Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Models for forecasting energy use in the US farm sector. Final report

Technical Report ·
OSTI ID:6044153
In this study econometric models are developed and estimated for the purpose of forecasting electricity and petroleum demand in US agriculture. A structural approach is pursued which takes account of the fact that the quantity demanded of any one input is a decision made in conjunction with other input decisions. Three different functional forms of varying degrees of complexity are specified for the structural cost function, which describes the cost of production as a function of the level of output and factor prices. Demand for materials (all purchased inputs) is derived from these models. A separate model which breaks this demand up into demand for the four components of materials is used to produce forecasts of electricity and petroleum is a stepwise manner. A great deal of attention is focused on precise measurement of the inputs and their prices since good data are critical for reliable results. Making use of recent advances in economic theory and in the measurement of economic data, sound measures of two classes of labor, capital, land, materials, and energy inputs (petroleum and electricity) are constructed. These data are used to estimate the parameters of the structural models, employing state of the art econometric techniques. The three models are compared on the basis of their ability to predict electricity and petroleum in both short run and long run ex ante forecasting experiments. The results indicate that the most complex model performs better for short run forecasts, but the simpler models may provide more accurate forecasts for longer time horizons.
Research Organization:
Data Resources, Inc., Lexington, MA (USA)
OSTI ID:
6044153
Report Number(s):
EPRI-EA-1956(Vol.1); ON: DE81904220
Country of Publication:
United States
Language:
English