Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Estimating the demand for electricity with a piecewise-linear budget constraint: Comparing the empirical results using OLS, complete budget constraint, and linearized budget constraint techniques

Thesis/Dissertation ·
OSTI ID:7076574
When a stochastic model is specified and estimated, the observed price (or wage) is endogenous since the price of electricity depends on the quantity purchased if there is block-rate pricing. Three techniques are used to estimate the parameters: OLS, linearized budget constraint (LBC), and a technique called complete budget constraint (CBC) which was developed in the labor literature to estimate labor-supply curves. The latter produces consistent and efficient estimates but is computationally more difficult. The LBC is an instrumental variable technique which may have a nonzero bias in finite samples. The CBC procedure, not used to estimate electric-demand equations until now, produced results significantly improved over the LBC results. Results confirm the extensive bias present in OLS. Increasing block-pricing results in positive price-elasticity estimates in some cases using both the OLS and LBC methods. Since the CBC technique is more complicated to implement an indication of how well the LBC technique will perform can be measured by calculating the correlation between the observed price and the instrument. A case study is done where data from one utility is used to estimate price elasticities using both the LBC and CBC techniques.
Research Organization:
Washington Univ., Seattle, WA (USA)
OSTI ID:
7076574
Country of Publication:
United States
Language:
English