Biased Technological Change in Agriculture: The Hayami-Ruttan Hypothesis Revisited

This paper presents a consistent explanation for biases in the direction of technical change in agriculture. The emergence of such biases had been pointed out by Hayami and Ruttan (1970) based on a study of long time series of agricultural inputs and technologies in Japan and in the US from 1880 to 1980. Land scarcity in Japan seems to have favored land-saving technologies like new crop varieties and the use of, while land abundance in the US made technological change more labour-saving due to the use of tractors and machinery. A theoretical explanation for the relationship between factor-biases in technological change and relative factor proportions can be given in the framework of the model of directed technical change developed by Acemoglu (1998, 2000, 2001). In this model the elasticity of substitution between factors of production is a key determinant of the direction of the technological bias. In our paper Acemoglu’s model of direct technical change is adapted to the analysis of agricultural production. The implications of the model are tested with data from 93 countries between 1961 and 2002. Based on this test we can confirm the Hayami-Ruttan hypothesis in its original form on a global level. When we test across regions, levels of income and levels of land inequality we find that those countries have developed in the most successful way, where the direction of technological bias in agriculture has been the least biased by market imperfections.