Résumé
Water managers are increasingly aware of the problem of water scarcity and the randomness of rainfalls. This problem is exacerbated by recurrent droughts observed in south-western France in recent years where water consumption exclusively by agriculture often exceeds 80% in summer. In the given context of climate hazard, some management companies have introduced new water pricing methods with very specific features aiming particularly at a certain anticipation of the demand of irrigation water. The objective of this research is to analyze the effect induced by the application of these different water pricing methods on water demand, especially in case of drought, on farmers' income and on the revenue collected by the management company. To undertake this analysis a stochastic model that simulates farmers' behavior and their response to different water pricing scenarios has been built. Empirical application of the model has been carried out with the help of an agronomic model of plant growth and data collected from Midi-Pyrénées (France). The results show that these pricing policies create a wide range of effects that can be searched by management companies according to their characteristics and their access conditions to the resource. These pricing systems prove to be powerful tools to mitigate the impact of drought. This study provides very useful lessons for the design of water management policies.