Regional Integration and Dynamic Adjustments: Evidence from a Gross National Product Function for Canada and the United States

We propose an empirical trade model to test for structural change and dynamic effects induced by free trade agreements for the Canadian and US economies. We estimated a translog Gross National Product (GNP) function along with output and factor shares and tested for structural change (abrupt or gradual) which is endogenously determined by the data. After this, we estimated Stolper-Samuelson (SS) and Rybcynski (R) elasticities, and assessed the stability of their sign and magnitude link to the structural change. The null hypothesis of no structural change is soundly rejected for both countries. For Canada, we found gradual structural change that started prior to the implementation of CUSTA and lasted for several years. In the US case, we found evidence of an abrupt structural change occurring in 1995, a year after NAFTA came into force. More interestingly, several SS and R elasticities experienced sign reversals and a magnification effect over the different sub-periods, implying that the categorization of goods in terms of friends or enemies of labour and capital changed during the transition.