Small-sample inference in rational expectations models with persistent data

This paper shows that, in rational expectations (REs) models, accurate small-sample inference in the presence of persistent variables can be explicitly derived using the mapping between the rational expectations hypothesis (REH) restrictions and the parameters on the stationary and persistent variables. An application to the New Keynesian Phillips curve (NKPC) demonstrates that the persistence in the data leads to large size distortion and power loss of commonly-used tests.