Buy me a river: Use of multi-attribute non-linear utility functions to address overcompensation in agricultural water buyback

The realization of buyback welfare enhancing opportunities is conditioned to the ability of government agencies to place bids consistent with the shadow price of irrigators. However, methods used to inform buyback programmes to date either rely on ex-post trading data that is not readily available in most regions worldwide; or compensate projected foregone income, and thus ignore the effects that buyback may have on other relevant attributes determining utility. This paper uses revealed preference methods to elicit the parameters of a multi-attribute objective function that mimics the observed behavior of irrigators in the overexploited Segura River Basin in SE Spain. Objective functions are used in a series of simulations in which water allocation is progressively constrained to ex-ante reveal the shadow price of water using two alternative compensation measures: i) the foregone income, a proxy of the shadow price typically used in the literature; and ii) the compensating variation that addresses foregone utility. Results show a relevant gap between the two methods For example, restoring the balance in the basin through purchase tenders would demand an investment of million 2400+ EUR (9.6+ EUR m−3) attending to the foregone income method, and million 950+ EUR (3.8+ EUR m−3) (−60.3%) with the foregone utility method.