Geography and Trade.

• Author: Khalil, Elias L.

Paul Krugman, the 1991 winner of the AEA John Bates Clark medal, has produced a lucid, short, and rich book which is fun to read. Krugman advocates a geographical approach to international trade theory. The approach emphasizes increasing returns to scale, positive feedbacks, path-dependent dynamics, and the importance of historical accidents which may justify protectionist policies in certain circumstances.

To put the book in one sentence, to which the author, I think, would not object, international trade theory should be founded on Adam Smith's discussion of specialization and positive feedbacks |2~ rather than on David Ricardo's theory of comparative advantage. For obvious reasons, Krugman does not mention Adam Smith. However, he appeals to many others like Allyn Young, Albert Hirschman, Nicholas Kaldor, and Brian Arthur--some of whom have referred to Adam Smith.

The book is the collection of three lectures (with four appendices) given in Belgium in 1991. The first lecture elaborates on how market dynamics engender geographical polarization between center and periphery. The second lecture discusses the sources of localization of industry. The third lecture, which is the weakest, brings in national boundaries and policy questions.

In building the core-periphery model, Krugman employs two basic ideas of cumulative feedbacks: internal economies of scale and the externality of demand. When a firm makes a decision where to locate it takes into consideration two factors: fixed cost and transportation cost. If fixed cost is greater than transportation cost, it would be cheaper to concentrate all production in one plant in order to capture the internal economies of scale. If transportation cost is greater than fixed cost, it would be better to distribute production across regions. However, this question cannot be answered that easily. The firm has to take into consideration the distribution of population which affects the total cost of transportation. This introduces the externality of demand. In order to cut the cost of transportation, firms tend to locate, ceteris paribus, in regions with greater share of the population. However, this engenders circular feedbacks where location of demand enhances location of production which, in turn, amplifies location of demand. Such cumulative dynamics could engender multiple equilibria. An accident of history may be the ultimate determinant of the exact core-periphery structure.

For example, the...