New Principles For Stabilization Policy

In a broad class of discrete-time rational-expectations models, I consider stabilization-policy rules making the policy instrument react with coefficient φ ∈ R to a (past, current, or expected future) variable at horizon h ∈ Z, possibly among other variables, possibly with inertia. Using two complex-analysis theorems, I establish analytically some simple, easily interpretable, necessary or sufficient conditions on φ and h for these rules to ensure local-equilibrium determinacy. These conditions lead to new, general principles for stabilization policy in terms of whether, and how strongly or weakly, to react to any variable, at any horizon, in any model, with any policy instrument. Building on these conditions, I characterize the scope of validity of (a generalized version of ) the long-run Taylor principle as a condition for determinacy. I apply all these results to standard interest-rate rules in 134 quantitative monetary-policy models, and find the new principles to be (either typically or occasionally) quantitatively relevant.