A review of mathematical programming models of irrigation water values

Abstract

By introducing the user-pays principle into the irrigation water pricing debate, the 1998 National Water Act created a demand for models to measure willingness-to-pay for irrigation water. Water values are traditionally simulated with mathematical programming models. Models differ in their treatment of crops, irrigation options and water constraints, and other firm-level characteristics but they all use shadow prices as an indication of water value. The 17 models reviewed here, report average annual water values of between $0.0042·m-3 and $0.1899·m-3. Crops modelled influence water values, but there is no apparent relationship between objective function specification and average value. Nor does the number of irrigation options seem to influence water value either. The policy implication is that while similar models for the same region produce consistent estimates, each region requires its own model that has to be updated regularly.