Does decentralized management of irrigation ensure efficient use of water? Theory and evidence from India

Abstract

In this project we propose to determine whether and how decentralized management of irrigation system may help achieve efficient extraction and distribution of water in agriculture. We plan to use the staggered constitution of locally elected irrigation management bodies, known as “Pani Panchayats” (PP), in the state of Orissa in India to determine if decentralization leads to better use of water in agriculture. Decentralization may lead to improved (static) distribution of water since farmers in a locality are presumably better aware of their local water requirements than the central planner, i.e., in our case, the state government. However, the central planner may have a better understanding of the inter-temporal trade-off of using a scare natural resource such as water. This is especially true if the local agencies are elected bodies, as is in our context of study, since they will be subjected to reelection concerns. Since election would incentivize the members of PP to ensure short run water availability, high level of reelection incentive may distort inter-temporal efficiency of water use by increasing current extraction. Following this logic, we hypothesize that extending the tenure of the elected PP members may improve inter-temporal allocation of water. We plan to test this hypothesis using an amendment of the relevant Act that increased the tenure of a random half of the PP members from three to six years before the second round of elections