Abstract:
The pattern of trade observed from firm-product-country data calls for a new generation of models. To address the unexplained variation in the data, we propose a new model of monopolistic competition where varieties enter preferences non-symmetrically, capturing both horizontal and vertical differentiation in an unprecedented way. Together with a variable elasticity of substitution, competition effects, varying markups and prices across countries, this results in a tractable model whose predictions differ from existing ones. Using the population of Belgian exporters, our model succeeds in explaining the hitherto unexplained variation. The implications call for a re-thinking of earlier results and measurement practices.