A Proof of Chaos for a Seasonally Perturbed Version of Goodwin Growth Cycle Model: Linear and Nonlinear Formulations

Abstract

We show the existence of complex dynamics for a seasonally perturbed version of the Goodwin growth cycle model, both in its original formulation and for a modified formulation, encompassing nonlinear expressions of the real wage bargaining function and of the investment function. The need to deal with a modified formulation of the Goodwin model is connected with the economically sensible position of orbits, which have to lie in the unit square, in contrast to what occurs in the model’s original formulation. In proving the existence of chaos, we follow the seminal idea by Goodwin of studying forced models in economics. Namely, the original and the modified formulations of Goodwin model are described by Hamiltonian systems, characterized by the presence of a nonisochronous center, and the seasonal variation of the parameter, representing the ratio between capital and output, which is common to both frameworks, is empirically grounded. Hence, exploiting the periodic dependence on time of that model parameter we enter the framework of Linked Twist Maps. The topological results valid in this context allow us to prove that the Poincaré map, associated with the considered systems, is chaotic, focusing on sets that lie in the unit square, and also when dealing with the original version of the Goodwin model. Accordingly, the trademark features of chaos follow, such as sensitive dependence on initial conditions and positive topological entropy.