General introduction:
Jonathan Donges and Wolfram Barfuss are researchers in the flagship project COPAN on Coevolutionary Pathways in the Earth System at the Potsdam Institute for Climate Impact Research (Germany). The COPAN team works on modelling social dynamics in an Earth system context from a complex systems perspective. The focus is on understanding various global change problems and the dynamics of sustainability transformation, including climate and social tipping elements and their networked interactions.
Talk Jonathan Donges: Will divestment burst the carbon bubble?
To achieve the ambitious aims of the Paris climate agreement, the majority of fossil-fuel reserves needs to remain underground. As current national government commitments to mitigate greenhouse gas emissions are insufficient by far, civil society actors such as the social movement on divestment from fossil fuels could play an import role in putting pressure on national governments on the road to decarbonization. Using an agent-based model of co-evolving financial market and investors' beliefs about future climate policy on a social network, we find that the dynamics of divestment from fossil fuels shows potential for social tipping away from a fossil-fuel based economy. Our results further suggest that socially responsible investors have agency: a small share of 10–20 % of moral investors is sufficient to initiate the burst of the carbon bubble, consistently with the Pareto Principle. These findings demonstrate that divestment has potential for contributing to decarbonization alongside other instruments, particularly given the credible imminence of international climate policy. Our analysis also indicates the possible existence of a carbon bubble with potentially destabilizing effects to the economy.
Talk Wolfram Barfuss: Modeling human behavior with Generally Applicable Behavioral Algorithms
Mathematical and computational models incorporating human behavior are of increasing importance in the social, economic and sustainability sciences. However, it is an open question how to meaningfully formalize social dynamics in the context of mathematical social and social-ecological systems modeling. Existing models often use a system dynamics approach of aggregated quantities, thereby not being able to account for complex social network effects, social stratification and human agency - all presumably central issues for global sustainability; other types of models incorporate these ideas, but put their focus rather on case specific systems and tend not to use a generalizable approach. In this project I combine the concept of Partially Observable Markov Games (POMGs) with a co-evolutionary social-ecological systems perspective by equipping the social agents with artificial intelligence inspired behavioral algorithms, making them capable of adapting and thereby being generally applicable to multiple different environments. The formalism of POMGs additionally guarantees the availability of real-world observations of human behavior through multiple behavioral experiments, for comparison against the behavioral algorithm. This is important because the two requirements: i) general applicability and ii) comparability to behavioral experiments enable ways forward to improved mathematical representation of human behavior, eventually making model conclusions more reliable for policy and business.