Towards an Economics of Irrigation Networks

Both the economics and the engineering of irrigation design are typically based on the assumption of a single source. The more general economic problem is to determine which sources should be developed and how water should be allocated and delivered to various receptor-farmers. This is a problem in network economics. We begin our exploration with the problem of allocating irrigation water from existing sources when the conveyance structures are already in place. Transporting water from a particular source to a farmer entails a conveyance loss such that only a fraction of water sent from the source is received by the farmer. Optimal allocation requires that irrigation demands are matched with the least-cost source, including conveyance losses. Economic networks are then defined as optimally-matched subnetworks. Allocation within each economic network is then determined by equalizing the marginal products of water across farmers, reckoned at the source. Different cases are considered depending whether the sources have similar or different cost functions. We provide a modest beginning to the problem of endogenous sources by examining the problem of locating a single source within the network. We also provide a possible reconciliation of equity and efficiency objectives.