What is the Fed’s preferred inflation measure?

HEADLINES ABOUT inflation in America typically refer to the country’s consumer-price index (CPI), the most widely used measure of changing prices. CPI inflation slowed in August to 2.5% year-on-year. But when America’s central bankers meet on September 17th to discuss cutting interest rates, they will focus on a different index. Since 2000 the Federal Reserve has used the personal-consumption-expenditures (PCE) price index, rather than the CPI, as its preferred measure of inflation. It is against this that the Fed’s target for inflation, 2%, is compared. What are the differences between the measures—and why does the Fed use the PCE?

Both CPI and PCE inflation are calculated by tracking the prices of a representative sample of goods and services, often referred to as a “basket”. Items in the basket are weighted according to the relative importance of each category.

One advantage of PCE is that it reflects changes in consumer spending quickly. CPI weightings are updated annually, and reflect spending from two years prior; PCE’s are adjusted each month. During the covid-19 pandemic it reflected the fact that lockdowns were causing people to spend less on services, such as eating out, and more on goods, like electronics. The frequent tinkering with the PCE means that it rapidly captures substitution effects: if, say, the price of cars soars, consumers might switch to buses, and the PCE would afford more weight to ticket prices the next month. The CPI, on the other hand, would assign the same emphasis to the rising price of cars until its weights were updated. That is one of the reasons that CPI inflation tends to be higher than inflation as measured by the PCE index (see chart).

A second difference is that the PCE is broader. The CPI only tracks spending in urban areas; the PCE index also includes purchases made in rural areas. And the CPI’s weights are only based on what consumers spend themselves—that does not include health insurance provided by their employer, say, or access to Medicare, which is funded by the government. The PCE includes all this. These differences affect the weightings of the indices’ baskets: in May 2024 energy accounted for 7% of the CPI, for example, but only 4% of the PCE index. When the price of a particular component rises fast, the gap between the two measures widens. In 2022, when the war in Ukraine sent global energy prices up sharply, the gap in the year-on-year change between the two indices grew to around two percentage points.

So why has the CPI stuck around? For one thing, new data typically come out two weeks ahead of the PCE measure. A hot CPI reading is almost always an indication that the PCE gauge will also be hot. And when the producer-price index, another measure of inflation that reflects manufacturers’ sales prices, is published, usually a day after the CPI, economists can combine both to predict with great precision what the PCE reading is likely to be. Finally, the CPI is arguably better at reflecting out-of-pocket perceived costs for consumers, and does a good job of capturing “sticker shock”. The post-pandemic surge in prices has forced Americans to pay close attention to inflation figures. But not all measures are created equal. ■