Do National Borders Really Matter? Canada–US Regional Trade Reconsidered

This paper reconsiders the Canada–US border’s effect on trade. The authors first test whether the findings of McCallum (1995) and Helliwell (1996)—that the border substantially decreases trade—change when better data are used. It is found that the “border effect” may be substantially less than previously measured—up to 50% smaller—but remains surprisingly large. An explanation of the border’s effect is sought. Transportation equipment offers a natural experiment, as North American trade has been completely liberalized for several decades. A higher border effect is found for these freely traded goods, which rules out standard protection as the border effect’s cause.