Recent research finds that the U.S. fiscal stance on debt is subject to unobserved and recurring changes that can affect inflation. How should monetary policy be conducted in the presence of recurring fiscal regime change? We study optimal interest rate rules in economies with recurring active fiscal regimes, and describe our results using a generalization of the Leeper (1991) conditions. Many empirically-relevant fiscal policies call for rules that do not depend on the current fiscal regime, including interest rate pegs. Still, we find that in some empirically-relevant cases, central banks will lose control of inflation unless the monetary stance tracks fiscal regime changes. These results hold under rational expectations, and in a novel adaptive learning model with unobserved policy regimes.
Does my model predict a forward guidance puzzle? (with Christopher G. Gibbs)
We provide sufficient conditions for when a rational expectations structural model predicts bounded responses of endogenous variables to forward guidance announcements. The conditions coincide with a special case of the well-known (E)xpectational-stability conditions that govern when agents can learn a Rational Expectations Equilibrium. Importantly, we show that the conditions are distinct from the determinacy conditions. We show how the conditions are useful for diagnosing the features of a model that contribute to the Forward Guidance Puzzle and reveal how to construct well-behaved forward guidance predictions in standard medium-scale DSGE models.
This paper addresses the relationship between determinacy and E-stability in a general class of Markov-switching DSGE (MS-DSGE) models with lagged endogenous variables. We prove that determinacy conditions from Cho (2016) and Cho (2019) imply the E-stability of the unique solution when agents condition their expectations on contemporaneous variables and use one-step-ahead decision rules. We therefore extend the main result of McCallum (2007) to a class of DSGE models with time-varying parameters. As with linear DSGE models, E-stability conditions are weaker than determinacy conditions, but we show that MS-DSGE models present new cases where indeterminate models feature E-stable MSV solutions. In particular, we employ a New Keynesian model with recurring exogenous interest rate regimes and show that indeterminacy can obtain when our E-stability condition, which coincides with the Long Run Taylor Principle, is satisfied.
New Keynesian models predict implausible responses to anticipated structural changes when interest rates are exogenous, a phenomenon called the forward guidance puzzle. We develop tractable necessary and sufficient conditions for when a model exhibits a forward guidance puzzle, and apply these conditions to characterize restrictions on fiscal policy that eliminate forward guidance puzzles. We show that permanent or recurring active fiscal policy regimes can severely dampen equilibrium responses to forward guidance when interest rates are pegged. Moreover, uncertainty about future fiscal policy by itself may severely dampen equilibrium responses to forward guidance.
Maturity, Determinacy and Policy Regimes (under revision)
This paper explores some determinacy properties of a New Keynesian model with recurring fiscal and monetary policy regimes. We find that the maturity structure of government debt matters for determinacy and the existence of stable equilibria in the switching model, which is not true in standard monetary models with fixed policy regimes. In particular, central banks can typically embrace more overall active monetary policy stances in models with recurring active fiscal policy regimes when the maturity structure of debt held by households is longer. We argue that this gives central banks an incentive to twist the maturity structure of debt in order to choose interest rate rules that more effectively stabilize inflation.
Work in Progress
Heterogenous Learning and the Value of Intergenerational Knowledge Transfers (with Tristan Nighswander)
Expectational Stability and Interest Rate Pegs (coming soon)
Adaptive Learning in Hidden Markov Models