earticle

영어

This paper develops a joint model of the demand for electricity and the demand for appliance by residences. Based on consumer's utility maximization, it is sought to find the simultaniety and endogeneity between electricity demand and appliance demand. The model which ignores these facts will result in biased and inconsistent estimation. It is known that a conditional demand model cannot allow for sample selection problem in the model. As an alternative for it, we used the sample selection model, using micro level data. The empirical results are supportive of the existence of sample selection bias. An important implication of the findings is that the conditional demand model may lead to the overestimation of true effect of appliance demand.