Estimating a small DSGE model under rational and measured expectations : some comparisons

Using European panel data and GMM system estimation, we explore the empirical performance of the standard three-equation New Keynesian macro model under different informational assumptions. As a benchmark, we consider the performance of the model under rational expectations and revised (final) data. Alternatively, instead of imposing rational expectations hypothesis we use realtime information, ie Consensus Economics survey data, to generate empirical proxies for expectations in the model and the current output gap in the Taylor rule. We demonstrate that, contrary to the assumption of rational expectations, the errors in measured expectations and real-time current output gaps are positively autocorrelated. We produce evidence that the use of real-time variables (including measured expectations) improves the empirical performance of the New Keynesian model. Relaxation of the rational expectations hypothesis makes a noticeable difference for the parameters of the New Keynesian model, especially in the Taylor rule. Keywords: DSGE model, survey expectations, GMM system estimation, expectations, estimation JEL classification numbers: C52, E52, E20