Short-run residential demand equations for electricity and gas are estimated in this study. Short-run demand depends on the appliance stock in existence. Use of the appliance stock is a function of the price of fuel, income, and the weather. The major difference between this study and others explicitly using appliance stock data is that appliances are not aggregated into a single stock measure. Demand consists of the sum of the individual demands for energy for each fuel-burning appliance type. Consequently, different price, income, and weather elasticities are estimated for each use of the fuels. The data consist of annual observations for each state for the years 1960-1975. Most of the appliance stock data were developed by Data Resources, Inc. These are supplemented by appliance data developed for use in this study. Two different methods of pooling time-series and cross-section data, the random and fixed effects models, are used, and a specification test is performed to test for consistency of the random effects model estimates. The results are somewhat mixed. However, they do suggest directions for further research. Fairly reasonable estimates in terms of average energy consumption for each type of appliance are obtained. The aggregate price and income elasticities fall in the range found in previous work. Price elasticities appear to vary among the demands for fuel for different end uses, but the differences are not statistically significant. Income elasticities for the individual fuel uses are disappointing; they are often of the wrong sign and magnitude. The most reasonable results are obtained for the appliances which consume the most fuel. Further work most likely would benefit from aggregation of the small appliances, leaving only for estimation the coefficients of demand for the major users of fuel and the residual aggregate appliance stock.
展开▼
