This paper examines the role of adaptive learning expectations in rationalizing stronger monetary policy responses in Brazil. The authors estimate a Dynamic Stochastic General Equilibrium (DSGE) model with different expectation formation processes for Brazil and the US, including a standard rational expectation process and an adaptive learning model. The inclusion of a labor market in the model helps to anchor inflation even in a situation of adaptive expectations, a positive output gap, and inflation above target. The estimation results show that the adaptive learning model does a better job in fitting the data in both Brazil and the US.