The Fed made its first mistake in 2021. It misread a supply shock as a transitory blip, fell behind, and spent two years engineering the steepest tightening campaign since Paul Volcker's early-1980s crackdown. Now, with April CPI at 3.8%, producer prices up 6.0% year over year, and fed funds futures pricing a meaningful chance of a rate hike before year-end, there is a real risk the institution overcorrects in the opposite direction and makes the second mistake to go with the first.
Inflation is real. The gasoline bill is real. The food shock is real. But the boogeyman -- the 1970s-style spiral feeding itself through wages and expectations -- is not in the April data.
The April producer price index did not calm those fears. Headline PPI for final demand jumped 1.4% in April, the largest monthly increase since March 2022, and now stands 6.0% above a year earlier. Excluding food, energy, and trade services, the underlying index rose 0.6% on the month and 4.4% over the past year, the firmest reading since February 2023. Pipeline inflation hot enough to light up screens and lift hike odds.
The mistake would be tightening into a supply shock that has no wage-price engine behind it, and treating a single month of hotter CPI and PPI as proof that one has returned. The April numbers are real. The spiral they are being taken to imply is not.
Much has been said about whether the April spike marks the inflation comeback. Less has been said about how it reads under Kevin Warsh's own preferred framework. Warsh, just confirmed as Fed Chair in the narrowest vote of the modern era, has argued that trimmed-mean and median measures do a better job isolating the broad trend in prices. They throw out the biggest price increases and the biggest price decreases each month rather than just excluding food and energy by category. By that standard, underlying inflation still looks far cooler than the headlines.
The Dallas Fed's trimmed-mean PCE -- the personal-consumption-expenditures price index, the Fed's preferred inflation gauge -- was 2.4% over the 12 months through March, versus 3.5% for headline PCE and 3.2% for core PCE. On the measure that cuts off the biggest price moves at both ends and focuses on the broad middle, inflation is already much closer to the Fed's 2% goal than the headlines suggest. If Warsh wants the market and the committee to look through noisy shocks, his own preferred gauge says they should.
Dallas staff has warned that its own measure can underweight tariff-driven price spikes when extreme price increases pile up at the top of the distribution, as they have now. Fair point. But the gap between the trimmed mean and headline is so wide that even adjusting for the tariff bias leaves the underlying trend well below where the headlines sit.
Powell made this case at Harvard on March 30, before April CPI scrambled the pricing. He emphasized that the Fed should not overreact to short-lived supply shocks and should focus instead on broader, slower-moving measures of underlying trend. Year-end hike odds dropped from better than 50% to 2.2% the day of his remarks. The framing was correct then. A single month does not change it.
The Loop Is Not Running
The 2021 to 2022 inflation surge was not just about supply chains and oil. It was turbocharged by a labor market where employers were begging for workers. They bid up wages. Businesses passed those costs to consumers. Workers demanded even higher wages to keep up.
That self-reinforcing loop is what made the period so hard to break and so expensive to stop. A spiral needs a mechanism -- workers demanding raises, employers granting them, prices passing through. Today's data show that engine sitting idle.
Wage growth is moderate and largely sideways. The Employment Cost Index, the BLS measure of labor costs, sits at 3.4% annually, well off the 5.7% pandemic peak. The Atlanta Fed's Wage Growth Tracker, which follows the same workers' pay over time, runs at 3.9%, down from the 6.7% peak of 2022. Unit labor costs rose only 1.2% in the first quarter, with productivity gains absorbing most of the wage increases.
NFIB reports only 18% of small business owners plan to raise compensation, well below the 32% peak of 2022. That forward-looking signal is the right one. There is no sign of employers gearing up for another wage chase.
The institutional channel is missing. In 1973, COLA clauses indexed union wages to inflation automatically, and pattern-setting contracts at the UAW and Steelworkers pulled the rest of the labor market with them. Today, private-sector union density is around 5%.
COLA clauses are rare. Workers can ask. The infrastructure that forced employers to grant the raise is gone.
The leverage channel is missing too. Unemployment is 4.3% and has hovered near multi-year highs. Hiring slowed in February to its lowest pace since 2011 outside the pandemic. Job openings remain well below 2024 highs. That is not a labor market where workers demand double-digit raises and credibly threaten to walk.
What is happening instead is demand destruction. Real average hourly earnings fell 0.5% in April and 0.3% year over year. Households filling 55 gallons a month against gasoline up more than a dollar year over year do not strike for raises. They cut spending.
That is disinflationary at the margin, not the first step in a spiral. It is cold comfort for families feeling the squeeze, but it matters for the inflation story.
The one channel worth watching is inflation expectations. The University of Michigan one-year measure eased to 4.5% in May from a seven-month high of 4.7% in April. The 5-year, 5-year forward breakeven, the market's bet on average inflation from five to ten years out, sits near 2.25%. Short-term expectations are easing, and long-term gauges remain anchored. Exactly the pattern of a transitory shock, not a regime change.
None of this means an energy and food shock can never turn into something worse. If gasoline and grocery prices stayed elevated long enough to push expectations up and embolden wage demands, the risk profile would change. The Fed's calculus would have to change with it. The point is that today's wage data, labor leverage, and expectations are not behaving that way. The loop that made 2021 to 2022 so dangerous is not running.
The April Numbers Have an Explanation
The April CPI was driven by energy. Energy prices rose 17.9% year over year as the U.S.-Israeli war with Iran that began in late February choked off the Strait of Hormuz and pushed Brent crude up sharply through the spring. Food compounded the pain. The food-at-home index jumped 0.7% in a single month, with fresh produce hit hardest by a convergence of weather, tariffs, and transportation costs.
Even on core CPI, the 12-month rate sits at 2.8%, well below the 3.8% headline. That gap is the supply-shock signature.
The April PPI tells the same story one step earlier in the pipeline. Final demand prices rose 1.4% in April; nearly 60% came from services, led by trade margins as firms moved earlier cost increases onto buyers. Prices for final demand goods rose 2.0%, with roughly three-quarters reflecting a 7.8% surge in energy. Goods excluding food and energy rose a comparatively modest 0.7%. An energy shock in the wholesale data, not a generalized cost spiral.
Warsh's trimmed-mean approach points the same way. A trimmed mean does not pretend the shocks are unreal. It asks whether they are broad enough to define the underlying trend. Right now the answer is no. Extreme price moves are hot, but the broad center of the distribution still looks much calmer.
Durable goods prices were essentially flat in CPI, and the non-energy portion of PPI moved only modestly. That is not how entrenched, broad-based inflation behaves. It is how a supply shock behaves.
Airline fares climbed 2.8% on the month, and postage and delivery rose. Both are real. Carriers and shippers are passing through higher energy costs, not signaling that something permanent has taken hold.
The Rent Spike Is a Statistical Artifact
The shelter index appeared to surge in April, and some will read it as evidence of sticky services inflation. They should not. The Bureau of Labor Statistics could not collect rent data during the 43-day government shutdown last fall and assumed zero shelter inflation for October. Because BLS recontacts the same units every six months, the missing October data showed up as a single April catch-up, not as a sudden turn in the rental market.
Market rents are moderating on national measures. In some metro areas, rents have ticked higher again, but the broad indexes landlords and tenants actually face are still cooler than the CPI shelter line suggests. The catch-up will reverse. This is a measurement artifact catching up to the real world, not a new burst of housing-driven inflation.
The Committee Is Not a Hike Committee
The April FOMC vote, Powell's last as Chair, split 8-4, the most dissents since 1992. But the nature of those dissents matters more than the count. Three of the four dissenters, Hammack of Cleveland, Kashkari of Minneapolis, and Logan of Dallas, did not push for a hike. They pushed against language in the statement that hinted at coming cuts: the phrase "additional adjustments," which in Fed-speak means cuts are next. The fourth, Governor Miran, dissented in the other direction, wanting a quarter-point cut.
They wanted a symmetrical statement leaving the next move as either a cut or a hike depending on incoming data. That is a call for optionality. It is not a call to tighten.
Three of the four 2026 rotating regional bank presidents lean hawkish by recent standards. Paulson of Philadelphia is the exception. She has been consistently dovish on labor-market risks and supported the rate cuts of late 2025.
The seven governors vote permanently, and they tilt the other way. The governors have shown more concern about labor-market fragility than about a single hot month.
Taken together, this is a hold committee with three regional presidents pushing for two-sided language and one governor pushing for a cut. The futures market is putting real odds on another hike. The committee is not signaling one.
Rate hikes cool demand. They can slow hiring and spending, and they bite especially hard when real wages are already under pressure. They do not fix oil disruptions or crop failures. Applied to a shock with no wage-price engine behind it, they slow an economy already softening. They squeeze the real wages of workers already losing ground at the pump.
And again, the irony is that Warsh's Fed will be hard-pressed to justify a hike on Warsh's own metric. If trimmed-mean PCE is running at 2.4%, then the broad center of the inflation distribution is already much closer to target than headline CPI or PPI suggests. Hiking into an energy shock under those conditions would not look data-driven. It would look reflexive.
The Fed's Own Officials Understand This
Senior Fed officials have already laid out the case for patience. Waller has said he can look through energy-driven inflation, knowing it will unwind, and remains more inclined toward cuts to support the labor market than toward hikes. Daly's April 3 blog post argued the speed limit of the labor market has changed and the Fed should not read low job growth as weakness. She carried that case into the May 8 Hoover Institution panel. Bowman has flagged underlying weakness in the labor market as a near-term risk.
Daly has also made the immigration argument explicitly and on the record. San Francisco Fed research finds net immigration collapsed from 2.2 million in 2024 to roughly half a million in 2025, with the Dallas Fed estimating an outright net outflow of unauthorized workers in the second half of 2025. The labor force is barely growing.
By some Fed estimates, the economy now needs fewer than 10,000 new jobs per month just to hold unemployment steady. Daly has said publicly that a zero or even slightly negative month of job growth could be consistent with full employment. That is not a sign of a tight labor market generating inflationary wage pressure. It is a sign of a structurally smaller workforce.
Waller made the same immigration point in an April speech, calling the shift in labor-force growth unprecedented in recent history. This is not a fringe view. It is coming from a Fed governor and a Reserve Bank president, voiced openly in the weeks before markets priced in a hawkish reaction to a single number.
In a labor market that needs very few new jobs to keep unemployment steady, extra tightening bites faster than the old rules of thumb assume. A rate move that ten years ago would have been a tap on the brakes is now a harder squeeze.
The Market Is Pricing a Reflex, Not a Forecast
The CME FedWatch tool is still reacting mechanically to the latest inflation headlines rather than weighing the underlying story. After the April CPI release on May 12, futures put hike odds at roughly 30% by year-end. The PPI release the next day pushed odds toward 40%. By Friday morning they sat just over 50%. That is what those markets are built to do: mechanically reprice as each number hits the tape, not a verdict that the inflation picture has changed.
The analysts who forecast inflation's actual path are telling a different story. Pantheon Macroeconomics noted core CPI's strength in April came largely from one-time factors: the shutdown rent catch-up, airline fare passthrough from higher fuel, and a Netflix price increase. Bank of America, among Wall Street's most hawkish, pushed its first cut to the second half of 2027, expecting inflation near target by then.
Goldman Sachs expects PCE to hover near 3% through 2026. Deloitte sees it easing to 2.1% in 2027. None are forecasting a permanent overshoot.
Forecast models have badly underestimated inflation persistence since 2021, so no one should treat any single house's path as gospel. The point is narrower: even the more hawkish shops now see inflation drifting back toward target over the next couple of years, not exploding into a new 1970s. The futures market is pricing the chance that the Fed reacts more aggressively to a string of hot prints. The forecasters are betting that the underlying drivers will not justify that response.
Brent crude has retreated from a wartime peak above $125 to the $105 to $110 range. If it sat near current levels for several years, with gasoline and diesel pinned where they are now, at some point it would show up in higher wage demands and stickier expectations. What matters for policy is not that prices are high today, but whether wages, expectations, and margins are behaving as if they will stay high indefinitely.
So far, wages are cooling, long-term expectations are anchored near 2.25%, and margins outside energy are not exploding. That is not how embedded energy inflation behaves. The rent catch-up from the shutdown will reverse. This is what a shock unwinding looks like.
The Second Mistake
The lesson the Fed took from 2021 was reasonable: do not misread the signal again. That lesson, hardened into reflex, produces its own error. Misreading a supply shock as a spiral and tightening accordingly is not proof the institution learned from 2021. It is proof it learned the wrong thing.
The pump pain is real. The grocery shock is real. The household squeeze is real. But on the measures Warsh himself has championed, the underlying trend is far less alarming than the headlines.
The first mistake was misreading the signal. The second would be overreading it.
No boogeyman here.
