Are regional trading arrangements trade creating? An application of extreme bounds analysis

In this study, we use extreme bounds analysis to test the robustness of the hypothesis that regional trading arrangements (RTAs) are trade creating. Extreme bounds analysis provides a more rigorous test of specification uncertainty than traditional econometric theory by incorporating prior information and using a systematic approach to testing the fragility of coefficient estimates. The results show that the trade creation effect of most RTAs is fragile. Using a least squares estimator where all weight is attached to the sampling distribution, eight or more of the twelve RTAs considered are found to be trade creating. When the relative weight in the estimation is shifted from the sampling to the prior distribution, the number of RTAs that are trade creating falls to six at the 95 percent likelihood ellipsoid. Moreover, when we specify that not all RTAs are trade creating, four RTAs increase trade at the 95 percent likelihood ellipsoid. At the extreme bounds, when all weight is attached to the prior distribution, none of the RTAs are found to be trade creating. As a result, we conclude that the pervasive trade creation effect found in the literature reflects not the information content of the data but rather the unacknowledged beliefs of the researchers. (C) 2003 Elsevier B.V. All rights reserved.