CPI vs. PCE: two ways to measure inflation

Two inflation numbers come out every month. The Consumer Price Index (CPI) from the Bureau of Labor Statistics arrives first, around the middle of the month. The price index for personal consumption expenditures (PCE) from the Bureau of Economic Analysis follows about two weeks later. They measure the same thing in different ways, and the gap between them is usually a few tenths of a point.

Whose spending they measure

CPI tracks what urban households pay out of their own pockets. PCE covers everything households consume, including what others pay on their behalf, such as employer-provided health insurance and Medicare and Medicaid payments to doctors and hospitals.

That changes the weights. Medical care counts for far more in PCE than in CPI. Housing works the other way: shelter is about a third of CPI but closer to a sixth of PCE, so rent increases move CPI more.

How the weights move

CPI updates its spending weights once a year. PCE is a chained index whose weights shift every month as people switch from items that got more expensive to ones that didn’t. That substitution, along with the scope differences, tends to make PCE inflation run lower than CPI inflation over time.

Which one matters for you

The Federal Reserve defines its 2 percent inflation goal in terms of PCE, so Fed officials watch it most closely. Most things that touch your money use CPI: Social Security raises follow CPI-W, I bond rates and TIPS follow CPI-U, and the IRS adjusts tax brackets with chained CPI-U.

Both indexes also come in a core version that leaves out food and energy. Those prices swing with weather and oil markets, so core figures give a steadier read on where inflation is heading, though food and energy are very much part of what households pay.