Trimmed Mean and Sticky-Price Inflation

Core inflation measures strip out food and energy because those prices swing wildly for reasons unrelated to broad inflation pressure, but even "core" can be jumpy if a handful of other categories spike or crash in a given month. Two further techniques address this. The trimmed mean ranks every price category by its monthly change and discards the tails — the most extreme risers and fallers — before averaging what's left, so no single erratic category can dominate the reading.

Sticky-price inflation takes a different angle: it separates categories whose prices change infrequently (rent, insurance, services) from "flexible" categories that reprice constantly (gasoline, some food). Because sticky prices adjust slowly, they're thought to better reflect where firms expect inflation to settle, making sticky-price inflation a forward-looking gauge rather than a backward-looking snapshot.

Worked example. In a month where used-car prices spike 8% and airfares drop 6%, the headline CPI jumps around on these two categories alone; the trimmed mean, having cut both extremes from the calculation, might show a much steadier 0.3% monthly gain that better reflects broad price pressure.

Neither measure replaces the headline number for policy purposes — regulators and central banks still target overall CPI or PCE — but both are used as cross-checks when the headline swings sharply on one or two categories, helping distinguish a genuine shift in inflation trend from a single volatile data point. Some regional Federal Reserve banks publish trimmed-mean and sticky-price series monthly precisely so analysts can triangulate between several competing views of "true" underlying inflation rather than relying on any single measure alone.