GRAVITY WITH GRAVITAS: Effect of the border

GRAVITY WITH GRAVITAS: Effect of the border Border — Canada is the exponential of the Canadian dummy variable coefficient, a5, which gives us the effect of the border on the ratio of inter-provincial trade to state-province trade after controlling for distance and size. Similarly, Border — US is the exponential of the coefficient on the US dummy variable, which gives the effect of the border on the ratio of inter-state trade to state-province trade after controlling for distance and size.
Four conclusions can be reached from the table. First, we confirm a very large border coefficient for Canada. The first column shows that, after controlling for distance and size, inter-provincial trade is 16.4 times state-province trade. This is only somewhat lower than the border effect of 22 that McCallum estimated based on 1988 data. Second, the U.S. border coefficient is much smaller. The second column tells us that inter-state trade is a factor 1.50 times state-province trade after controlling for distance and size. We will show below that this large difference between the Canadian and U.S. border coefficients is exactly what the theory predicts. Third, these border coefficients are very similar when pooling all the data. Finally, the border coefficients are also similar when unitary income coefficients are imposed. With pooled data and unitary income coefficients (last column), the Canadian border coefficient is 14.2 and the U.S. border coefficient is 1.62. add comment
The bottom of the table reports results when atheoretical remoteness variables are added. We use the definition of remoteness that has been commonly used in the literature following McCallum’s paper. The regression then becomes
w8079-2
where the remoteness of region i is
w8079-3
This variable is intended to reflect the average distance of region i from all trading partners other than j. Table 1 shows that adding remoteness indices for both regions changes the border coefficient estimates very little and also has very little additional explanatory power based on the adjusted R2.