Both indexes track the price of a fixed idea, the cost of living, but they build it from different sources and update it on different schedules. The Consumer Price Index (CPI) comes from a survey of what households say they buy. The Personal Consumption Expenditures price index (PCE) comes from business receipts, and it covers everything households spend money on, whoever technically pays the bill, including spending made on a household's behalf by an employer or an insurer.

Why the weights move differently

The CPI holds its basket of goods fixed for a period and only revises it periodically. The PCE lets its basket shift each month as spending patterns change, so if beef prices rise and shoppers switch toward chicken, the PCE picks that substitution up sooner. Over any stretch of years this alone is enough to pull the two measures apart, even before the coverage differences are counted.

A gap that does not close

Because the two indexes are built from different data and different weighting rules, there is no reason to expect them to land on the same number in any given month or year, and they have not. A reader comparing a CPI figure against a PCE figure for the same period is comparing two different constructions of the same idea.

The mechanism behind each index, and the document where the Federal Reserve explains its own choice of the PCE, is covered in the sections that follow. The glossary entry for the dual mandate sets out what the Fed is weighing inflation against when it reads either series.

The report itself

In a Monetary Policy Report to Congress in 2000, the Federal Open Market Committee set out why it had started to watch the Personal Consumption Expenditures price index when judging inflation. It named four specific ways the two indexes are built differently, and on each one it judged the Personal Consumption Expenditures measure the better fit for setting policy. A reader who wants to see how a document like this is structured can work through How to Read a Central Bank's Own Words.

What separates the two indexes

The four differences all come down to questions of construction. Whose spending the index actually counts, how it treats a shopper who switches from a good that has become expensive to a cheaper substitute, which survey supplies the weight given to each category of spending, and how often those weights are refreshed to match current habits. The Consumer Price Index and the Personal Consumption Expenditures price index answer each of these questions differently, and the answers do not always pull the inflation reading in the same direction. The Central Banking Glossary carries a short entry on each term for a reader who wants the definition on its own.

Why it settled the question

No single one of the four points would have been reason enough on its own to change which index a central bank follows. Taken together, they gave the Federal Open Market Committee a measure that adjusts faster to how people are actually spending and that covers a wider base of spending, including some spending a household never sees a bill for. That is the case the 2000 report made, and it is the reasoning the Federal Reserve has stood behind since.

A saver checking their return usually reaches for the Consumer Price Index (CPI), published monthly and widely quoted in the press. The Federal Reserve, when it assesses whether inflation sits near its target, looks at the Personal Consumption Expenditures price index (PCE) instead. The two indexes measure the same broad thing, the rising cost of living, but they are built differently, so they rarely produce the same number for the same month.

Why the same month can show two different inflation rates

CPI is built from a fixed survey of what households say they buy. PCE is built from business sales data across the whole economy, covering spending done on a household's behalf by employers and government programmes as well as spending the household makes directly. PCE also updates its basket of goods more often, so it can pick up a shift toward cheaper substitutes sooner than CPI does.

The practical effect is that a saver comparing a savings rate against CPI, and the Fed weighing the same rate against PCE, can each be looking at a different real return for an identical account in an identical month. Neither figure is wrong; they answer slightly different questions about what a household's money bought.

Where to see the gap on your own numbers

The Real Return Calculator takes an amount, a period, and the interest rate earned, and returns the real value after CPI and after PCE side by side. Running your own savings rate through both measures shows whether the difference matters much in your case or barely moves the answer.

The Central Banking Glossary covers the dual mandate the Fed weighs PCE against, and What the Projections Are and What They Are Not explains how PCE inflation appears in the FOMC's own published forecasts.