A Time-Varying Finance-Led Model for U.S. Business Cycles

Abstract

This paper empirically assesses predictions of Goodwin's cyclical growth model regarding demand and distributive regimes when integrating the real and financial sectors. In addition, it evaluates how the effects of distributive, demand, and financial shocks over six different U.S. business-cycle peaks. It identifies a parsimonious Time-Varying Vector Autoregressive model with Stochastic Volatility (TVP-VAR-SV) containing the labor share of income, the employment rate, residential investment, and the term spread as endogenous variables. Using Bayesian inference methods, key results suggest (i) a combination of profit-led demand and profit-squeeze distribution; (ii) weakening of these regimes after the Great Moderation years; and (iii) significant connections between the standard Goodwinian variables and residential investment as well as term spreads. Findings presented here broadly conform to the goal of reducing macroeconomic volatility after the 1980s.