Evidence on Goodwin cycles across US golden age and neoliberal era

This paper provides a set of baseline empirical results on Goodwin cycles for the US post-war macroeconomy. Our sample consists of Hamilton-filtered series for nonfarm business sector output and labor share as well as the employment rate. Vector autoregressions (VARs) yield strong evidence for the Goodwin mechanism, i.e. profit-led activity and profit-squeeze distribution, in two (output or employment rate cycle vs. labor share cycle) and three (output, employment rate and labor share cycles) dimensions. We focus solely on the cycle rather than underlying trends, discuss how theory and empirical methods relate, and investigate whether and how the Goodwin mechanism has changed over time. To do so, we estimate split samples with standard recursive VARs for the 'golden age' (GA) and 'neoliberal era' (NE), as well as time-varying parameter VARs. Results indicate (i) overall a persistence of the Goodwin mechanism over the entire post-war period; but (ii) generally less pronounced cyclicality during NE than GA, and (iii) a weakening of the profit squeeze vis-`a-vis the employment rate over time as well as, on balance, a weakening of the profit led regime of both activity variables.