Estimation of a TVP Demand for Residential Electricity in Iran using State-Space Approach

Own-Price, cross-price and income elasticities of demand are of great importance in policy making and forecasting. Previous studies have considered these quantities as constants. Since generating parameters of these quantities are unstable over time, the calculated elasticities are biased and inconsistent due to misspecification. In this paper, we try to estimate own-price, income, temperature and cross-price elasticities by using Kalman filter within a space-State model and by performing Hansen test to reject the stability of parameters. We estimate the elasticities of demand for residential electricity in Iran over the period 1970-2011. Variables under study include electricity consumption per capita, real electricity price, per capita income, temperature and real gasoline price as a proxy for electricity in residential sector. Results show that price, income and temperature elasticities of demand for electricity are time varying parameters (TVP). In addition, some overshoots are observable in these elasticities over time, so that absolute value of price elasticity indicates an increasing jump in 1980. The income elasticity has an overshoot from zero in 1973, and then it decreases and approaches to more stabilized trend. Finally, temperature elasticity has recorded an upward jump from 1980 onward.