Dylan Munson (4th year PhD) - UPEP
Environmental policy in the context of complex systems: statistical theory and practice
Date: 02/14/2025 (Fri)
Time: 2:00pm- 3:00pm
Location: This seminar will be held both on-site and remotely. The on-site location will be: Rubinstein 149- Sanford School. It will also be held remotely via Zoom. (Please sign in to see the link.)
Organizer: Alex Herrera, Paula Sarmiento and Xingchen Chen
Meeting Schedule: (Not currently open for scheduling. Please contact the seminar organizer listed above.)
All meetings will be held in the same location as the seminar unless otherwise noted.
2:00pm - Seminar Presentation (2:00pm to 3:00pm)
Additional Comments: Abstract: Coupled human-environment systems are increasingly being understood as complex adaptive systems (CAS), systems in which micro-level interactions between components lead to emergent behavior and components change their behavior in response to signals from other agents and the system. Practically speaking, this means that makers of environmental policy must embrace new techniques, such as agent-based modeling (ABM), in order to properly understand the impacts which interventions may have on ecologies, geographies, and anthropological systems. In this paper, we develop, using a simple spatial ABM of resource harvesting, a framework for optimizing over various policy options when one of these goods has a negative effect on the environment. The model is Axtell and Epstein’s classic “Sugarscape” simulation with trade and pollution, and as policy levers we include taxes and caps on the dirty good as well as reinvestment schemes into the clean good. We use techniques drawn from the statistical literature to run ensembles of the model and optimize over these policy options, in particular Bayesian (optimization). We demonstrate that, despite several challenges, such a technique provides reasonable policy results and performs better than a random sampling technique. The work thus contributes to the literature on spatial ABMs as well as coupled human-environment policymaking more generally. Stage: Late stage work Feedback: Any and at any time welcome