On Fiscal Multipliers: Estimates from a Medium Scale DSGE Model

This paper contributes to the debate on fiscal multipliers, in the context of a structural model. I estimate a micro-founded dynamic stochastic general equilibrium model, that features a rich fiscal policy block and a transmission mechanism for government spending shocks, using Bayesian techniques for US data. I find the multiplier for government spending to be 1.12, and the maximum impact is when the spending shock hits the economy. In addition, the estimated model predicts a positive but small response of private consumption to increased government spending. The multipliers for labor and capital tax on impact are 0.13 and 0.33, respectively. The effects of tax cuts, on the other hand, take time to build, and exceed the stimulative effects of higher spending at horizons of 12-20 quarters. The expansionary effects of tax cuts are primarily driven by the response of investment. I carry out several counterfactual exercises to show how alternative financing methods and expected monetary policy have consequences for the size of fiscal multipliers. I also simulate the impact of the American Recovery and Reinvestment Act of 2009 in the estimated model.