Simulating Dynamic Complexity in a Water Quality Trading Market: Insights for Design of Nutrient Trading Programs

Nutrient credit trading is a market-based policy currently proposed in several countries to mitigate nutrient pollution underpinning eutrophication and anoxic “dead zones” in coastal waters and lakes. Under nutrient credit trading programs regulated sources of nutrient pollution with high discharge compliance costs, such as wastewater treatment plants, are allowed to meet discharge restrictions by purchasing credits representing discharge reductions from other sources with low compliance costs, such as agricultural operations. The potential for cost savings is huge; also nutrient credit trading can, in theory, incentivize traditionally non-regulated sources, such as agriculture, to voluntarily adopt better nutrient management practices and thereby reduce nutrient pollution at an aggregate scale. Despite its intuitive appeal, and the successes of emissions trading markets for some atmospheric pollutants, the performance of nutrient credit trading is widely considered to be disappointing. A recent international survey by the World Resources Institute found over 50 nutrient trading programs at some early stage of development, yet very few trades had actually occurred. Many of the reasons offered by researchers for the apparent stagnation of nutrient credit trading appear to be due to learning impediments. Nutrient credit trading presents a difficult learning environment featuring multiple agents with disparate decision making criteria, complex feedback and stock-and-flow structures, and lengthy time lags between causes and effects. This research develops a system dynamics model to facilitate understanding of the dynamic complexity in nutrient credit trading systems. It is anticipated that following models developed on similar principles, and configured into appropriate user-friendly formats, may provide valuable insights to designers of these systems. The model explicitly takes into account feedback, stock-and-flow effects, time lags, and agent decision-making processes. A guiding principle in the model development was to capture the essential dynamic complexity of the nutrient trading system while striving to maximize model transparency and comprehensibility. A range of policy experiments were made with the primary objective of developing a realistic design that attains aggregate nutrient loading targets recommended by researchers. These experiments focus on settings and interactions between policy elements that include loading caps, trading ratios, exceedance penalties, investment of funds accumulated from exceedance penalty payments. Specific design recommendations are offered; however, these recommendations should not be considered definitive but exploratory in nature. The primary value of the research is viewed to be demonstration of the value of the system dynamics modeling approach to support trading policy design through simulation experiments and enhanced learning