St. Louis Fed Research: Core PCE Would Be Better by Excluding Only Energy Goods
N.R. Finch
St. Louis Fed adviser Fernando Martin argued on July 15 that core PCE should exclude only energy goods — not food and all energy — yielding a May headline PCE of 3.36%, well below the current 4.1% reading. The proposal feeds directly into the Fed's ongoing inflation-framework review and could reshape the benchmark behind rate decisions.
What does current core PCE actually strip out?
The current core PCE removes three categories: food at home, energy services (utilities), and energy goods (gasoline, etc.) — together about 13% of consumer spending.
This means → the Fed filters out nearly one-eighth of the consumer basket just to smooth volatility.
Martin's key finding: food and energy services are no more volatile than other consumer categories. Only energy goods — under 3% of the basket — swing wildly and track global oil prices.
Why is stripping only energy goods enough?
Martin compared his approach to a "scalpel" — cutting only the genuinely anomalous slice while keeping maximum spending data intact.
In plain terms = the current method bandages an entire hand because one finger is hurt; Martin says just bandage the finger.
The resulting measure is smoother and more representative, and "does not overreact to short-term shocks."
How large is the data gap?
Using the latest May data: Martin's revised headline PCE reads 3.36% versus the current 4.1%; his revised core reads 3.42%.
This means → under Martin's method the Fed would see a noticeably milder inflation picture, potentially shifting the timing of rate hikes or cuts.
Why is the "trimmed mean" approach under question?
Fed Chair Kevin Warsh has voiced support for the "trimmed mean" — a method that automatically strips the most volatile items each month.
Martin flagged two problems: common trimmed-mean calculations exclude up to 55% of the consumer basket, and at key turning points they show "significant lag."
He added that over the past 12 months trimmed-mean inflation "continued to decline rather than rise, performing poorly overall."
What does this research mean for rate decisions?
The Fed has set up five working groups to review its monetary-policy framework; the choice of inflation measure will directly affect the benchmark for rate decisions.
This reflects a real internal divide over a fundamental question: which thermometer should the Fed use to take the economy's temperature?
In plain terms = different rulers yield different readings — Martin's research hands policymakers a concrete alternative ruler for that debate.
Content is for reference only, not financial advice.