On Technical Progress and the Speed of Adjustment

Abstract

This paper contends that greater embodiment of technical progress need not always lead to faster given an initial disequilibrium within a neoclassical growth model. Furthermore it is suggested that changes in the mean age of the capital stock over time may provide a means for reconciling an apparent paradox relating to the question of time. The paradox arises from the fact that the mathematics of the long run indicates that embodied and disembodied models will close a given disequilibrium in the same period of time even though the embodied model will presumably always have a higher initial growth rate. In an article by K. Sato (1966) dealing with the question of time in neoclassical growth models it was concluded that the adjustment process is completed in a shorter period of time if . . . technical improvements are more embodied rather than disembodied. The position taken here is that Sato's result is a special case rather than the general case as suggested by his article. It should be noted that nlo exception is taken with respect to his formal analysis. Sato (1966) develops the vintage neoclassical growth model made familiar by Solow (1960). Assuming a Cobb-Douglas production function, and denoting the rates of embodied and disembodied technical change as A and y respectively, the following expression is obtained: