ESTIMATION OF RESIDENTIAL DEMAND FOR ELECTRICITY WITH THE CROSS-SECTIONALLY CORRELATED AND TIME-WISE AUTOREGRESSIVE MODEL

Following Taylor (Taylor, L.D., 1975, The demand for electricity: a survey, Bell Journal of Economics and Management Science 6, 74-110) and Houthakker (Houthakker, H.S., 1980, Residential electricity revisited, Energy Journal 1, 29-41), we estimate residential demand for electricity for five Southern states during 1981-90 with the cross-sectionally correlated and time-wise autoregressive model (CSCTWA) that simultaneously takes into consideration heteroscedasticity, autocorrelation, and spatial correlation that are likely to be detected when a pooled sample is used. There are a number of interesting findings. The CSCTWA model yields more efficient estimates with smaller standard errors than the OLS and the CSHTWA models. The price of natural gas is insignificant of the CSCTWA model is not considered. The log-linear form can not be rejected, whereas the linear form can be rejected. Thus, the elasticity of demand remains constant, but the slope is expected to vary.