This paper presents an econometric analysis of residential electric appliance holdings and consumption, focusing on the relationship between appliance choice and electricity demand. The authors use a subsample of 313 households from the 1975 Washington Center for Metropolitan Studies survey to test the statistical exogeneity of appliance dummy variables in electricity demand equations. They argue that if appliance choice and electricity use are related decisions, ignoring this relationship can lead to biased and inconsistent estimates of price and income elasticities. The study develops a unified model of the demand for consumer durables and derived demand for electricity. It introduces a joint water-heat space-heat choice model to estimate electricity demand equations. The authors find that the demand for electricity is influenced by both the choice of appliances and their usage intensity. They also test the consistency of electricity demand equations under different assumptions and find that ordinary least squares estimation can introduce classical bias due to the correlation between explanatory variables and equation errors. The paper discusses the implications of these findings for energy policy and highlights the importance of using appropriate statistical methods, such as instrumental variables, to account for potential endogeneity in the model. The authors conclude that the estimation of unit electricity consumption (UEC) by ordinary least squares can lead to severe bias, and that appropriate statistical methods are necessary for accurate estimation of electricity demand. The study provides estimates of price and income elasticities for electricity demand, showing that income elasticity is smaller than previously reported, and that cross-price elasticity with respect to gas is higher when using consistent estimation methods. The results suggest that portfolio shifts significantly influence price sensitivity in electricity demand.This paper presents an econometric analysis of residential electric appliance holdings and consumption, focusing on the relationship between appliance choice and electricity demand. The authors use a subsample of 313 households from the 1975 Washington Center for Metropolitan Studies survey to test the statistical exogeneity of appliance dummy variables in electricity demand equations. They argue that if appliance choice and electricity use are related decisions, ignoring this relationship can lead to biased and inconsistent estimates of price and income elasticities. The study develops a unified model of the demand for consumer durables and derived demand for electricity. It introduces a joint water-heat space-heat choice model to estimate electricity demand equations. The authors find that the demand for electricity is influenced by both the choice of appliances and their usage intensity. They also test the consistency of electricity demand equations under different assumptions and find that ordinary least squares estimation can introduce classical bias due to the correlation between explanatory variables and equation errors. The paper discusses the implications of these findings for energy policy and highlights the importance of using appropriate statistical methods, such as instrumental variables, to account for potential endogeneity in the model. The authors conclude that the estimation of unit electricity consumption (UEC) by ordinary least squares can lead to severe bias, and that appropriate statistical methods are necessary for accurate estimation of electricity demand. The study provides estimates of price and income elasticities for electricity demand, showing that income elasticity is smaller than previously reported, and that cross-price elasticity with respect to gas is higher when using consistent estimation methods. The results suggest that portfolio shifts significantly influence price sensitivity in electricity demand.