An agent-based model of agricultural water market structure and function

An agent-based model (ABM) is developed to simulate the function of lease markets for surface water irrigation rights operating at the watershed scale. The ABM examines the effect of drought on trading activity and gains from trade under three market models: bilateral trading with random matching, a “smart market” that actively matches bidders based on bids and asks, and a full-information optimization model that represents the efficient benchmark against which to assess other market models. The model allows for variation in farm characteristics, water rights seniority, diversion location, and includes a watershed accounting system as a novel component. Results suggest gains from trade and market participation are largest with moderate drought severity than with severe or weak drought. Smart markets provide larger gains than bilateral trade, but both are lower (half or less) than the efficient benchmark. We also show that higher watershed complexity and certain distributions of agents across the watershed can significantly reduce market participation and gains from trade. The applications presented in this paper illustrate the broad applicability of the framework to water resource problems under appropriative water rights systems of the Western United States.