Time-varying endogenous productivity growth dynamics

This paper provides an empirical analysis of the dynamic determinants of US labor productivity growth by considering that the latter is an endogenous outcome, mainly influenced by changes in the size of the economy and relative labor costs. Specifically, we consider that changes in GDP, real wages, wages relative to the price of capital, and investment effect simultaneously the evolution of labor productivity growth. We focus on studying whether the response of the latter to these effects has been stable or time-varying by adopting a flexible hybrid time-varying parameter Bayesian vector autoregression with stochastic volatility empirical framework. This allows us identify whether none, some, or all lagged and contemporaneous coefficients in the equations in the model are constant or time-varying via model selection. We find: (i) evidence supporting the view of time-varying endogenous labor productivity growth dynamics; (ii) that the response of labor productivity growth to GDP growth has tended to increase over time; and (iii) that the response of labor productivity growth to real wage growth has tended to decrease over time. Our findings have important policy recommendations that can help to improve the future performance of labor productivity growth in the USA.