Originally published on 04/05/2007
Q. I recently read an editorial from the New York Times [dead link] that was written in 2004 by Charles Schumer (senior senator from New York ) and Paul Craig Roberts (assistant secretary of the Treasury for economic policy in the Reagan administration). In it they argued that the case for free trade which is founded on the principle of comparative advantage is no longer valid since one of the assumptions made by Ricardo, that factors of production are immobile, is no longer valid. Does the fact that factors of productions are now easily mobile across national borders really ruin the whole logic of comparative advantage?
A. This article inspired quite a few quick responses. (1), (2) [dead link] Rather than covering the same ground, I’ll make just a few points. First of all, it is common for people to question the validity of a model or theory’s assumptions. Because models are simplifications of the world, the assumptions made are ALWAYS unrealistic in many ways. However, these unrealistic assumptions are made NOT because the model makers believe them to be true, but rather because it is the only way to make the mathematical or graphical model easy enough to work with. Introduce too many real world assumptions and a model becomes unintelligible due to complexity.
A common argument made by critics of economic theories is to say something like Schumer and Roberts did. Namely, the theory of comparative advantage assumes that factors don’t move across countries, but because that assumption is not true, the result of the model, that countries both benefit from trade, is therefore false. This is a logical fallacy. We cannot conclude that if one assumption is invalid the conclusion is false. We can, however, logically conclude that the result is uncertain, but not that the result is false. This is an important difference.
Indeed this highlights the way knowledge is developed using economic models. Many advances in economic theory have arisen because someone questions the validity of a model’s assumptions. But the way an economist develops such a criticism is to change the offending assumption to something more realistic and see if, in the context of the same type model, the results change. In this way we can understand whether and how the results are dependent on that one assumption.
Schumer and Roberts simply proclaim the results must be different but don’t actually show or explain why. Bhagwati, Panagariya and Srinavasan however, develop the more appropriate response by incorporating factor mobility into a Ricardian model to show that it doesn’t change the result that free trade is better for both countries. (see page 10 in this paper).
Understanding international trade theory requires one to understand the limitations of each model and to recognize how the results are affected as assumptions are changed. Sometimes the assumptions are critically important, many times they can be shown to be immaterial. For example, one Ricardian model result is that EVERYONE benefits from trade. But this result is dependent on the assumption of one factor of production. Change the assumption to include both capital and labor and this result changes. Thus, economists don’t believe (at least they shouldn’t) that free trade benefits all people even though the Ricardian theory of comparative advantage says so. On the other hand it has been virtually impossible to reverse the Ricardian insight that if all countries shift production to their comparative advantage goods, overall economic efficiency must increase.
