity model consistently estimates a border effect because it is essentially misspecified. I also present empirical evidence to support two salient aspects of the model: first, that differences in regions' types significantly increases bilateral trade volumes (i.e. comparative advantage is creating trade); and second, that border costs may be as small as transport costs. Based on evidence presented in the second chapter, I claim that these transport costs are no larger than 12%.; In the second chapter, I use numerical simulation of a general equilibrium model to decompose the Canada-US border effect into several causal factors. These factors include tariffs, border costs, endowments, technologies and preferences. The degree to which each are significant depends on the elasticity of substitution (sigma). If goods are not considered close substitutes (sigma = 1.1), then primitive preferences explain most of the observed trade pattern. Alternatively, if Canadian and US goods are close substitutes (sigma = 20), there is an alternative explanation. Preferences still play a key role, but tariffs alone can account for up to 40% of the border effect. When tariffs are combined with a 10% border cost, the border effect is entirely explained.; The third chapter provides an explanation for the 40% decrease in the Canada-US border effect between 1990 and 1996. Using a general equilibrium model to simulate Canada-US trade patterns, I use a new technique of calibrating the model to two data points. By restricting the possibilities of parameter changes between the two years, I decompose the decreasing border effect into various causal elements. I find that when goods are close substitutes, the reduction in tariffs can explain the entire decrease in the border effect. If goods are not close substitutes, then primitive preferences must dramatically change to support the new trade pattern. I claim that the former case is more acceptable.; What made the large border effects so surprising is that Canada and the US are similar, thereby offering few restrictions to trade. What has been overlooked, until now, is that this similarity produces substitutable goods which are highly sensitive to price. In this context, I conclude that the large Canada-US border effects are not that surprising after all. |
