Fiscal Policy Multipliers in a New Keynesian Model under Positive and Zero Nominal Interest Rate

This paper uses a simple new-Keynesian model (with and without capital) and calculates multipliers of four types. That is, we assume either an increase in government spending or a cut in sales/labor/capital tax that is financed by lump-sum taxes (Ricardian evidence holds). We argue that multipliers of a temporary fiscal stimulus for separable preferences and zero nominal interest rate results in lower values than what is obtained by Eggertsson (2010). Using Christiano et al. (2009) non-separable utility framework which they used to calculate spending multipliers we study tax cuts as well and find that sales tax cut multiplier can be well above one (joint with government spending) when zero lower bound on nominal interest binds. In case of a permanent stimulus we show in the model without capital and assuming non-separable preferences that it is the spending and wage tax cut which produce the highest multipliers with values lower than one. In the model with capital and assuming that the nominal rate is fixed for a one-year (or two-year) duration we present an impact multiplier of government spending that is very close to the one in Bernstein and Romer (2009) but later declines with horizon in contrast to their finding and in line with the one of Cogan et al. (2010). We also demonstrate that the long-run spending multiplier calculated similarly to Campolmi et al. (2010) implies roughly the same value for both types of preferences for particular calibrations. For comparison, we also provide long-run multipliers using the method proposed by Uhlig (2010).