Working Paper: NBER ID: w30763
Authors: Robert G. King; Yang K. Lu
Abstract: A parsimonious model of shifting policy regimes can simultaneously capture expected and actual US inflation during 1969-2005. Our model features a forward-looking New Keynesian Phillips curve and purposeful policymakers that can or cannot commit. Private sector learning about policymaker type leads to a reputation state variable. We use model inflation forecasting rules to extract state variables from SPF inflation forecasts. US inflation is tracked by optimal policy without commitment before 1981 and by optimal policy with commitment afterward. In theory and quantification, the interaction of private sector learning and optimal policy within regimes is central to expected and actual inflation.
Keywords: inflation; commitment; reputation; macroeconomic policy
JEL Codes: D82; E52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Type of monetary policy (without commitment) (E60) | Inflation outcomes (pre-1981) (E31) |
| Type of monetary policy (with commitment) (E63) | Inflation outcomes (post-1981) (E31) |
| Reputation of policymakers (D72) | Inflation bias (during the Great Inflation of the 1970s) (E31) |
| Evolution of reputation (Z13) | Serial correlation in inflation forecast errors (C22) |
| Evolution of reputation (Z13) | Inflation expectations (E31) |
| Private sector learning (J24) | Optimal policy (C61) |