A Model-Based Evaluation of the Debate on the Size of the Tax Multiplier

Abstract

The SVAR and narrative approaches to estimating tax multipliers deliver significantly different results. The former yields multipliers of about 1 percent, whereas the latter produces much larger multipliers of about 3 percent. The SVAR and narrative approaches differ along two important dimensions: the identification scheme and the reduced-form transmission mechanism. This paper uses a DSGE-model approach to evaluate the hypothesis that the different tax multipliers stemming from the SVAR and narrative approaches are due to differences in the assumed reduced-form transmission mechanisms. The main finding of the paper is that in the context of the DSGE model employed this hypothesis is rejected. Instead, the observed differences in estimated multipliers are due either to both models failing to identify the same tax shock, or to small-sample uncertainty.