Abstract

Rational expectations (RE) has been dominant both in the economic literature and in the macromodels routinely used in central banks. The RE assumption has recently come under attack as one of the drawbacks of the Dynamic Stochastic General Equilibrium (DSGE modeling) paradigm. This study attempts to investigate whether other ways of modeling expectations would necessarily find a better support in the data. 

We investigate the relevance of the RE assumption by introducing regime switching into the expectations formation of an otherwise standard DSGE model by Justiniano and Preston (2010). In our model, expectations switch between RE and Adaptive expectations (AE). The model is estimated on UK data using Bayesian techniques.

By introducing a switching mechanism, the model explains the data better than both the pure RE and the pure AE models. Expectation formation switches to AE during changes in monetary policy and the financial crisis. 

The dynamics of the economic system is different under the two expectation regimes. Hence, should the UK economy switch to an AE regime after Brexit, or as a consequence of the COVID-19 pandemic, the dynamics of the economic system could be substantially more uncertain than under RE, given the model.