CPI vs PCE: why the US has two main inflation measures

By the Business News Analysis desk · Published September 29, 2026

The Consumer Price Index and the PCE price index both track inflation, but they differ in scope, weights and method. Here is why the Federal Reserve targets PCE.

When the news says inflation rose or fell, it usually means the Consumer Price Index. When the Federal Reserve talks about its 2% goal, it means a different measure: the price index for personal consumption expenditures, or PCE. The two usually move together, but not always by the same amount, and the differences explain some of the confusion around inflation headlines.

Who produces them

The Consumer Price Index (CPI) is published monthly by the Bureau of Labor Statistics. It measures the average change in prices paid by urban consumers for a basket of goods and services. The BLS collects prices for tens of thousands of items across the country and weights them using spending patterns from the Consumer Expenditure Surveys. CPI is also used to adjust Social Security benefits, tax brackets and many private contracts.

The PCE price index is published monthly by the Bureau of Economic Analysis as part of its personal income and outlays report, usually a couple of weeks after the CPI. It measures price changes for the goods and services consumed by households, and by nonprofit institutions serving households, as measured in the national accounts. Much of its price information actually comes from the CPI and the Producer Price Index.

Difference one: scope

CPI covers what urban consumers pay out of pocket. PCE covers spending by or on behalf of households. The biggest example is health care. CPI mostly captures what people pay directly, including their share of insurance premiums. PCE also includes medical care paid for by employer-provided insurance, Medicare and Medicaid. As a result, health care carries a much larger weight in PCE than in CPI.

Difference two: weights

Because the scope is different, the weights are different. Housing is the clearest case. Shelter, which includes rent and an estimate of what homeowners would pay to rent their own homes ("owners' equivalent rent"), makes up about a third of the CPI basket. In PCE, housing's share is considerably smaller because the overall basket is broader. When rents are rising quickly, CPI tends to run hotter than PCE, and when shelter inflation cools, CPI tends to fall faster.

The weights are also sourced differently. CPI weights come from household surveys; PCE weights come from business surveys and other data used in the national accounts.

Difference three: formula

CPI uses a formula that holds the basket fixed for a period between weight updates, although it uses geometric means within detailed categories to allow some substitution between similar items. PCE uses a chain-type formula that updates the weights every period. In practice, that means PCE reflects more quickly when people shift spending away from items whose prices are rising, for example buying chicken instead of beef. This "substitution effect" is one reason PCE inflation has tended to be somewhat lower than CPI inflation over long periods.

Why the Fed prefers PCE

In its statement on longer-run goals, first adopted in January 2012, the Federal Open Market Committee defined its inflation objective as 2% measured by the annual change in the PCE price index. The Fed has explained its preference in terms of the differences above: PCE covers a wider range of spending, its weights adjust as spending patterns change, and past PCE data can be revised as better information arrives, which makes it more consistent over time. CPI, by contrast, is generally not revised apart from seasonal factors.

"Core" measures

Both indexes have a "core" version that excludes food and energy. Food and energy prices are volatile and are strongly affected by weather, geopolitics and supply shocks. Excluding them does not mean they do not matter to households; it means the core measure is often a better guide to where overall inflation is heading. Policymakers watch core PCE particularly closely. Analysts also track other underlying measures, such as the Dallas Fed's trimmed mean PCE and the Cleveland Fed's median CPI, which exclude the most extreme price changes each month instead of fixed categories.

How to read the two together

• CPI comes first. Because PCE borrows much of its price data from CPI and PPI, forecasters can estimate PCE fairly accurately once those two are published.
• Expect a gap. CPI inflation has usually been a few tenths of a percentage point higher than PCE inflation, though the gap varies, and it widens when shelter costs are rising fast.
• Compare like with like. Year-over-year CPI and month-over-month PCE are not comparable. Check which measure and which period a headline uses.
• Think about what you buy. Neither index is anyone's personal inflation rate. A renter in an expensive city, a retiree with high medical costs and a commuter with a long drive will all experience different price changes from the average.

The bottom line

CPI is the better measure of out-of-pocket price changes for urban households and is the one used for cost-of-living adjustments. PCE is broader and adapts faster to changes in spending, which is why the Fed sets its target in terms of PCE. Watching both, and their core versions, gives a fuller picture than either one alone.