LONG-RUN DETERMINANTS OF INPUT PROPORTIONS IN THE ELECTRIC UTILITY INDUSTRY

We study the long-run determinants of input proportions in the production process of the electric utility industry. We analyze the effects of changes in relative input costs, environmental pollution regulations, plant location, load variation, cooperation among the utility firms, removal of oil import quotas, and other factors on input shares of coal, oil, gas, capital, and labor inputs. We also investigate their effects on long-run plant and fuel choices.;Demands for inputs are found responsive to their own prices and prices of other inputs. Demands for coal, oil, and gas are elastic with respect to their own prices; and demands for capital and labor are very inelastic. Demand for oil is most price elastic. Pollution emission regulations have resulted in 2.87 percent higher capital shares (in coal-burning plants) and substitution of low-sulfur coal for high-sulfur coal. Plants located in the Atlantic states have higher oil shares than plants in the rest of the country; those located in petroleum-producing regions have higher gas shares; and plants located in coal-producing regions have higher coal shares even after accounting for prices of inputs. Decontrol of oil import quotas in 1966 has led to significant rise in oil share, particularly in the Atlantic states. Gas share in new plants has decreased substantially in recent years. Cooperation among utilities in the form of sale of electricity to one another for resale has positive impact on capital share, but cooperation in the form of joint ownership of plants does not. Joint ownership of new plants has risen rapidly to about 30 percent of total. Plant output is positively correlated with coal share and negatively correlated with oil and gas shares.;Choices of plants in terms of their design and equipments to use specific fuel(s) and initial choice of fuel(s) depend, in a similar way, on input prices and competitive position of fossil fuels in various regions.;An econometric study of input shares requires the derivation of input share equations. We derive a system of input share equations from translog cost function applying the results of Duality theory. The derived demand equations contain only input prices and plant output as exogenous variables. Variables representing other influencing factors have to be defined and included in the model. In the econometric model, error terms are correlated across equations, and fuel share values are clustered at zero. We apply Iterative Zellner and tobit estimation techniques to obtain efficient and unbiased parameter estimates. The study of plant and fuel choices requires the specification and estimation of logit models.