Output and Input Biases Caused by Public Agricultural Research and Extension in Japan, 1957–1997

Abstract

The present paper investigates the magnitude and the Hicksian output and input biases of technological change brought about by investments in public agricultural research and extension activities in Japan. Given this objective, it estimates a translog multiproduct cost function for 1957–1997. Empirical results show that the cost reduction effects were fairly large. The Hicksian (1932) biases were found to be: (i) livestock‐augmenting; (ii) labor‐ and other‐inputs‐saving; and (iii) machinery‐ and intermediate‐inputs using. Except for other inputs, the directions of the biases are consistent with the Hicksian (1932) induced innovation hypothesis, which supports the public‐sector‐induced‐innovation model proposed by Hayami and Ruttan (1985) and Ohtsuka (1982).