Gross Product, Population Distribution and Heterogeneity of Border Effects in Gravity Models of Trade.

In this paper, I estimate a gravity model that provides strong evidence that border effects are heterogeneous by region and test for the presence of the spatial attenuation effect and international market access costs. I introduce GDP, importer population density, and importer metropolitan statistical area count as determinants of the border effect. I find the strong and robust result that border effects are decreasing in the GDP of the importing and exporting region: a 10% increase in importer GDP results in a 3.24% to 3.67% increase in international trade relative to domestic trade while a 10% increase in exporter GDP results in an increase in international trade relative to domestic trade by between 2.41% and 2.68%. These results are robust to a variety of specifications and indicate the existence of the spatial attenuation effect. However, this paper finds no indication of market access costs differing significantly when trade crosses the U.S.-Canada border. [ABSTRACT FROM AUTHOR]