Knowledge that Transforms

David Ricardo and Thomas Malthus saw the problem of population growth and attendant diminishing returns to labor as a major economic problem. Diminishing returns to labor led, of course, to the gloomy Malthusian law of population growth limited only by the subsistence wage of labor. In contrast, neoclassical growth theory which is characterized by its optimistic outlook on an economy's ability to achieve balanced growth shunts aside the problem of population growth and diminishing returns to labor as a mere qualification to the main argument. The difference between the conclusions of classical and neoclassical growth theories stems from their assumptions concerning the production function. In classical theory, land enters the production function as a factor limiting the level of output; its fixity in supply then characterizes the production function by diminishing returns to scale. The neoclassical conclusion of a stable balanced growth, on the other hand, is based on the assumption that production processes proceed by constant returns to scale with respect to capital and labor; land does not enter the production function as a factor limiting the level of output. The neoclassical conclusion may be valid in a situation in which land is a-free good: its supply increases in proportion to other factors at zero price. The United States in the early nineteenth century represented such a situation when the frontier was defined simply as uncleared land. The neoclassical conclusion may also be applied to a highly indtustrialized state of an economy in which the proportion of agriculture is small and in which the agricultural sector is highly modernized. In today's less developed countries, however, a traditional agricultural sector is dominant in the national economy. For such an economv, classical theory, rather than neoclassical, is the more relevant. Conventional theories of development, such as those developed by Arthur Lewis, John Fei and Gustav Ranis, Dale Jorgenson, Paul Zarembka, Ryuzo Sato and the author, have specified the agricultural production condition by the classical assumption of diminishing returns to scale.' Conclusions derived by these studies clarified the problem of development as one of the balance between two forces: productivity increases in agriculture caused by various forms of technical progress; and the forces of diminishing

Since Paul Samuelson's factor-price equalization theorem that commodity trade could serve as a perfect substitute for factor mobility, the mirror image problem of the conditions under which factor mobility could perfectly substitute for commodity trade has attracted the attention of such trade theorists as Robert Mundell, Ernest Nadel, and most recently Frank Flatters. The latter, in analyzing the case where factor-owners move with their exported productive factors, does not consider the question of the uniqueness of the equilibrium obtained. This is the subject of the present paper. It will be demonstrated that in the case where factor prices differ between countries in closed economy equilibrium, and where such difference is due to different factor endowment ratios in the two countries, f actor-price equalization is consistent with an infinite number of combinations of factor flows, the determination among which on a priori grounds is impossible. There will, however, be a single determinate mirror image solution in the case where trade is due to differences in tastes between the two countries.