Exact solutions of the simplified March model for organizational learning

Abstract

March's celebrated agent-based simulation model for organizational learning [J. G. March, , 71 (1991)] has been extensively studied in recent decades. Yet the model has not been fully understood due to the lack of analytical solutions of the model. We simplify the March model to take an analytical approach using master equations. We then derive exact solutions for some of the simplest yet nontrivial cases, and perform numerical estimation of master equations for more complicated cases. Both analytical and numerical results are in good agreement with agent-based simulations. These results are also compared to those of the original March model. Our approach enables us to rigorously understand the results of the simplified model as well as the original model, to a large extent. Published by the American Physical Society 2024