Sources universitaires et de recherche
Discretion versus Policy Rules in Practice
Taylor's original rule.
Taylor's original calibration uses coefficients of 0.5 on the inflation gap and 0.5 on the output gap. Accessed 2026-08-20.
Carnegie-Rochester Conference Series · 1993
The Science of Monetary Policy
Clarida, Gali, and Gertler on monetary-policy rules.
Journal of Economic Literature · 1999
Policy Rules and How Policymakers Use Them
Official comparison of Taylor, balanced-approach, ELB-adjusted, inertial, and first-difference rules.
Federal Reserve policymakers consult simple policy rules as benchmarks rather than mechanically following one formula. Accessed 2026-08-20.
Board of Governors of the Federal Reserve System · 2018
Monetary Policy Rules and Macroeconomic Stability: Evidence and Some Theory
Clarida, Gali, and Gertler on why responding more than one-for-one to inflation stabilizes it.
Quarterly Journal of Economics · 2000
An Historical Analysis of Monetary Policy Rules
Taylor's historical analysis compares U.S. policy episodes through interest-rate rules. Accessed 2026-08-20.
Taylor, John B. · 1998
Fondements de recherche
Éléments probants et données
Compare the conditional prescription with the observed policy rate using a documented data vintage, inflation measure, output-gap estimate, and neutral-rate estimate.
Calibrage ou mesure
The neutral rate, inflation target, current gaps, and response coefficients determine the prescription. On this graph, phi_y sets the line's slope while the other current inputs determine its vertical position.
Limites
- The rule calculation is static and does not forecast inflation or output.
- It does not identify monetary-policy shocks or solve a welfare problem.
- Neutral-rate and potential-output revisions can dominate the rule residual.
Conseils d’utilisation
Quand le modèle suffit
Calculating and decomposing a transparent policy-rate benchmark for a stated set of current inputs and coefficients.
Quand il reste une esquisse
Do not treat a rule residual as a causal shock or the displayed rate as an optimal policy result. Those claims require an identification design or a dynamic model with a stated objective.
Quand changer de modèle
Switch to forecast-targeting rules (Svensson) or to a fully estimated DSGE policy rule when forward-looking inputs matter. Switch to a financial-conditions rule when stress in credit markets is the binding constraint.
Signal de réfutation
Test any empirical fit claim out of sample after fixing the data vintage, gap measures, neutral-rate series, and coefficients. Persistent rate errors larger than a stated alternative reject that fit claim; revising unobserved inputs after seeing the errors makes the comparison circular.