A Keynes-Kalecki Model of Cyclical Growth with Agent-Based Features

Throughout his career, Malcolm Sawyer has both encouraged and contributed to the development of a Kaleckian alternative to conventional macroeconomic theory. In the spirit of this endeavour, we construct a Keynes-Kalecki model of cyclical growth with agent-based features. Our model is driven by heterogeneous firms who, confronting an environment of fundamental uncertainty, revise their “state of long run expectations” in response to recent events. Model simulations generate fluctuations in the rate of growth that are aperiodic and of variable amplitude. We also study the size distribution of firms resulting from our simulations, finding evidence of a power law distribution that we have no reason to anticipate from the basic structure of our model. Finally, we reflect on the potential advantages of combining aggregate structural modelling with some of the methods and practices of agent-based computational economics.