The residential demand for electricity in New England,

Abstract

The residential demand for electricity, studied on the national level for many years, is here investigated on the regional level. A survey of the literature is first presented outlining past econometric work in the field of electrical energy demand. An econometric model is then presented using a 1970 cross-section of sixty-seven New England electric utilities and their service areas as the data base. The special nature of the market for electricity, characterized by decreasing block rate structures, requires that both a supply price equation and a demand equation be estimated. A suitable technique, two-stage least squares, is used to obtain consistent coefficients, and elasticities of demand are obtained.

The residential demand for electricity is found to be significantly correlated with its average price, family income, family size, heating degree days, and the ownership -- public or private -- of the electric utility. Price and income are found to be the most important determinants of demand. The cross-elasticities of demand with respect to the average prices of natural gas and fuel oil are unexpectedly negative. This is thought to be due to poor data in the case of fuel oil and the unavailability of natural gas in much of the region. Another explanation may be the inability of the model to portray what might be a non-linear cross-elasticity function. The supply price is correlated with the quantity of electricity consumed, utility operation and maintenance costs, total number of customers, degree of urbanization, and the ownership of the utility. The structural supply and demand equations are also estimated using the ordinary least squares procedure. The coefficients derived in this manner are not significantly different from those obtained using the two-stage least squares technique, a rather surprising result considering the simultaneity of the demand and supply price equations.

Three methods are investigated for obtaining the total elasticities of demand from the reduced-form of the demand equation. These elasticities, contrasting with the direct elasticities derived from the structural equations, are presented for comparative purposes. Policy implications of the results of the equation estimation are discussed as are suggestions for further research in this field.