CPI vs. PCE: Why the Fed's Inflation Gauge Isn't the Famous One

Two federal agencies measure US inflation every month. Financial television counts down to one of them. The Federal Reserve targets the other. The number that makes headlines and the number that actually steers policy come from different reports.

Two reports, two agencies

The Consumer Price Index comes from the Bureau of Labor Statistics, which publishes it monthly, usually in the second week of the following month. Our guide to reading a CPI print covers what lands on that day. The Personal Consumption Expenditures price index, or PCE, comes from a different agency, the Bureau of Economic Analysis, and arrives later — tucked inside the monthly Personal Income and Outlays report near the end of the month. Both track the prices households pay, and both publish a “core” version that strips out food and energy. From there they diverge.

What each one counts

The biggest difference is coverage. CPI measures what urban consumers pay out of pocket. PCE casts a wider net: it also counts spending made on a household’s behalf, most importantly medical care paid by employers and by government programs like Medicare and Medicaid. That single choice reshapes the basket. Health care carries a much heavier weight in PCE than in CPI, because most of the country’s medical bill is not paid at the register.

Where the weight sits

Housing tilts the other way. Shelter (rent, and the rental-equivalent cost of owning a home) makes up roughly a third of CPI, but well under a fifth of PCE. So a hot rent reading pushes CPI around far more than it moves PCE. When the two indexes disagree in a given month, the split between housing and health care is often the reason.

The formula gap

The two indexes are also built differently. CPI uses a fixed basket of goods whose weights are refreshed only periodically. PCE uses a chained formula that updates weights more often, which lets it capture substitution: when steak gets expensive and shoppers buy more chicken, PCE registers the shift sooner. The practical effect is steady. CPI inflation tends to run a few tenths of a percentage point above PCE inflation year over year. PCE is also revised as fuller data comes in, while CPI is generally left as first reported.

Why the Fed watches PCE

The Fed picked the broader gauge. Its 2% inflation goal, formalized in January 2012, is defined in terms of PCE, and core PCE in particular is the measure policymakers cite most when they explain a rate decision. Wider coverage and more current weights make PCE, in the Fed’s view, a cleaner read on the underlying price trend.

One number moves the tape, the other grades the Fed

This leaves anyone following markets with two figures to track, on two different dates. CPI is the market event: it lands first, trades hard in the minute after release, and sets the tone. It also previews the other report, because the Bureau of Economic Analysis builds much of the PCE index from the same consumer-price and producer-price source data. A CPI surprise often signals which way PCE will lean weeks before it prints. PCE is the scorecard. It arrives quietly weeks later, but it is the figure the Fed measures itself against when it decides whether inflation is really back to target. Follow the loud one for the immediate reaction. Watch the quiet one to see what the Fed is actually thinking.

This is a general-market summary for information only — not investment advice, and not a recommendation regarding any security.