CPI vs. PCE: Reading US Inflation Reports
US CPI is higher than PCE—which is correct, and will the Fed immediately raise rates? This article separates index meanings, percentage calculations, and limits of interpreting one month's news.

CPI and PCE measure different scopes of inflation
Short answer: CPI and PCE measure US consumer price changes with different spending scopes, weights, and formulas. Different readings need not mean an agency made an error. Subtracting one index’s inflation rate from the other does not show a decline over time.
CPI (Consumer Price Index) in news reports usually means CPI-U, produced by Bureau of Labor Statistics (BLS) tracks price changes in a representative basket bought by urban consumers. It helps explain household price pressures without matching every household’s spending.
PCE price index (Personal Consumption Expenditures Price Index) prepared by Bureau of Economic Analysis (BEA) covers a broader spending scope, including household purchases and spending paid on consumers’ behalf, such as some employer- or government-funded health services. It is not limited to direct household outlays.
Both indices use US data, not Thai inflation data. Comparisons with living costs in Thailand require separate consideration of spending patterns and exchange rates.
How CPI and PCE differ in coverage and weighting
| Issue | CPI-U | PCE price index | Meaning when reading the news |
|---|---|---|---|
| Organization | BLS | BEA | Make sure the report name matches the news source. |
| scope | Purchases of goods and services by households in urban areas according to the index definition | Personal consumption, including spending paid on someone's behalf | Health expenses may be weighted differently. |
| Product weight | Based on consumer expenditure surveys, with periodically updated weights | Based on expenditure data in national accounts, with weights adjusted as spending changes | Categories with equal price increases may contribute differently to the overall index |
| Index aggregation method | Uses a modified Laspeyres approach at the aggregate level | Uses a Fisher chain-weighted index | The formulas differ and respond differently to changes in consumption behavior |
| What it is used to assess | Price changes faced by urban consumers within the basket | Price changes across a broader scope of consumption | No single index answers every question |
BEA explains CPI/PCE differences through formulas, weights, scope, and other factors such as seasonal adjustment and input prices. See BEA and Federal Reserve Bank of Cleveland directly
For example, if housing has a larger weight in one basket and housing prices rise quickly, that index may increase faster. This illustrates a weighting effect; it does not establish housing as the cause of every monthly CPI/PCE gap.
What do Headline and Core say?
Both CPI and PCE have figures for Headline which includes food and energy, and the figure Core excludes the food and energy categories defined for that index to reduce volatile price effects and help reveal underlying trends. It does not mean food and energy are unimportant to households.
According to Description of BEA's Core PCE The excluded food category means food and beverages purchased for consumption away from the point of sale; it does not straightforwardly exclude all restaurant services. Energy covers defined fuel and utility categories. Read the definitions rather than using everyday meanings.
For a headline saying “US inflation 3%,” identify CPI or PCE, headline or core, MoM or YoY, and the reporting month. Each changes what the figure means.
Interpret MoM, YoY, and annualized rates correctly
MoM (month over month) means the change from the previous month, while YoY (year over year) measures the change from the same month a year earlier. These figures answer different questions even when released on the same day.
When numeric price-index values are available, the formula is Percentage change = (new-period index ÷ base-period index − 1) × 100 Use the previous month for MoM or the same month last year for YoY, with seasonal adjustment appropriate to the comparison
Hypothetical example, not actual published CPI or PCE data: If the previous month's seasonally adjusted index is 100.000 and the latest month's is 100.300, MoM = (100.300 ÷ 100.000 − 1) × 100 = 0.30% If the non-seasonally adjusted index was 100.000 in the same month last year and is 103.200 in the latest month, YoY = (103.200 ÷ 100.000 − 1) × 100 = 3.20% The two sets of numbers are separate examples to show the formulas. It is not assumed that they are the same index series.
Annualizing a 0.30% monthly rate under the assumption that it repeats for 12 consecutive months gives ((1 + 0.003)^12 − 1) × 100 ≈ 3.66% This value is Hypothetical annualized rate not an actual measured YoY rate or a forecast of 3.66% inflation next year
Short-term MoM comparisons often use seasonally adjusted data to reduce recurring patterns; 12-month CPI headlines usually use unadjusted series. Check table headers. BLS says seasonal factors can be recalculated and prior data revised; do not mix data vintages or adjustment methods without disclosure (BLS: Seasonal Adjustment)
A higher comparison base can lower YoY inflation even as current prices rise
Hypothetical example Use the same set of indices and do not adjust for seasonality: Last year's August index was 100 and September 102; this year August is 104 and September 105. August YoY therefore equals (104 ÷ 100 − 1) × 100 = 4.00% But September YoY equals (105 ÷ 102 − 1) × 100 ≈ 2.94% Even though the last month's index increased from 104 to 105.
This shows the base effect: the YoY rate can decrease because the base month of the previous year also increased. It is therefore worth looking at both the index level and the seasonally adjusted monthly trend together. This example is for illustrative purposes only. Non-seasonally adjusted MoM is not used to replace short-term trends in actual reports. The rounding in the example is performed to two places.
Hypothetical example: Different numbers do not necessarily contradict each other.
Suppose one month's news reports Headline CPI up 3.2% YoY while Headline PCE rises 2.7% YoY. Both figures are hypothetical, not the latest readings at the time of writing . They do not mean inflation fell from 3.2% to 2.7%: the indices differ even when they cover the same 12-month period.
| Fictional headline | can be read as | Do not conclude that |
|---|---|---|
| Headline CPI +3.2% YoY | The CPI-U index is higher than the same month last year 3.2% | Every item increases in price 3.2% or every house pays the same increase. |
| Headline PCE +2.7% YoY | The PCE index is higher than the same month last year 2.7% | Inflation fell from CPI 3.2% to PCE 2.7% |
| Core PCE +0.3% MoM | The PCE index excluding food and energy rose 0.3% from the previous month | Core PCE YoY is equal to 0.3% or the whole year will definitely increase 3.6% |
Percent and percentage points differ: The difference between 3.2% and 2.7% is 0.5 percentage points, not 0.5%. Here it compares two different indices, so it does not represent a decline over time in the same inflation series.
Another precaution is Slower inflation does not mean falling prices If YoY falls from 4% to 3%, the overall price level is still rising from a year earlier, just more slowly. An actual fall in the price level is deflation over the measured period. Check the price index, rather than merely whether the inflation rate fell from its previous reading.
Why the Fed monitors PCE without relying on one number
In Federal Reserve Monetary Policy Framework revised 22 August 2025 the FOMC affirmed a longer-run 2% inflation objective measured by the annual change in the Overall PCE price index not an official target requiring Core PCE to equal 2% every month
The Fed considers more than PCE alone. Policy documents specify price stability and maximum employment, with multiple indicators considered. When reading inflation news, distinguish Announced information from Market expectations and FOMC decision which is a different step
Faster monthly Core PCE growth may warrant closer attention to price pressures, but it is insufficient to conclude that the Fed will raise rates at its next meeting. Several months of data, employment conditions, and other economic indicators also matter. A stock or bond price move after the release does not establish that a single inflation figure caused it.
How to verify figures at their source and avoid mistakes
- Start with the filing name: See CPI at BLS CPI Report and PCE in Personal Income and Outlays from BEA report page separately. Do not use charts without an identified agency.
- Record the reference month as well as the release date: Reports released one month may describe the previous month's prices, so reading "inflation this month" may be inaccurate.
- Note the index type and comparison base: CPI-U or PCE, Headline or Core, MoM or YoY, seasonally adjusted or unadjusted
- See trends rather than single points: Compare months using consistently defined datasets and check for revisions
- Separate facts and forecasts: The numbers from the agency are the announced results. As for the word "higher than expected," you need to know from whom and when the forecast was made. It is not a fact of the index itself.
Do not confuse PCE which means the value of consumption expenditures and PCE price index used to track prices In the same report there are also income, expenses, and values adjusted for price effects. Therefore, you must read the table names and units before selecting numbers to use. BEA: PCE price index
For Thai readers following US stocks, baht-dollar exchange-rate changes can make baht returns differ from dollar returns. CPI and PCE do not predict exchange rates or future stock returns, so inflation figures should not directly predict portfolio outcomes.
Read the full index name and reporting period before interpreting
News-reading exercise: If the text says only “Inflation 0.3%,” record only the stated rate. You cannot identify CPI versus PCE or MoM versus YoY yet. Leave missing fields blank and open the primary report for the series, month, and seasonal adjustment. Do not annualize to 3.66% until confirming that 0.3% is monthly, and state the assumption of repeating that rate for 12 months.
When reading US inflation headlines, check at least four things: What index? Does it include food and energy? What time period is it compared to? And what set of data is used? The CPI gives a reading of the prices faced by urban households, while the PCE covers consumption expenditures more broadly and is the measure the Fed uses to target long-term inflation. A difference in two numbers is not a mistake. But it is the result of different definitions and calculation methods.
Understanding the difference supports sound economic interpretation without treating one month's data as a certain forecast for interest rates or asset prices. For inflation's everyday purchasing-power impact, read How inflation affects savings which is explained with an example of calculation in baht
Sources
- BLS: Consumer Price Index Frequently Asked Questions
- BEA: What accounts for the differences in the PCE price index and the Consumer Price Index?
- Federal Reserve Bank of Cleveland: CPI versus PCE Price Index
- BEA: What is the Core PCE price index?
- Federal Reserve: 2025 Statement on Longer-Run Goals and Monetary Policy Strategy
- BLS: Using seasonally adjusted and unadjusted data
Principles and sources reviewed October 12, 2026. Figures are hypothetical calculation examples, not current CPI/PCE data or forecasts. General education, not personalized investment or financial advice.