Core Inflation

Core inflation is the rate of increase in prices excluding food and energy. It is not a separate phenomenon from inflation and not a separate index family; it is the same price data with two volatile components taken out, in order to make the underlying trend easier to see. The entry on inflation covers what inflation is and how it is measured; this entry is about why a second version exists, who uses it, and where it misleads.

The name is worth pausing on, because "core" is a word neither agency that produces the numbers actually uses for them. The Bureau of Labor Statistics publishes an index it calls all items less food and energy; the series title carried by the Federal Reserve Bank of St. Louis for the same data reads "Consumer Price Index for All Urban Consumers: All Items Less Food and Energy in U.S. City Average." The Bureau of Economic Analysis labels its version the PCE price index excluding food and energy. BLS acknowledges the popular name at arm's length, writing that some users of its data "want to focus on what they perceive to be the 'core' or 'underlying' rate of inflation." The word belongs to the people reading the statistics rather than to the people producing them, and the Federal Reserve is the reader that made it stick.

The Fed's reason for looking at core is about signal, and it has been explained in plain terms on the record. Asked at the September 2023 press conference why the Committee looks past a rise in headline inflation driven by energy, the Federal Reserve's then-Chair, Jerome Powell, answered: "The reason why we look at core inflation, which excludes food and energy, is that energy goes up and down like that. And it doesn't—energy, energy prices mostly, mostly don't contain much of a signal about how tight the economy is, and hence don't tell you much about where inflation's really going." The Federal Reserve's July 2026 Monetary Policy Report puts the same idea in its own words, describing core PCE prices as excluding "often-volatile food and energy prices" and being "generally considered a better guide to future inflation developments."

Note what that reasoning does not say. It does not say food and energy prices matter less, or that they should be discounted when thinking about a household budget. In the same answer, Powell said the opposite about the household side: "energy prices are very important for the consumer. This, this can affect consumer spending. It certainly can affect consumer sentiment." Core exists because energy is noisy as a predictor, not because it is minor as a cost. Reading it as a claim about importance is the commonest error made with the number.

The Fed's target is not set on core, and this trips up careful readers. The Committee's Statement on Longer-Run Goals and Monetary Policy Strategy commits to "inflation at the rate of 2 percent, as measured by the annual change in the price index for personal consumption expenditures." That is the total index, food and energy included. Core is what the Committee watches on the way there, on the reasoning above, and total is what it has promised to deliver. So a core reading below 2 percent is not the goal being met, and a core reading above it is not the goal being missed; it is an input to a forecast about the number that counts.

There are two core measures, not one, and they are built differently. BLS produces a core version of the Consumer Price Index and BEA produces a core version of the PCE price index. Because the two parent indexes use different formulas, different weights and different definitions of what counts as consumer spending, their core versions normally report different rates for the same month, and neither is wrong. A source that says "core inflation was X" without saying which one has left out the part that would let a reader reconcile it against another source.

Excluding food and energy is not the only way to strip out noise, and it is a blunt one. The exclusion is defined by category, so it removes every food and energy price whether it moved that month or not, and leaves in any other price that happened to swing. Statistical agencies and Reserve Banks publish alternative trimmed and median measures that select by how much a price actually moved rather than by what it is: the Federal Reserve Bank of Cleveland's median Consumer Price Index and the Federal Reserve Bank of Dallas's trimmed mean PCE inflation rate are the two best known. Core is the version that became conventional, not the only defensible one.

The construction is subtraction, not a different survey. Both agencies collect prices for a full basket, then publish an additional aggregate that leaves out the food and energy components. Nothing else changes: the same collection, the same methods, the same publication date. BLS releases its version with the monthly Consumer Price Index and BEA releases its version with the monthly Personal Income and Outlays report. Current readings are at BLS.gov and BEA.gov, and this entry states none, because they change monthly while the mechanics do not.

A hypothetical example of why a household's experience diverges from the core number. Suppose food and energy account for 20 percent of what a household spends, and everything else accounts for the other 80 percent. Over a year:

• Prices of everything except food and energy rise 2 percent.
• Food and energy prices rise 10 percent.

Core inflation, which sees only the first line, reports 2 percent. The household's actual cost increase is the weighted average of both: 0.80 multiplied by 2 percent gives 1.6 percent, plus 0.20 multiplied by 10 percent gives 2.0 percent, for a total of 3.6 percent. The core reading is not wrong. It is answering a question about the trend in the 80 percent, and the household is living in all of it.

This is why a plan built on core inflation understates the growth in spending it needs to fund. A retirement projection, a college savings target or a withdrawal plan in retirement should be inflated at the headline rate, because groceries and gasoline are in the budget. Core is the right number for anticipating what a central bank is likely to do next, which is a different use with a different audience.

What core inflation is good for

• Filters out the two categories that swing hardest for reasons unrelated to how tight the economy is, which is why the Federal Reserve leans on it.
• Reduces the risk of reacting to a temporary energy shock as though it were a change in trend, in either direction.
• Published on the same schedule and from the same data as the headline number, so no extra source or lag is involved.

Where it misleads

• It is not a household's cost of living. Food and energy are real spending, and a plan inflated at the core rate understates future costs.
• The exclusion is by category, not by volatility, so it drops a stable food price and keeps a wildly moving one from any other category.
• Two different core measures exist, from two agencies, and they normally disagree for the same month.
• The Federal Reserve's 2 percent commitment is set on total inflation, so scoring the target against core compares the wrong series to the goal.
• Neither agency uses the word, which makes it easy to assume "core CPI" and "core PCE" are official product names rather than shorthand.