Do learning models of expectations formation capture actual expectations? In this study we
address this question by using data from the Survey of Professional Forecasters in the
estimation of a dynamic stochastic general equilibrium model under learning. Exploiting the
moments of this data allows us to identify the parameters of the model better, particularly
those related with the formation of the expectations. We find that, without the use of survey
data, inflation expectations estimated under the assumption of learning capture some features
of the data not related with actual expectations, such as changes in the trend of inflation. Once
surveys are included, learning not only matches actual expectations but also emerges as a key
determinant of inflation persistence, explaining around 30 percent, while exogenous shocks
only account for approximately 15 percent.