Irrigation water management under risk: An application to Cyprus

In this paper we provide empirical evidence that attitudes towards risk are important when assessing the impact of conservation policies on production choices. In the case of an irrigation water quota, deriving water demand responses from a multi-input farmer decision model under risk has to account for unobserved risk attitudes. We first follow the approach of Antle (1987) based on higher-order moments of profit, which enables flexible estimation of the stochastic technology without ad hoc specification of risk preferences. In a second step, the impact of the environmental agency's quota policy on farmer decisions can be solved, using risk aversion and technology parameter estimates. The advantage of our method is that it allows direct simulation of a wide range of quota policies, using only profit and input level information. Application to a farm-level data-set from the agricultural region of Kiti in Cyprus, shows that the second and third moments (variance and skewness) of the profit distribution influence farmers behaviour under risk. Consequently, variability considerations are not sufficient and should be complemented by the inclusion of attitudes towards downside risk as well.