What are the Effects of Fiscal Policy Shocks?

Abstract

We investigate the effects of fiscal policy surprises for US data, using vector autoregressions. We overcome the difficulties that changes in fiscal policy may manifest themselves in variables other than fiscal variables first and that fiscal variables may respond "automatically" to business cycle conditions. We do so by using sign restrictions on the impulse responses as method of identification, extending Uhlig (1997), and by imposing orthogonality to business cycle shocks and monetary policy shocks. We find that controlling for the business cycle shock is important, but controlling for the monetary policy shock is not, that government spending shocks crowd out both residential and non-residential investment but do not reduce consumption, that a deficit spending cut stimulates the economy for the first 4 quarters but has a low median multiplier of 0.5, and that a surprise tax increase has a contractionary effect on output, consumption and investment. Our results differ from the benchmarks of Ricardian equivalence and tax smoothing, and are more in line with theories which allow for intergenerational redistribution with limits to the compensating effects of bequests. The best fiscal policy for stimulating the economy appears to be a deficit-financed tax cut.

Keywords: Fiscal Policy, Vector Autoregression, Bayesian Econometrics, Agnostic identification

JEL Classification: C32, E60, E62, H20, H50, H60