Chris Dillow writes,
Now, in saying this I don’t mean that the theory of comparative advantage is correct. It isn’t. There’s less (pdf) trade in some ways than the theory predicts, and more trade within industries and between neighbouring countries. Some guy got a Nobel in 2008 for analyzing these issues.
In case you didn’t follow the links he provides us, Dillow is referring to Krugman. Krugman is known for his work in trade theory, specifically the application of the concepts of monopolistic competition and returns to scale. This was motivated by the fact that empirical evidence did not support a straightforward Ricardian story, and early marginalist developments (viz. Heckscher-Ohlin) of that theory were not empirically successful either.
But, my interpretation of Krugman’s work is different from Dillow’s. While it’s easy to say that the former’s research seems to disprove the empirical significance of comparative advantage, I interpret Krugman’s research as making the theory of comparative advantage more complete. Ricardian theory tells us that the agent with the lowest cost associated with some line of production will specialize there. Krugman told us that there’s more to the determination of these costs than innate characteristics and factor endowments, and that we need to consider return to scale and economies of scale. These latter factors also influence costs, and trade patterns can reflect the role economies of scale play in determining where firms locate themselves and which firms will have competitive advantages.
I’ve been meaning to write a post on this subject for some time and, while I haven’t found the time yet, I should mention that I interpret Krugman’s work in trade theory as a dynamic model within the general framework of Ricardian trade. The phenomena he’s trying to explain is clearly dynamic, in that the effects of economies of scale take time to reveal themselves — current trade patterns reflect economic geographies that took time to come into place. In fact, the post I have in mind is meant to be somewhat controversial, because I think there is an affinity between Krugman’s trade theory and the Austrian concept of the market process.
Unlike Dillow, I don’t think that Ricardian trade theory has been proven false. I think that basic Ricardian theory has been proven to be incomplete — it does not consider costs that we now know are relevant. Far from being a substitute to Ricardian theory, Krugman’s research should be seen as complementary.