Addressing agricultural salinity in the American West : harnessing behavioral diversity to institutional design

Abstract

Salinity accumulation in the Lower Arkansas Basin (LAB) of Colorado threatens environmental quality, the agricultural economy and the potential for efficient reuse of water. Salinity is a threat to "hydraulic sustainability", since it will affect societal and environmental sustainability in a system heavily dependent on engineered structures for its water supply. Institutional solutions are preferable, being usually cheaper, quicker, and more reversible than infrastructure. Market institutions - water quality trading markets - have been often applied in the past to deal with salinity problems, but have been largely ineffective despite theoretical promise. Explanations for such institutional failure typically assume that stakeholders are boundedly- rational economic actors, but I review evidence that this is empirically unjustified, may be insufficiently explanatory, and precludes consideration of more innovative behavioral change solutions. Through collaborative work with basin stakeholders, I developed an agent-based model - "ArkAgent" - which simulates a water quality trading market; the water use and market interactions of basin actors; and basin hydrology. I conduct experiments to show that a simulated neoclassical market institution is less effective at reducing salinity when we make more realistic provisions for attitudinal and behavioral heterogeneity among resource users. I show that the use of post-hoc informational feedbacks as alternative non-monetary institutional incentives can address this performance issue, even in the face of conflicting economic pressures. I further demonstrate that exploiting social networks in non-economic incentive design can go even further in improving sustainability benefits. This work makes new theoretical contributions by showing how our models of institutional performance are critically dependent on behavioral assumptions; and that consequently our institutions for addressing hydraulic sustainability challenges may have incentives poorly matched to real behavioral complexity. This work also shows how an appropriately designed market institutional intervention in the LAB could achieve salinity reduction benefits over an 8 year period. Many of the model's practical insights are also relevant to large salinity-threatened basins across the western United States. The ArkAgent model provides an example of how we can use collaborative systems modeling and empirically-based behavioral assumptions to develop more robust institutions for sustainability.