Fiscal stimulus and monetary and fiscal policy interactions in Brazil

Abstract

This article estimates the dynamic impacts of a fiscal stimulus on Brazilian macroeconomic variables between 1999 and 2017. It adopts the Dynamic Stochastic General Equilibrium (DSGE) model, with Markov switching regimes in monetary and fiscal policies, developed by Davig and Leeper (2011). In the model, the expectations of economic agents are endogenously affected by the existence of – and probability of switching – regimes in fiscal and monetary policies, and there are price rigidity and monopolistic competition. The probability of occurrence of regimes and the parameters of the Brazilian monetary and fiscal policy rules, used in the calibration of the chosen DSGE model, were estimated through a Markov-switching structural bayesian vector autoregression model (MS-SBVAR). In the period considered, is quite robust the evidence that the monetary policy was always active with changing intensity in its degree of activity. There is also a weaker evidence that the fiscal policy was always passive. There is a marked alteration, in the different regimes, of the response of both monetary and fiscal policies to the output gap. An unanticipated shock in government spending has a multiplier inferior to one and causes, in all combinations of the two identified regimes in the monetary and fiscal policies, an increase in the output gap and in the real interest rates, a reduction in private consumption and inflation, and an increase in debt and taxes. The results about the inflation rate response, are strongly dependent on the response of monetary policy to the output gap. In one of the monetary policy regimes there is strong uncertainty about this response to the output gap, and it cannot be rejected that it is zero. In the combination of regimes where the response to the output gap is imposed to be zero, there is a positive response of the inflation rate to a positive fiscal innovation in government spending.