Description:
We estimate the fiscal (spending) multiplier using quarterly U.S. data, 1986-2017. We define government spending shocks as actual minus expected expenditure growth, the latter obtained from the Survey of Professional Forecasters. We employ the ST-VAR model with the local projections method. A key testable conjecture is that the effects of positive and negative spending shocks have numerically different effects (the latter being stronger). Although we cannot formally reject the null of equality, the conjecture does hold in general. We also find evidence of state-dependence of multipliers as previously pointed out.