Fed Under Warsh May Swap Its Core Inflation Gauge for a Trimmed-Mean Measure, and Critics Are Already Calling It Score-Keeping by Different Rules

The Federal Reserve is reportedly moving toward changing how it measures inflation. According to an Axios report published July 6, the Fed is considering shifting from the traditional core Personal Consumption Expenditures (PCE) index to trimmed-mean or median inflation metrics. These measures strip out the highest- and lowest-inflation items each month before calculating an average. The result, by design, is a smoother number.

What Warsh Actually Said

This is not a sudden idea. Kevin Warsh, Trump's nominee to chair the Federal Reserve, laid out his preference during Senate confirmation testimony. According to CNBC's April 22 coverage, Warsh said: "The measures I prefer are looking at things that are called trimmed averages. We take out all of the tail-risks, all of the one-off items, and we ask ourselves whether the generalized change in prices is having second-order effects on the economy."

Warsh's framing is coherent on its face. Volatile one-time swings in energy prices or used cars have historically distorted short-term PCE readings. A trimmed measure is specifically designed to reveal underlying trend inflation, not noise.

The Dallas Fed already publishes exactly this kind of metric. RSM chief economist Joe Brusuelas noted that the Dallas Fed PCE trimmed mean currently sits at 2.4% and is expected to tick up modestly in the May reading. That figure excludes roughly 55% of all price changes each month, including the top 31% of high-inflation items and the bottom 24% of low-inflation items.

The Legitimate Criticism

Critics are raising substantive concerns. Haver Analytics described the trimmed-mean approach as producing an "alternative reality" and called it a "statistically manipulated version of reported inflation." The core objection is that by excluding outliers, you also exclude the very price spikes that consumers actually feel. During a supply shock, the trimmed mean has a documented tendency to understate inflation, a point Brusuelas flagged directly, noting the metric "tends to understate inflation during supply shocks."

That matters right now. Core service-sector prices remain sticky, and Brusuelas cited AI-driven competition for scarce resources and U.S. defense-sector demand as ongoing pipeline pressures. Whether underlying inflation is cooling or merely being measured differently is not an academic question when the Fed is deciding whether to cut rates.

Bank of America ran historical comparisons and found the trimmed-median gauge actually ran higher than core PCE in both 2019 and 2020. BofA analyst Bhave made a pointed observation: "To preserve Fed credibility and avoid optics of cherry picking, Warsh will need to stick with his preferred metrics even when they are outpacing the core." That is a genuine accountability standard, not a partisan attack.

The Strongest Case for the Change

Proponents of trimmed-mean measurement, including researchers at the Dallas Fed and the Cleveland Fed, have published evidence that their measures outperform headline PCE as predictors of future inflation trends, not just as current snapshots. The Cleveland Fed publishes its own CPI trimmed mean alongside the Dallas Fed measure, and both have academic credibility going back decades. The argument is not that the Fed wants to lie about inflation. Rather, core PCE has its own distortions, and a trimmed approach corrects for different flaws than it introduces.

Warsh would also need the full Federal Reserve Board to agree. As ASA Connect member Chris Barker noted in an April 26 forum post summarizing the Senate testimony, all seven Fed governors would need to vote on a formal methodology change. Warsh cannot do this unilaterally.

What Actually Changes, and What Does Not

The numbers in your grocery cart stay exactly the same. What changes is the single figure the Fed watches most closely when deciding whether inflation is "at target." If trimmed-mean PCE reads differently than traditional core PCE, the Fed's declared progress toward its 2% goal looks markedly different depending on which ruler it picks up.

Market participants are treating this as a potential rate-cut signal. The implicit logic: if the Fed can show inflation closer to target using its new preferred measure, it has more political and institutional cover to cut rates without appearing to have surrendered on price stability.

The unresolved question is the one Brusuelas flagged directly. Warsh should expect to be pressed at his next Fed press conference on the asymmetry built into the Dallas Fed metric, specifically whether a measure that excludes 55% of monthly price changes is reliable enough to steer the world's most consequential interest-rate decisions during a period of active supply-side pressure.