Purpose: This paper aims to analyze the macroeconomic effects of a monetary policy shock considering that fiscal policy is under fiscal constraints. For that, a dynamic stochastic general equilibrium (DSGE) model was developed for Brazil, which was estimated through Bayesian econometrics. Design/methodology/approach: In the basic model, the government does not have any type of fiscal restriction. The other two estimated models, however, consider that the fiscal authority implements some kind of fiscal rule. One of these rules is the Constitutional Amendment 95/2016 (EA 95/2016), which includes a limitation for government spending. The other Alternative Rule seeks to represent the characteristics of a more austere fiscal rule, as proposed by Wesselbaum (2017). Findings: It was possible to verify in this paper that the implementation of EA 95/2016 by the Brazilian government does not produce statistically different results and that it reduces the welfare of the households in relation to the scenario without fiscal rule. Thus, the proportionate benefit of EA 95/2016 is less than the cost associated with this fiscal rule (less welfare). If the government adopts a fiscal constraint similar to the Alternative Rule, it is possible to considerably reduce the interaction between fiscal and monetary policy, thereby reducing the fiscal dominance policy over monetary policy. However, the cost in terms of welfare is much higher than the baseline scenario. Thus, the fiscal authority is subject to a trade-off among public debt stabilization and household welfare. Originality/value: The study intends to contribute to the literature on three specific points. First, the monetary–fiscal policy interaction within a representative model of the Brazilian economy is discussed. In addition, the study considers that the government can adopt EA 95/2016 and the Alternative Rule, used in the US economy. Second, the impacts of EA 95/2016 and the Alternative Rule on household welfare will be quantified. Finally, two types of individuals (Ricardian and non-Ricardian agents) and two sectors of production (wholesalers and retailers) are considered. In this paper, the DSGE model is estimated, since the previously mentioned authors performed simulations [ABSTRACT FROM AUTHOR]