An econometric estimation of industrial water demand in France

This study investigates the structure of industrial water demand by estimating the derived demand for water on a sample of industrial establishments located in the south-west of France. Production technologies are represented by short-term variable cost functions and approximated by a translog form. Industrial water use is modeled as having three components: the quantity of water bought to a water utility, the quantity of autonomous water and the quantity of water treated prior to use. We include in this framework water effluents, which are considered as a by-product of the production process, emitted by firms. Each of the three water components is treated as a separate input and all are estimated as a system of simultaneous equations. The model is estimated on a sample of 51 industrial plants in the Gironde district observed from 1994 to 1996 using Seemingly Unrelated Regression ( SUR) and Feasible Generalized Least Squares (FGLS). Results of estimations show that industrial firms are sensitive to water price inputs. Network water elasticity is estimated at -0.29. It varies from -0.10 to -0.79 according to the type of industry considered. Autonomous water price elasticity is not significant. Elasticity for treated water is evaluated at -1.42 at the mean-sample and varies from -0.90 to -2.21 according to the industry considered.