Monetary policy indeterminacy in the U.S.: results from a classical test

We work with a newly developed method to empirically assess whether a specified new-Keynesian business cycle monetary model estimated with U.S. quarterly data is consistent with a unique equilibrium or multiple equilibria under rational expectations. We conduct classical tests to verify if the structural model is correctly specified. Conditional on a positive answer, we formally assess if such model is either consistent with a unique equilibrium or with indeterminacy. Importantly, our full-system approach requires neither the use of prior distributions nor that of nonstandard inference. The case of an indeterminate equilibrium in the pre-1984 sample and of a determinate equilibrium in the post-1984 sample is favored by the data. The long-run coefficients on inflation and the output gap in the monetary policy rule are found to be weakly identified. However, our results are further supported by a proposed identification-robust indicator of indeterminacy