Aggregation and the Factoral Content of Trade

Abstract

With the publication of Vanek's pioneering extension (1968) of the Heckscher-Ohlin theorem to many goods and factors, it became clear that paradoxical trade patterns were possible in a well-behaved neoclassical world. In a recent article in this Journal, Edward Leamer (1980) has employed Vanek's model to demonstrate that Leontief's paradoxical trade patterns are consistent with a very restrictive set of Heckscher-Ohlin-Vanek (HOV) assumptions.' His analysis suggests that Leontief's and other investigators' appeals to ad hoc assumptions to explain the Leontief paradox are unnecessary and based on an overly restrictive view of the model's ultimate predictions. In the general HOV model employed by Leamer, it is assumed that capital and labor are two of the many factors possessed by countries and used in production. But an issue that has been overlooked and that would seem to be equally important when one increases the number of factors in the analysis is the model's predictions when different factors are aggregated. Clearly, capital and labor are categories of factors, rather than each being a single factor of production.