A plain-English guide to the Taylor rule.

The Taylor rule is the most-cited policy benchmark in monetary economics. It was first published by John Taylor in 1993 as a description of what the Federal Reserve had been doing through the 1980s; it has since become both a descriptive yardstick (what is the rule recommending today?) and a normative one (should the Fed follow it?). With today's 3.4 percent core PCE print and Warsh's published views that future framework documents may make the rule more explicit, this is the explainer on what the rule actually is, the parameter choices that matter, what it recommends today, and where the Committee deviates and why.

The rule in one line

i = r* + π + 0.5(π − 2) + 0.5(y − y*)

Where:

• i = the nominal policy rate (the Fed funds target). What the rule recommends the Fed sets.
• r* = the real neutral rate. The real interest rate consistent with full employment and stable inflation. We cover r-star in a separate reference piece; current estimates cluster between 1.0 percent (Laubach-Williams) and 2.0 percent (Lubik-Matthes).
• π = current inflation. Taylor originally used CPI; current academic and Fed practice uses core PCE.
• π − 2 = the inflation gap. Difference between current inflation and the Fed's 2 percent target.
• y − y* = the output gap. Difference between current GDP and potential GDP. Negative when the economy is below capacity; positive when running hot. The CBO publishes the canonical output-gap series.

Plug in numbers. With r* at 1.0, core PCE at 3.4, and an output gap near zero, the rule recommends:

i = 1.0 + 3.4 + 0.5(3.4 − 2) + 0.5(0) = 5.1 percent

The current FF target midpoint is 3.625 percent. The rule recommends roughly 150 basis points higher than current policy. The Fed is below the rule's recommendation, which in Taylor's terms is "accommodative".

The four parameter choices that matter

1. The inflation measure. Headline CPI, core CPI, headline PCE, core PCE. Each produces a different recommendation. The 2026 cycle has had headline CPI run roughly 60 basis points above core PCE; that gap maps to a 90 basis-point gap in the rule's recommended rate (because of the 0.5 weight on the inflation gap).
2. The output-gap measure. CBO output gap, Federal Reserve Board real-time estimate, Atlanta Fed GDPNow-derived gap, unemployment-rate-based estimate. Each carries different real-time noise. Most central banks use the unemployment-rate-based version because it is published more frequently with less revision.
3. The neutral-rate estimate. A 100 basis-point r* difference (LW vs LM) translates directly to a 100 basis-point difference in the rule's recommendation.
4. The reaction coefficients. Taylor's original specification used 0.5 weights on both gaps. The "Taylor 1999" specification doubled the output-gap weight to 1.0. Janet Yellen's preferred version added a higher unemployment-gap weight. Each version produces different prescriptions.

What the rule has recommended over time

The 2022 "behind the curve" episode

The Taylor rule's most famous moment in the 2020s was 2022. With core PCE at 5.4 percent and the FF target at 1.5 percent, the rule recommended a rate of roughly 5.5 percent: a 400-basis-point gap. The forward guidance the Committee had committed under FAIT constrained the Committee's ability to catch up; the "behind the curve" criticism became politically and academically acute.

The 2022 episode is the cleanest illustration of why rule-based policy and discretion-based policy disagree. A pure Taylor-rule Committee would have hiked aggressively in late 2021 as inflation surprised; the actual Committee was constrained by FAIT's "make-up" framing and waited. The lag cost real credibility and contributed to the 2022 framework debate that Warsh has since led.

Why the Committee deviates from the rule

Three legitimate reasons to deviate, two illegitimate. The legitimate first:

1. The rule is mis-specified for the regime. The rule assumes inflation and output gaps adequately summarise the state of the economy. In a crisis (2008, 2020) the policy challenge is liquidity provision or financial-stability action, not inflation-output trade-off management. The rule does not apply.
2. The neutral rate is uncertain. A 100 basis-point uncertainty band on r* translates to a 100 basis-point uncertainty on the rule's recommendation. Operating in the centre of the band against a politically observed rate is reasonable.
3. Lags matter. The rule prescribes a contemporaneous response to contemporaneous data. Policy operates with a 12 to 18 month lag. A forward-looking Committee should deviate from the contemporaneous rule toward its forecast of where the rule will be 12 months from today.

The two illegitimate reasons (per the rule-based critique):

1. Political pressure. A Committee that avoids hikes because of political cost is operating sub-optimally. The Taylor critique of 2010s policy makes exactly this charge.
2. Optimism about r*. If the Committee systematically uses lower r* estimates than the data supports, the rule's recommended rate falls systematically, justifying easier policy. Warsh's published view (that r* is structurally higher than the Committee assumes) is a charge of exactly this kind.

What the rule recommends today, against the framework

With today's 3.4 percent core PCE and r* at 1.0 (Laubach-Williams), the rule recommends 5.1 percent. With r* at 2.0 (Lubik-Matthes, closer to Warsh's stated view), it recommends 6.1 percent. The actual FF target is 3.625. The Fed is 150 to 250 basis points below the rule depending on which r* is used.

Three reads on what this means:

• The Committee is constrained by transmission lags. The Fed expects inflation to fall toward 2 percent over the next 12 to 18 months (the SEP median PCE for 2027 is 2.5). A forward-looking rule prescribes a lower rate than the contemporaneous version because future inflation should be lower.
• The longer-run dot in the SEP partially captures this. The longer-run dot at 3.125 is implicitly the Committee's view on where the rule will eventually settle. The drift toward 3.25 in Warsh's hinted direction would partially close the gap.
• The political-economy critique applies modestly today. The Committee's failure to deliver the contemporaneous Taylor-rule rate is not entirely a forward-looking judgment; it is also a recognition that sharp hikes would have second-order costs. Warsh has published that the cost should be borne; the broader Committee currently disagrees.

Warsh's framework hint

Warsh's published research has argued that the next Fed framework should make a Taylor-rule benchmark explicit. The proposal: the SEP would include a Taylor-rule recommendation alongside the actual median dot, with deviations explicitly justified. This would force the Committee to publicly defend departures from the rule, which would discipline communication and anchor expectations.

The 2026 framework review is the venue. If Warsh delivers on this, the Taylor rule moves from a critic's yardstick to an institutional tool. The cross-asset implications would be substantial; OIS curves would price closer to rule-implied rates, and the real-yield differential against G10 would structurally widen.

What the rule does not tell you

1. The path. The rule prescribes a level, not a path. A Committee that follows the rule perfectly can still produce policy mistakes through bad timing.
2. The right r*. The single largest source of disagreement between rule-based prescriptions is the neutral-rate estimate. Different r* estimates produce different rules.
3. The cross-asset reaction. The market prices the Committee's actual reaction function, not a rule. A Committee that follows the rule is not the same as a Committee that the market expects to follow the rule.
4. The credibility cost of changing course. The rule does not capture the credibility tax that sharp rate adjustments impose. The 2022 lag was partly an attempt to avoid that tax.

Related reading

• Today's analysis applies this directly: Core PCE printed hot. The dollar didn't extend.
• The r-star piece: A plain-English guide to r-star
• The framework piece that mentions Taylor-rule benchmarking: A plain-English guide to monetary policy frameworks
• The forward-guidance piece on what the rule replaces: A plain-English guide to forward guidance