Quantitative evidence of Goodwin's non-linear growth cycles

Abstract

The article searches for quantitative evidence in favor of an extended version of Goodwins predator-prey model of endogenous distribution-employment cycles for 16 countries. The model is extended to include several harmonics in the workers' share of national income and the employment rate. The model fits all the observations both within sample and out of sample for half of the countries, with a 10% error, and for 4 countries with a 5% error. This suggests that the model may be useful to investigate the behavior of dynamical economic systems.