Agrarian Structure and Productivity in Developing Countries

The main argument of this book is that in most developing countries small farmers produce more per unit of land than large farmers and are at least comparable in total factor productivity. This is true after adjustment for differences between large and small farms in the quality of land. This situation reflects a socially inefficient use of agricultural resources resulting in large measure from the present high concentration of agricultural land in the hands of a relatively small number of large farmers. A strategy to redistribute land in favor of small farmers not only would reduce income inequality among farmers but would also increase the total product of agriculture. Such a strategy, therefore, would move societies toward achievement of both equity and efficiency goals for agriculture. Let me make clear at the outset that while I have some reservations about the theoretical underpinnings of this argument, the highly professional job Berry and Cline have done of collecting and presenting the empirical evidence convincingly supports their policy conclusion favoring a small farm strategy for the developing countries. The theoretical argument leading to the expectation that small farms will have higher yields than large farms has two components: (1) the inputs used in developing country agriculture are sufficiently divisible (machinery services can be rented on a custom basis or small machines developed for use on small farms) to yield approximately constant returns to scale; and empirical studies of agricultural production functions in developing countries in fact have found approximately constant returns. (2) However, small and large farmers confront quite different patterns of marginal rates of return or factor prices. The marginal product of family labor on small farms is less than the wage paid hired labor on large farms, and the prices of both land and capital are higher to small farmers than to large farmers. Consequently, while the technical characteristics of the production function indicate approximately constant returns to scale, the pattern of effective factor prices systematically induces the small farmer to use more labor per unit of land, and therefore to produce more per acre, than the large farmer. Two questions naturally occur with respect to this argument: (1) Why do these differences in factor prices or marginal returns persist, and in particular why does labor not move from small to large farms in response to the difference in marginal labor income, thus reducing the intensity of land use on small farms, increasing it on large farms and eliminating,