The Asymmetric Effect of Government Spending in the United States

Abstract

In this paper, we test for the presence of asymmetric effects of government spending on output and employment rate using U.S. postwar states panel data. We estimate two “endogenous” asymmetry: (i) whether negative and positive government spending change has different multiplier impact on real output and employment; (ii) whether big or small change has different effects. Our empirical results show there are asymmetric effects in terms of positive and negative government spending, but the asymmetric effect of positive and negative government spending can vary for different dependent variables and there is no significant asymmetric effect on employment. For the asymmetry of big and small government spending, we observe the same asymmetric property for output and employment. In both cases, the differences of the effect of big and small change are non-significant. Our main results are robust to intersection of “endogenous” factors effects, alternative specification (or other additional controls) and longer change interval, but asymmetric effect might differ among different production sectors.