Why PCE Is Closer to Real Inflation Than CPI

If you've watched US inflation headlines, you've probably seen two numbers swap in and out: CPI and PCE. CPI drops and the news goes loud. PCE comes out a few days later and barely moves the headlines. But here's the twist — when the Fed actually decides on policy, it only looks at PCE. Same kind of gauge, different starring role. Let's walk through why both exist and why the Fed picks PCE.

---

CPI and PCE — what's actually different?

Start with the names. CPI is the Consumer Price Index, published every month by the BLS, the labor statistics arm of the US government. PCE is the Personal Consumption Expenditures Price Index, published monthly by the BEA, which sits under the Commerce Department.

Both gauges are trying to answer the same question: how much have the prices of stuff people buy gone up, on average? They start from the same place. But the way they measure ends up pretty different.

Two big differences. First, how often the weights refresh. CPI updates its weights once a year. PCE updates them every single month. Second, what they cover. CPI counts only the money urban households pull out of their own wallets. PCE goes wider — it also includes things like the health insurance your employer pays on your behalf.

Combine those two differences and the same month of inflation shows up as slightly different numbers in each gauge.

---

Why does updating weights monthly matter?

Weights in a price index are basically the ratios for "how much should each category count." Should gasoline get 5% of the weight or 7%? That kind of question.

Where people actually spend money changes month to month. Gas gets expensive, so people drive less. Dining out gets pricey, so they cook at home more. To measure real inflation, the weights have to chase those shifts.

CPI refreshes weights once a year. So whatever ratios were locked in last December stick around until next November. Even as habits shift mid-year, the index keeps computing with the old mix. PCE refreshes monthly. If dining out shrinks this month, next month's calculation reflects it right away.

That's why PCE tracks closer to what people are actually doing with their money right now.

---

The "steak-to-chicken" effect

Make it concrete. Picture a grocery run where beef prices suddenly jump 30%. What would you actually do? Most people switch to chicken or pork. Beef went up 30%, but your real grocery bill doesn't climb 30% because you found an alternative.

Economists call this the substitution effect. PCE bakes it in every month. If beef's share shrinks and chicken's share grows, that shift hits next month's calculation. The weight on the now-expensive item shrinks on its own.

CPI can't do that. It keeps beef at last year's 30% weight, so the same price jump shows up bigger. That's why CPI numbers usually print slightly higher than PCE.

Look at the March 2026 release: CPI ran about 3.5% year over year while PCE ran about 3.2% at the same moment. A 0.3-percentage-point gap sounds small, but when you're setting interest rates, that's a meaningful gap.

---

Why does the Fed only look at PCE?

The Fed officially picked PCE back in 2012. Before that, it weighted CPI too. Two reasons for the switch.

First, that substitution effect is more accurate. Monetary policy needs to react to inflation people are actually feeling. PCE gets closer to that experience.

Second, the coverage is wider. Things like the health insurance your employer pays on your behalf are still benefits you receive. PCE counts them. CPI doesn't. In the US, healthcare is such a huge chunk of spending that this gap matters.

Former Fed chair Alan Greenspan once called PCE "the best consumer price index by far." After that, PCE basically became the official policy yardstick.

The Fed targets 2% PCE inflation over the long run. So when headlines say "core PCE at 3.2%," markets read it as "still 1.2 percentage points above target — rate cuts won't come fast."

---

Is CPI useless then?

Not at all. CPI still has its job.

CPI directly drives Social Security adjustments, wage negotiations, and the coupons on inflation-protected Treasuries. The annual bump retirees get from the US government is pegged to CPI. CPI also publishes about two weeks earlier than PCE. So the market always sees CPI first.

The rhythm works like this. Mid-month, CPI drops and markets twitch. Toward month-end, PCE arrives and markets react again — this time guessing what the Fed will make of it.

Same month of inflation, looked at twice from different angles. CPI for everyday contracts and the market's first reaction. PCE for the Fed's policy call.

---

So what do you actually watch?

Three things will do.

First, if you care about everyday prices or wage and pension issues, watch CPI. It tracks closer to what you feel at the cash register.

Second, if you want to know when the Fed might cut or how markets might shift, watch PCE — especially core PCE, which strips out food and energy. That one is the direct policy yardstick.

Third, watch the months where the two gauges spread far apart. When CPI runs much hotter than PCE, the substitution effect did heavy lifting that month. People shifted away from the items that got pricey, and that's a real signal about the economy.

Headlines lead with CPI because it prints first and lands harder. The Fed picks PCE because it's more accurate. They're not competitors — they just do different jobs.

Next time an inflation number lands in the news, check which gauge it came from. The reason the same month can show different numbers will finally make sense.