Headline inflation is the percentage change in a broad consumer-price index including all components within that index's defined coverage. It describes the overall measured price change before the exclusions or filtering used in a core inflation measure. For a quant reading an economic release, “headline” tells you to use the full named index; it does not identify the index, comparison period or seasonal basis by itself.
In the United States, headline CPI usually refers to the all-items Consumer Price Index for All Urban Consumers (CPI-U). Food and energy are included. BLS also publishes the separate all-items-less-food-and-energy index: their exclusion from that series does not remove them from headline. See the BLS explanation.
“All items” means all categories in scope, not every price in the economy. Consumer indexes have defined populations and consumption coverage; they are not indexes of stock prices or every asset purchase. Nor does the aggregate rate have to match a particular household's experience. BLS documents the CPI's population, coverage and weighting.
A fast-rising component is not the headline rate
Consider a fictional fixed-quantity basket observed at two dates exactly one year apart, with no seasonal adjustment. Its two blocks exhaust the basket. At the first date, Block A costs 80 USD and Block B costs 20 USD. The same quantities cost 81.60 USD and 22 USD one year later because their prices rise 2% and 10%, respectively.
| Component | Starting expenditure weight | Annual price change | Contribution to headline |
|---|---|---|---|
| Block A | 80% | 2% | 1.6 percentage points |
| Block B | 20% | 10% | 2.0 percentage points |
| Whole basket | 100% | 3.6% | 3.6 percentage points |
For this fixed basket, the prior-period expenditure weights give the exact decomposition:
The cost calculation checks the same result:
Original fixed-quantity example at two dates one year apart, with no seasonal adjustment. Component rates use percent; additive contributions use percentage points of the whole basket’s annual change.
Block B rises five times as fast as A, yet its contribution is only slightly larger: its starting expenditure share is smaller. Its 10% component rate and 2.0 percentage-point contribution answer different questions. The unweighted average of the two rates is 6%, which treats the blocks as equally important and is wrong for this basket.
The identity is exact under these teaching assumptions. Official indexes use their own aggregation, linking and weight-update methods. For release analysis, use the publisher's contribution tables or a decomposition consistent with that index and interval; do not assume a convenient pair of published weights makes a rounded official annual rate exactly additive. The ABS methodology explains weighting, comparisons and rounding differences.
The index name changes the meaning
Headline CPI and headline PCE inflation are both broad consumer-price measures, but they are not interchangeable observations. The US personal consumption expenditures (PCE) price index covers spending by and on behalf of the personal sector, including households and nonprofit institutions serving households. CPI focuses on household out-of-pocket spending. Their formulas and expenditure weights also differ; BEA groups the differences into formula, weight, scope and other effects. “Headline” does not erase those distinctions.
Other jurisdictions use other named measures. Australia's headline measure is the original All groups CPI, with national data derived as a weighted average of eight capital cities; the ABS also publishes a separate seasonally adjusted all-groups series. The September 2026 methodology describes both. The euro area's broad consumer-price measure is the Harmonised Index of Consumer Prices (HICP), compiled by Eurostat and national statistical institutes under a common methodology, as the ECB explains.
Read the horizon before interpreting the signal
For a monthly index , the month-on-month rate is ; the twelve-month rate is . They can move in different directions because their comparison observations differ. An annual-average change compares averages, while an annualised monthly rate extrapolates one month's growth conventionally. Neither is automatically the reported twelve-month rate.
Seasonal adjustment also matters for short-term interpretation. A seasonally adjusted monthly rate and an unadjusted twelve-month rate should retain those labels in a dataset. Keep the release date, series identifier and vintage as well: later seasonal revisions were not available to an earlier trading decision. Otherwise a backtest can introduce look-ahead bias.
Finally, headline is an aggregate, not a measure of how many categories are rising or a guarantee of persistence. One component can dominate its movement. Read its contributions alongside core measures and the broader distribution of price changes. A lower positive headline rate is disinflation; prices in the measured basket are still rising overall. Deflation requires a negative change over the stated interval.
Sources
- BLS: CPI Frequently Asked Questions — all-items CPI-U, coverage, weights, comparison intervals and seasonal adjustment.
- BLS: Common Misconceptions about the Consumer Price Index — food and energy remain in headline CPI.
- BEA: Differences between the PCE price index and CPI — scope, formula, weight and other effects.
- ABS: CPI methodology, September 2026 — All groups CPI, weights, original and adjusted series, rounding and revisions.
- ECB: Measuring inflation and consumer prices — euro-area HICP and its compilation. Official sources checked 28 October 2026; numerical data above are fictional.