Tax multipliers and monetary policy: Evidence from a threshold model

Romer and Romer (2010) use the narrative record to generate a time series of exogenous shocks to fiscal policy. They report a tax multiplier of 3.0. We extend their analysis and allow for nonlinearities between their shocks and the effects on output by estimating a threshold regression model. Using Hansen’s (1997) procedure, we find the best fitting threshold is changes in the federal fund rate with a delay of two quarters. Moreover, we find that the tax multiplier is approximately 4.3 if accompanied by an accommodative monetary policy and approximately 1.2 under tight monetary policy.