PCE inflation is the percentage change in the price index for personal consumption expenditures, the Commerce Department's broadest gauge of what consumer goods and services cost in the United States. The Bureau of Economic Analysis (BEA) publishes the index each month in its Personal Income and Outlays report, scaled so the 2017 average equals 100. It prices everything US residents consume, including purchases made for them by employers, government programs and nonprofit groups, so it reaches well beyond the out-of-pocket spending tracked by the consumer price index. Headline PCE inflation covers the whole basket, while core PCE leaves out food and energy. The Federal Reserve defines its 2% inflation goal on the headline measure, which is why the release moves the rates discussed across the Treasuries hub.
How is the PCE price index calculated?
BEA weights detailed price indexes, most of them CPI components, by spending in each category and chains the months with a Fisher formula that lets weights change.
The spending side comes from the national accounts. BEA estimates consumer spending mainly from business and government sources, such as Census Bureau retail and services surveys, rather than from a household survey, and in September 2026 it ran at an annual rate of $22.45 trillion. For prices, BEA's handbook says the closest match for most categories is a CPI or a group of CPIs, with producer price indexes and other sources filling gaps. The two gauges therefore share much of their raw price data and differ mainly in what they cover and how they combine it.
Each monthly change is a Fisher index: the geometric mean of a Laspeyres index, which prices last month's purchases at this month's prices, and a Paasche index, which does the same with this month's purchases.
Fisher price change = √(Laspeyres change × Paasche change)
Chaining those monthly links produces the index level, and inflation over any period is the ratio of two levels:
PCE inflation (%) = (index this month ÷ index 12 months earlier − 1) × 100
Worked example: How the Fisher formula handles a switch from beef to chicken
Illustrative numbers. Last month a household bought 10 pounds of beef at $10 and 20 pounds of chicken at $5, spending $200. This month beef costs $12, chicken is still $5, and the household buys 8 pounds of beef and 24 of chicken.
- Laspeyres, last month's quantities: (10 × $12 + 20 × $5) ÷ $200 = 1.100, a 10.0% rise
- Paasche, this month's quantities: (8 × $12 + 24 × $5) ÷ (8 × $10 + 24 × $5) = $216 ÷ $200 = 1.080, an 8.0% rise
- Fisher: √(1.100 × 1.080) = √1.188 = 1.090, a 9.0% rise
A formula that holds quantities fixed reports the full 10%, while the Fisher result gives partial credit for the move to cheaper chicken. BEA calls the resulting gap between the two national gauges the formula effect.
Why does the Fed target PCE inflation?
The Fed says it uses the PCE index largely because it covers a wide range of household spending; the 2% goal has been tied to PCE since 2012.
The choice dates to February 2000, when the Fed's Monetary Policy Report to Congress moved its inflation forecasts from the CPI to the PCE chain-type index. The report gave three reasons: the formula follows the changing mix of spending and so avoids some of the upward bias of a fixed basket, the weights rest on a more comprehensive measure of spending, and past data can be revised so the series stays consistent. Every framework statement since has kept the same gauge:
| Date | Document | What it says about the goal |
|---|---|---|
| January 25, 2012 | First Statement on Longer-Run Goals and Monetary Policy Strategy | 2% inflation, measured by the annual change in the PCE price index |
| August 27, 2020 | Revised statement | Seeks inflation that averages 2% over time, aiming moderately above 2% after persistent shortfalls |
| August 22, 2025 | Revised statement | Returns to flexible inflation targeting and drops the makeup strategy; the 2% PCE goal stays |
| January 27, 2026 | Annual reaffirmation | The August 2025 text is reaffirmed unanimously, without changes |
Chair Powell explained the 2025 change by saying the idea of an intentional, moderate overshoot had proved irrelevant once inflation surged in 2021. The current statement adds that the Committee is prepared to act forcefully to keep longer-term inflation expectations anchored at 2%. The goal applies to headline PCE. Core measures help policymakers separate lasting trends from swings in food and fuel, and the Fed also tracks the CPI and producer prices. How those judgments feed into the federal funds rate and bill yields is covered in how Fed rate decisions affect T-bill rates.
How does PCE inflation differ from CPI inflation?
PCE counts more kinds of spending, takes its weights from business data, uses a Fisher formula and is revised over time; the CPI-U prices urban households' out-of-pocket purchases.
| Feature | PCE price index | CPI-U |
|---|---|---|
| Publisher | BEA, Commerce Department | BLS, Labor Department |
| Spending covered | All US residents, plus spending for them by employers, governments and nonprofits | Out-of-pocket purchases of urban consumers, over 90% of the population |
| Source of weights | National accounts built from business and government data | Consumer Expenditure Surveys of households |
| Housing | 15.6% of spending (September 2026) | 35.6% for shelter (December 2025 weights) |
| Health care | 17.2% of spending | 8.4% for medical care |
| Formula | Fisher, with weights updated every month | Modified Laspeyres, weights fixed for a year |
| Revisions | Revised as fuller data arrive | Unadjusted index final when published |
| Reference base | 2017 = 100 | 1982-84 = 100 |
| 12 months to September 2026 | 3.6%, core 3.1% | 3.7%, core 2.6% |
The health care gap is the scope difference at work. PCE includes employer-paid health insurance and medical care financed by government programs, while the CPI counts only what households pay themselves. BEA's Table 9.1U splits every monthly gap into four parts: the formula effect, the weight effect, the scope effect and "other effects," which cover seasonal adjustment and differences in the prices used.
Worked example: BEA's bridge from PCE to the CPI for September 2026
The seasonally adjusted CPI-U rose 0.6% in September 2026 and the PCE price index rose 0.5%. Table 9.1U assigns the gap in percentage points, with a plus sign where the effect lifted the CPI above PCE:
- Weight effect: +0.16, led by gasoline at +0.12 and rent of shelter at +0.06
- Formula effect: +0.03
- Items only PCE covers, led by hospital and nursing home care: −0.05
- Items only the CPI covers: +0.02
- Seasonal adjustment and price differences: −0.01
- Total: 0.16 + 0.03 − 0.05 + 0.02 − 0.01 = +0.15 point, consistent with the published 0.6% and 0.5% once rounding is allowed for
Gasoline jumped 9.1% in September in the seasonally adjusted CPI, and motor fuel carries more weight in the CPI than in PCE, so the CPI ran hotter that month. Over 12 months the order flips for core prices: core PCE rose 3.1% against 2.6% for core CPI. Shelter, up 3.2% over the year in the CPI, has less than half the weight in PCE, so faster increases elsewhere in the basket count for more. In 2000 the Fed noted a tendency for the CPI to rise faster than PCE; in 2026 the core measures ran the other way.
What is the latest PCE inflation reading?
The latest figure is 3.6% in the 12 months to September 2026, or 3.1% excluding food and energy (US Bureau of Economic Analysis, as of Nov 2, 2026).
Worked example: Annual PCE inflation to September 2026 from the index levels
BEA released these index levels, with 2017 = 100, on October 29, 2026.
- Headline: 132.257 in September 2026 and 127.601 in September 2025, so 132.257 ÷ 127.601 = 1.036489, a 3.65% rise published as 3.6%
- Core: 130.902 ÷ 126.940 = 1.031212, a 3.12% rise published as 3.1%
- Distance from the goal: 3.6 − 2.0 = 1.6 percentage points, or 160 basis points
- September alone: 132.257 ÷ 131.611 = 1.0049, the 0.5% monthly rise in the release
| Month, 2026 | PCE, 12-month change | Core PCE, 12-month change |
|---|---|---|
| June | 3.5% | 3.0% |
| July | 3.4% | 3.0% |
| August | 3.4% | 3.0% |
| September | 3.6% | 3.1% |
These rates come from BEA's index levels as of the October 29, 2026 release; an older release may show a slightly different figure for the same month because PCE data are revised. The October 2026 report is due November 25, 2026, and unlike the CPI it will include a 12-month change. BLS collected no October 2025 prices, so BEA filled that month with the geometric mean of the September and November 2025 CPIs.
Where do you see PCE inflation?
You meet it in BEA's monthly release, in every set of Fed projections, in FOMC statements and press conferences, and in the bond-market moves that follow a surprise.
- The BEA release. Personal Income and Outlays comes out at 8:30 a.m. Eastern about four weeks after each month ends, roughly two weeks after the CPI for the same month.
- Fed projections. The Summary of Economic Projections, published four times a year, has rows for PCE and core PCE inflation. At the September 15-16, 2026 meeting the median participant projected 3.7% PCE inflation for 2026, 2.3% for 2027 and 2.0% in the longer run, each measured from the fourth quarter of one year to the fourth quarter of the next.
- Rates and yields. An inflation surprise shifts expectations for the federal funds rate, now 3.75% to 4.00%, and moves bill and note yields with it. What moves Treasury yields and the quantitative easing entry cover the other forces on long rates.
- Household planning. At 2% a year, prices rise about 21.9% over a decade; at 3.6% they rise about 42.4%. The compound interest calculator runs that arithmetic for any rate, and real returns and inflation shows how to take inflation out of a yield.
Which PCE figures are most often misread?
The usual errors are treating core PCE as the target, mixing up the price index with the spending total, and comparing figures built on different bases.
- Core is not the goal. The 2% objective is written for the full index, food and energy included; core is a trend gauge.
- Spending is not prices. "PCE" also names the spending total. In September 2026 current-dollar spending rose 0.5% and prices rose 0.5%, so real spending barely moved, and a headline saying "PCE rose 0.5%" could mean either.
- Projection years are not 12-month rates. The Fed's 3.7% for 2026 compares fourth-quarter averages, so it will not match the 12-month change printed for any single month.
- PCE indexes nothing you hold directly. TIPS and I bonds follow the CPI-U and Social Security follows the CPI-W; PCE reaches savers through Fed policy and interest rates.
- Old releases go stale. BEA revises earlier months as source data improve, so a figure copied from an old release may not match today's series, while the unadjusted CPI-U never changes once published.
- The makeup rule is gone. Since August 2025 the Fed no longer says it will aim above 2% after a stretch of low inflation; the goal is 2% from here, with no catch-up for past misses.