Macroeconomic Effects of Sectoral Shocksin Germany, The U.K. and, The U.S.: A VAR-GARCH-M Approach

Abstract

A VAR-GARCH-M model for aggregate employment and employment shares is developed to explore the macroeconomic effects of sectoral shocks. Using U.S., U.K. and German data, three main issues are investigated: the relevance of shocks volatility; the amount of aggregate employment growth variation accounted for by re-allocation shocks and the amount of aggregate innovation volatility explained by sectoral components. Bayesian methods are used for estimation model selection and innovation accounting – Bayes factors for model selection and MCMC for estimation. The results favor the VAR-GARCH-M model. A significant GARCH-M component indicates the presence of volatility clustering and the feedback of volatilities on aggregate employment and sectoral shares growth rates. The innovation analysis supports sectoral shocks as a triggering force for aggregate employment fluctuations. In all three countries, 45% to 55% of aggregate employment variation is accounted for by sectoral innovations.