ABSTRACT
This chapter investigates the ordinary least squares bias in estimates of water demand functions under decreasing and increasing rate structures. However, opinions concerning the appropriate methodology for estimating water demand differ significantly. The chapter compares the instrumental variables, ordinary least squares, and two stage least squares methods in estimating water demand using a unique data set containing both decreasing and increasing block rates. It presents the data set on water consumption, which consists of a random sample of 101 individual customers' monthly water use records from the city of Denton, Texas for the summer months of 1976-1980 for the decreasing block rate period, and 1981-1985 for the increasing block rate period. In the decreasing block rate scenario the ordinary least squares (OLS) technique resulted in a statistically negative price effect, whereas the increasing block rate scenario resulted in a statistically positive price effect.
