From decentralized to centralized irrigation management

Abstract

Centralized organizations can internalize transaction costs and externalities, addressing possible inefficiencies of decentralized management. In practice, however, centralized organizations can have its own inefficiencies and decentralized management can do relatively well. Empirically testing relative efficiency is difficult because distinct organizations emerge endogenously in various contexts. This paper, in contrast, draws upon the unique history of New Mexico, leveraging a natural experiment to assess how the partial transition in the early 20th century from the original small decentralized communal Spanish irrigation systems (acequias) to centralized quasi-public irrigation districts altered agricultural development and production. Surface water irrigators in arid regions confront public good issues for building and maintaining shared infrastructure as well as common-pool resource issues to appropriate the surface water. My results confirm that irrigation districts can significantly improve outcomes when investing in costly infrastructure to expand irrigated acreage, increasing farmland values up to 33%. However, I find no broader evidence that the centralized control of water distribution provides any gains to acreage previously under irrigation by the decentralized acequias.