Search for: All records

Studying social-ecological systems, in which agents interact with each other and their environment are important both for sustainability applications and for understanding how human cognition functions in context. In such systems, the environment shapes the agents' experience and actions, and in turn collective action of agents changes social and physical aspects of the environment. Here we review current investigation approaches, which rely on a lean design, with discrete actions and outcomes and little scope for varying environmental parameters and cognitive demands. We then introduce a multiagent reinforcement learning (MARL) approach, which builds on modern artificial intelligence techniques, and provides new avenues to model complex social worlds, while preserving more of their characteristics, and allowing them to capture a variety of social phenomena. These techniques can be fed back to the laboratory where they make it easier to design experiments in complex social situations without compromising their tractability for computational modeling. We showcase the potential MARL by discussing several recent studies that have used it, detailing the way environmental settings and cognitive constraints can lead to the emergence of complex cooperation strategies. This novel approach can help researchers bring together insights from human cognition, sustainability, and AI, to tackle real world problems of social-ecological systems. 

Rule enforcement is critical in democratic, self-governing societies. Many political disputes occur when citizens do not understand the fundamental rationales for enforcement (e.g., COVID-19 pandemic). We examined how naïve groups learn and develop wise enforcement systems. Based on theories from behavioral economics, political science, psychology, and education, we predicted that groups need to experience failure of an enforcement system, but be guided on restorative justice principles to collectively learn from this failure. Undergraduate students (N= 288) from a Midwestern U.S. metropolitan university self-governed a simulated common-pool resource with real financial payoffs. Groups began with one of three conditions designed to create different experiences with enforcement and regulatory failure: (a) no enforcement (no communication or peer sanctioning), (b) lax enforcement (communication with peer-sanctioning), or (c) regulatory abuse (peer sanctioning without communication). Half then received facilitated guidance on restorative justice principles (e.g., discuss whether/why to use sanctions). To examine cooperation, we measured how well participants maintained the resource. To examine group learning, we created a novel coding system, which tracked groups’ constitutional decisions about conservation agreements and enforcement, conceptual understanding, and the enforcement systems they created. The no-enforcement and lax-enforcement conditions quickly yielded moderate cooperation via voluntary agreements. However, such agreements prevented groups from discovering how and why to use enforcement (peer sanctioning) to improve performance. Initial exposure to regulatory failure had different effects depending on facilitation. Unfacilitated groups fixated on initial misconceptions, causing them to abandon or create less sophisticated enforcement systems, hindering cooperation. Facilitated groups learned from prior failure—discovering principles of wise enforcement (e.g., collective efficiency, self-restraint)—and created more sophisticated enforcement systems (e.g., coordinated sanctions) that improved cooperation. Guidance on restorative justice principles and experience with regulatory abuse may be necessary preconditions for naïve individuals to understand and develop wiser collective enforcement systems. 

There is limited research about how groups solve collective action problems in uncertain environments, especially if groups are confronted with unknown unknowns. We aim to develop a more comprehensive view of the characteristics that allow both groups and individuals to navigate such issues more effectively. In this article, we present the results of a new online experiment where individuals make decisions of whether to contribute to the group or pursue self-interest in an environment with high uncertainty, including unknown unknowns. The behavioral game, Port of Mars is framed as a first-generation habitat on Mars where participants have to make decisions on how much to invest in the shared infrastructure to maintain system health and how much to invest in personal goals. Participants can chat during the game, and take surveys before and after the game in order to measure personality attributes and observations from the game. Initial results suggest that a higher average social value orientation and more communication are the key factors that explain why some groups are more successful than others in surviving Port of Mars. Neither other attributes of players nor the group’s communication content explain the observed differences between groups. 

Governance of shared resources needs to overcome collective action problems. Relational values and decision-making play a critical role in this process. Approaches are needed to stimulate self-governance, taking relational values into account. We review the literature on the use of collective action games as a tool to stimulate social learning and self-governance. We emphasize the importance of legitimacy in decision-making and the risk of crowding out internalized motivations — for instance, based on relational values — with instrumental incentive mechanisms. We further highlight the need to include ecological outcome indicators in the game design to allow the activation of relational values. Our review concludes that games used as part of a set of participatory activities enable communities to come together to identify relevant problems and craft potential solutions.