Trimmed mean inflation vs core PCE: Dallas Fed method shapes Federal Reserve policy

Fresh US inflation data is giving the Federal Reserve conflicting messages. One measure sits well above the 2% goal. Another suggests price pressure is nearer to target. This split is drawing market attention as investors track US interest rates. It also revived debate on which inflation gauges deserve more weight in policy decisions.

The argument gained urgency after Kevin Warsh raised concerns at an April confirmation hearing. Warsh urged policymakers to rethink how they read volatile price shocks. Warsh said the Fed should rely more on measures like the "trimmed mean" index. That approach strips out the most extreme price changes each month.

The consumer-price index from the Labor Department often drives headlines because it arrives earlier. CPI also feeds into Social Security adjustments and inflation-linked bonds. Many private contracts also refer to CPI. Fed officials, though, lean more on the personal-consumption expenditures price index, or PCE, from Commerce.

PCE sets the Fed’s 2% target and covers more household spending categories. It also adjusts for shoppers switching between items as relative prices change. In April, overall PCE inflation measured 3.8% from a year earlier. Policymakers often stress "core" prices, which exclude food and energy due to sharp swings.

Commerce reported core PCE inflation at 3.3% over the past year. That level still sits above the target. A trimmed mean inflation reading, which cuts the largest rises and falls, was much lower at 2.3%. Warsh criticised the core metric and called it a "rough swag". Warsh argued it leaves too many one-off distortions.

Fed officials already review several filters for large price moves. Examples include wireless phone bill cuts or temporary tariff effects. Trimmed mean inflation differs because it uses a fixed monthly rule. It removes the most extreme changes to show an underlying trend. Warsh told lawmakers, "What I'm most interested in is what's the underlying inflation rate, not what's the one-time change in prices because of a change in geopolitics or a change in beef,".

The main April readings discussed by economists are listed below.

Measure	April reading	Notes
Overall PCE inflation	3.8% year-over-year	Headline PCE from Commerce
Core PCE inflation	3.3% year-over-year	Excludes food and energy
Trimmed mean inflation	2.3% year-over-year	Removes extreme increases and decreases
Dallas Fed gap estimate	0.7 percentage points	Trimmed mean below core PCE in April
Employ America symmetric trimmed mean	3% year-over-year	Cuts equal shares of highs and lows
Employ America ex-housing and other categories	2.8% year-over-year	Excludes lagging or indirectly measured items

Dallas Fed method and trimmed mean inflation: how the index is built

Warsh did not name a preferred trimmed mean inflation series. The best-known version comes from the Federal Reserve Bank of Dallas. It removes categories that together exceed half of spending each month. It drops both sharp rises and steep falls. Dallas Fed research says this method often predicted future inflation better than core.

Yet 2021 weakened confidence in the index. During the pandemic rebound, inflation rose faster than many expected. The trimmed mean measure increased more slowly. Some officials treated that slower rise as evidence inflation would fade. That judgement later looked too optimistic because inflation stayed elevated.

The Dallas Fed design helps explain that period. Between 1977 and 2009, price falls were often steeper than rises. Equal trimming could have pushed the average trend too high. The Dallas Fed therefore removed the largest 31% of increases each month. It removed only the biggest 24% of decreases.

That setting worked in earlier decades but struggled during the pandemic. Price rises then became steeper than price falls. Cutting a larger slice of increases removed more upward pressure. The index therefore understated how strongly inflation was building in 2021. This supported the view that the surge would be temporary.

Tariffs, AI spending, and trimmed mean inflation: why the gap persists

Recent readings show trimmed mean inflation again below broader measures. Dallas Fed researchers said April’s trimmed mean was 0.7 percentage points under core PCE. They linked the gap to lower weight on goods excluding food and energy. Those goods are most exposed to tariffs. Tariff increases have raised costs across imported products.

Another influence is higher spending tied to artificial intelligence. That demand has lifted prices in software and computing. Economists are debating whether these moves are temporary shocks. Others see signs of lasting demand. Analysts at Employ America built alternative trimmed measures using the same PCE data to test this.

Employ America’s symmetric trimmed mean inflation cuts equal shares of extremes. It narrowed much of the gap with core PCE. That gauge stood at 3% in April. The think tank also tracks a series excluding housing and other lagging categories. Examples include some medical costs not directly measured.

That second Employ America gauge was 2.8% in April. It rose year over year for 13 straight months. Riccardo Trezzi, a former Fed economist, now runs an inflation-research firm. Trezzi said the distribution of prices has missed the 2% goal since 2021. Trezzi also said it drifted higher in recent months.

These measurement choices matter for policy because the Fed may "look through" some shocks. That means holding rates steady if pressures may fade. Warsh’s view links to shocks from tariffs, geopolitics, and heavy AI investment. "The question is whether 'looking through' becomes a principled framework or simply a way to down-weight inconvenient inflation prints when needed," said Trezzi.

The April split between core PCE and trimmed mean inflation has kept the debate active. Core PCE showed inflation still above target, while trimmed mean looked closer. Alternative trims landed between those signals and showed persistent upward drift. For markets, the key issue remains which gauge the Fed prioritises when setting interest rates.