Reading data · 7 min read · 5 steps · Updated Invalid Date

How to read CPI components: going beyond the headline number

Every major CPI release includes a detailed component breakdown — not just the headline number. For the US:

The BLS CPI news release includes Table 1 with all major categories and their contributions to the annual change.
The BLS CPI data tables provide month-by-month series for every sub-index going back decades.
FRED carries all BLS CPI series with a convenient chart and download interface.

For the euro area, the Eurostat HICP database (PRC_HICP_MANR) provides all components. For the UK, the ONS MM23 dataset is the master file.

CPI is typically divided into several broad groups, each with different inflation dynamics:

Food at home (groceries): Driven by agricultural commodity prices, supply chains, and weather. Often volatile month-to-month but mean-reverting over 12–18 months. The FAO Food Price Index is a leading indicator.
Food away from home (restaurants): Driven by labour costs — much stickier than grocery prices because wages are slow to fall. Closely tracks services inflation.
Energy: The most volatile component. Tracks global oil and gas prices with a short lag. The US EIA gasoline price tracker gives a real-time read on where this component is heading.
Core goods (non-energy industrial goods): Electronics, clothing, furniture. Driven by global supply chains. Surged in 2021–2022 as supply chains broke; normalised by 2024.
Shelter: Rent and owners' equivalent rent. Largest US CPI component (~36%). Lags market rents by 12–18 months — see our housing inflation explainer.
Core services (ex-shelter): Healthcare, transport, education. The Fed's preferred underlying inflation gauge — closely tied to wage growth.

To understand which components are driving headline CPI, calculate each component's contribution:

Contribution = component rate × component weight in the basket

Example: If shelter CPI is 5.5% and shelter weight is 36%, shelter's contribution to headline CPI is approximately 5.5% × 0.36 = 2.0 percentage points. The BLS Table 1 in each news release does this calculation for you — look for the "Unadjusted percent change" column alongside the "Contribution to index" column. When a single component (like energy) contributes 2+ percentage points in either direction, the headline number is telling you more about that one price than about the broad economy.

A year-on-year rate compares this month to the same month last year. If last year's comparison month had an unusual price spike or drop, it creates a "base effect" that distorts the current reading. The most common source is energy:

If energy prices were very high in the comparison month (e.g., March 2022 after the Ukraine invasion), the year-on-year energy comparison will be flattering (prices look lower) in March 2023.
If energy prices were very low in the comparison month (e.g., April 2020 during COVID lockdowns), the year-on-year comparison will be unflattering (prices look higher) in April 2021.

You can identify base effects by looking at the month-on-month change in parallel. If headline is rising year-on-year but month-on-month is flat or negative, a base effect is doing the work. The FRED CPI energy component (CPIENGSL) helps you see historical energy price spikes that will affect future year-on-year comparisons.

Once you've broken down the components and identified base effects, you can tell the real story behind the headline. A useful framework is to classify the inflation as:

Supply-driven vs demand-driven: Supply shocks (energy, food) tend to hit headline but spare core. Demand-driven inflation tends to be broader, hitting core services, and more persistent.
Transitory vs persistent: Goods inflation tends to be transitory (global supply chains normalise); services inflation is stickier (wages are the main input).
Broad or narrow: The Cleveland Fed Median CPI and 16% Trimmed Mean CPI strip out the highest and lowest components each month to give a picture of the "typical" price change — a better signal of underlying inflation breadth.

Our inflation explorer lets you track the headline trend across all countries. For deeper component analysis of your specific country, the national statistics agency's data portal is always the definitive source.

Sources & further reading

Terms in this guide

core inflation headline inflation consumer price index cpi base effect seasonal adjustment

Frequently asked questions

What is the difference between headline and core inflation?

Headline inflation is the all-items CPI change — it includes food and energy. Core inflation strips out food and energy on the grounds that they are volatile and supply-driven. Core is typically more useful for judging persistent inflation pressure. The BLS explains core CPI and its uses in detail. Note: there is no single universal definition of "core" — some measures exclude only energy; others exclude both food and energy; the ECB's "supercore" also excludes volatile services.

How do I find what drove inflation in a specific month?

For the US: the BLS CPI news release includes Table 7 showing contributions of each component to the 12-month change in percentage points. This directly answers "shelter contributed X pp; energy contributed Y pp" etc. For the euro area: the Eurostat HICP database includes decomposition tables. For the UK: the ONS CPI bulletin includes a contribution chart in every release.

What are the most volatile CPI components?

Across most economies, the most volatile components are: (1) energy — especially motor fuels, which track oil prices; (2) fresh food — especially vegetables and fruits, which are weather-sensitive; (3) airfares — highly seasonal and react to fuel costs. The least volatile components are typically: shelter/rent, healthcare services, and education — all of which are either regulated or driven by sticky wages. The Cleveland Fed Median CPI and trimmed-mean measures exclude the most volatile components each month automatically.