Investment Behavior in the U.S. Electric Utility Industry, 1949-1968
Volume: Volume 3, No. 2
Issue: Autumn 1972
Pages: pp. 645-664
Authors: U. Sankar
Title: Investment Behavior in the U.S. Electric Utility Industry, 1949-1968
Abstract: Jorgenson and Handel (J-H) applied their recently developed econometric model of investment behavior in U.S. regulated industries to four subindustries of the regulated sector. In their derivation of the demand for capital, they assumed constant returns to scale and unitary elasticity of substitution between capital and labor. For application to the U.S. electric utility industry the author finds it desirable to relax these two assumptions. Econometric studies on production and cost functions pertaining to this industry show that economies of scale are quite important and that the scope for substitution between capital and labor is limited. Using a CES production function, the author derives a more general expression for desired capital stock. This expression is incorporated in a logarithmic version of a rational distributed lag function in a manner suggested by Eisner and Nadiri. As in the J-H model, it is assumed that replacement investment is proportional to net capital stock.
The distributed lag function is applied to annual data relating to investor-owned electric utilities in the United States for the period 1949-1968. The estimates of long-run elasticities of capital stock with respect to relative price and output are of reasonable magnitudes and are in accord with results obtained from production function studies. The author finds that an investment function derived from a Leontief-type model is suitable for analyzing investment decisions in this industry. As for the characterization of the time structure of investment, he concludes that Joregenson's rational distributed lag function is preferable to geometric distributed lag functions.