Estimating the demand for irrigation water in the Central Valley of California

This paper computes estimates of the demand for surface irrigation water directly from disaggregated profit functions for fields in the San Joaquin Valley of California. It finds that treating delivered surface water and pumped ground water as separate, imperfectly substitutable inputs to production matters a great deal. We find substantial ranges of inelastic demand for delivered water, and thresholds across which demand then becomes highly elastic. The results imply that moves toward freer water markets could lead to large quantities reallocated from agriculture to urban uses in the Western U.S., but would require large price increases and would induce extensive ground water mining and major changes in cropping patterns. While these results are dependent on our particular model and simplifying assumptions, evidence exists that they may be robust.