This paper provides a selective survey of over half a century of research linking the neoclassical trade model to the data. Three lessons stand out. First, competitive and new trade theory models are complementary rather than competing ways to look at many existing empirical regularities. Second, the Ricardian formulation has proved to be a useful framework for structural estimation regarding the pattern of international specialization. Third, empirical confirmations of the core predictions of the model provide scientific support for employing the competitive trade model in structural estimation.
testing and estimation of neoclassical trade theory