I’ve been working on refinements and extensions of the Goodwin (1967) model of the growth cycle (here is a short set of notes on it: ECON705_Goodwin) for a while. The Goodwin model is a very elegant, stylized framework representing the dynamics of employment (v) and income shares (omega) in a growing economy where capital accumulation is driven by savings out of profit income earned by asset-owners (capitalists). Goodwin argued that his model was an attempt to formalize the dynamics of accumulation and the reserve army of labor described by Marx.
The key point is that the model generates endogenous cycles, that is fluctuations that are independent of shocks (such as monetary policy, fiscal policy, or supply shocks). Employment and distribution move counterclockwise around a medium-run equilibrium. The story goes as follows: a low labor share means higher profitability and therefore faster accumulation by firms. Investment in new capital stock increases employment, and after some point the labor market will become tight enough that real wages will rise (expansion phase). At this point, firms will respond to further wage increases by cutting down on labor demand. Slackness in the labor market (that is, high unemployment) will create downward pressure on real wages (recession phase), until profitability is restored and the cycle can pick up again.
Although Goodwin did not think his model would provide an accurate representation of actual cycles, it is remarkable that Goodwin-type cycles can be found in the data. Below, I plot HP filtered quarterly series of the labor share (2001=100) in the non financial business sector and the employment rate in the United States (data after 2005 are altered by the filtering procedure and the peculiarity of the 2008 crisis, and therefore I left them out).
It is important to point out contemporary macroeconomics -both RBC and New Keynesian, whose benchmark is the Neoclassical growth model (NGM) of Cass and Koopmans in its stochastic version by Brock and Mirman- cannot make sense of these cycles endogenously, but it requires exogenous shocks to generate fluctuations.
There are serious shortcomings in the Goodwin model (I will talk about them in later posts), but the fact that this simple setup can make sense in an intuitive way of economic fluctuations in a growing economy makes it, to me, a compelling reason to take it seriously, refine it and improve it.