Capital Formation and Technological Change in United States Manufacturing

T HE present study represents an attempt to apportion increases in output per manhour between increases in capital employed per man-hour and a somewhat nebulous constellation of forces referred to as It is hoped that a quantitative estimate of the relative importance of these two factors in contributing to an increase in the average productivity of labor in the past will help policy-makers determine what proportion of our investment resources should be devoted to improving the technology, rather than to expanding existing types of capital equipment and structures. My procedure is to examine the annual increases in output per man-hour of labor in the manufacturing sector of the United States economy between I9I9 and I955. I shall try to determine what proportion of these annual increases can be attributed to increases in capital input per man-hour, attributing the residual to technological change. The classification of causal forces is thus exhaustive, for technological change serves as a catch-all category. The implications of this will be discussed below. As a point of departure, I shall employ the model developed by Robert Solow,' which represents technological change as a shift in the aggregate production function. In its most general form, the production function can be written (using Solow's notation):