US core inflation has conveniently settled at its lowest annual pace in four years — an outcome as reassuring as it is suspect. According to a November report from the Bureau of Labour Statistics, core consumer prices rose by 2.6% year-on-year, marginally below expectations and the weakest reading since early 2021. Headline inflation followed closely at 2.7%. On paper, it looks like progress. In practice, it looks like a number produced under laboratory conditions — or worse, during a blackout.

The longest government shutdown on record prevented the BLS from collecting a meaningful portion of October price data. As a result, November’s figures are less a clean snapshot than a statistical reconstruction exercise. Monthly inflation dynamics were largely unreadable, and even the annual figures may have been quietly distorted. When the data collector is partially blind, the precision of the diagnosis becomes an article of faith.

Still, the report hints at moderation. Core CPI rose just 0.2% over the two months to November, helped by falling prices for hotels, leisure activities and clothing. Furniture and personal-care goods moved the other way, but not enough to spoil the narrative. More troubling, however, is what did not move at all: key housing components — among the stickiest and most consequential drivers of US inflation — showed virtually no change. Outside a recession, such sudden stillness is statistically exotic.

Capital Economics’ Paul Ashworth put it diplomatically: either inflation pressures are genuinely evaporating, or the data are lying. A sudden freeze in shelter inflation, particularly in services, is not something one encounters casually in an economy still growing. His conclusion — shared quietly across the profession — is that December’s data will have to decide whether November was the beginning of something real or merely an artefact of administrative paralysis.

The BLS, for its part, acknowledged the issue. Housing prices were imputed using carry-forward methods, effectively assuming no change from the last observed data point. October rents, for instance, were rolled over from April. The result: indices remain unchanged by construction. This is methodologically defensible in extremis — and analytically dangerous when markets treat it as a signal.

Bloomberg Economics sees enough smoke to suspect fire, cautiously arguing that underlying inflation may indeed be cooling and that the odds of a January rate cut have increased. Their baseline now assumes up to 100 basis points of easing in 2026. The Fed, however, remains split. Having cut rates for the third consecutive meeting last week, policymakers are torn between a labour market that is losing momentum and inflation that refuses to die quietly.

Markets wasted no time. Equities rose, Treasury yields eased, and the dollar slipped on release. Investors are now pricing at least two Fed cuts next year. Jerome Powell himself poured cold water on excessive enthusiasm, warning that CPI data may be “distorted” by the shutdown and that conclusions should be drawn carefully — a rare moment of institutional self-awareness.

Beyond the headline numbers, goods inflation continues to decelerate, services inflation is easing unevenly, and housing — the long-standing villain — is finally showing signs of fatigue, albeit under questionable measurement conditions. Real wages are rising again, up 0.8% year-on-year, offering households some relief and policymakers a new complication.

In short, the inflation story is improving — but through frosted glass. The Fed is being asked to calibrate policy using data partially reconstructed, partially delayed and possibly skewed by Black Friday discounts and truncated surveys. Disinflation may be winning. Or statistics may simply be catching their breath.

For now, prudence demands restraint — from policymakers and investors alike. When inflation looks this good in a report this flawed, the correct response is not celebration, but suspicion.