Abstract
This paper develops a quantitative computational framework for analyzing robustly optimal policies within a mean-field control setting. Drawing on dynamic decision theory, asset pricing, mean-field theory methods, and advanced deep reinforcement learning techniques, we present an approach to compute prudent social policies, evaluate the underlying economic mechanisms, and quantify the relative importance of alternative uncertainty channels influencing policy decisions. Using a climate change application, we demonstrate that incorporating uncertainty aversion into our mean-field framework systematically tilts robustly optimal abatement and R&D investment strategies, thereby mitigating inequities in climate damage outcomes.
The seminar will be held in room 750 (7th floor) at Eilert Sundts Hus. The address is Moltke Moes vei 31.