PCE vs CPI: Why America Has Two Inflation Numbers — And Why the Gap Between Them Has Never Mattered More

On June 10, 2026, the Bureau of Labor Statistics reported that the Consumer Price Index rose 4.2% over the previous twelve months. On May 28, the Bureau of Economic Analysis reported that the Personal Consumption Expenditures price index rose 3.8% over the previous twelve months. Both numbers measure US inflation. Both are produced by the US government. And they disagree by nearly half a percentage point — a gap that, in the current interest-rate environment, is worth hundreds of billions of dollars in monetary policy and trillions in global currency valuations.

The Federal Reserve, which has more influence over the global economy than any other single institution, ignores the higher number. It targets the lower one. The Fed's 2% inflation target, set in January 2012, refers to PCE — not CPI. This means the policy rate that determines borrowing costs for every country on earth, compresses Japan's GDP by depressing the yen, and sets the floor for emerging-market interest rates, is calibrated to a number that most Americans have never heard of. Understanding why requires understanding what each measure actually captures — and what it misses.

What Each Measure Actually Measures

The CPI, published monthly by the Bureau of Labor Statistics since 1913, is the older and more familiar measure. It tracks the price of a fixed basket of goods and services purchased by urban consumers, based on household surveys of what Americans report buying. When you read a headline about “inflation,” it almost always refers to the CPI.

The PCE price index, published monthly by the Bureau of Economic Analysis as part of its personal income and outlays report, takes a fundamentally different approach. Instead of surveying consumers about what they buy, it surveys businesses about what they sell. And instead of measuring only out-of-pocket costs, it includes spending made on behalf of consumers by third parties: employer-paid health insurance premiums, Medicare and Medicaid payments, and employer contributions to retirement plans. If your employer pays $18,000 per year for your health insurance and that premium rises 8%, CPI does not see it. PCE does.

The scope difference alone explains much of the divergence. CPI is narrower — it measures what consumers pay out of pocket. PCE is broader — it measures all consumption, regardless of who pays. In a country where employer-sponsored health insurance covers 155 million people, this distinction matters enormously.

The Three Technical Divergences

Beyond scope, three methodological differences drive the gap between CPI and PCE.

Difference 1: Shelter weighting.CPI assigns approximately 36% of its basket to shelter costs — rent and the imputed rent that homeowners would pay if they rented their own homes (owners' equivalent rent, or OER). PCE assigns roughly 15%. When rent inflation is high, CPI runs significantly above PCE. When rent inflation is moderate, the gap narrows. In 2026, shelter inflation is sticky at around 5.4% in CPI, which alone contributes about 1.9 percentage points to headline CPI — nearly half the total 4.2% reading. In PCE, the same shelter inflation contributes only about 0.8 percentage points, because of its lower weight.

Difference 2: Healthcare coverage. PCE captures healthcare spending that is invisible to CPI: the $18,000 annual premium your employer pays, the Medicare reimbursement for a hospital visit, the Medicaid-funded prescription. Total US healthcare spending exceeds $4.5 trillion per year — about 17.6% of GDP. Only a fraction of that shows up in CPI. All of it shows up in PCE. When medical costs rise rapidly, as they have in 2026 (post-pandemic catch-up in hospital costs, Medicare Part D drug pricing changes, insurance premium adjustments), PCE's core measure can actually exceed CPI's core measure — an unusual reversal.

Difference 3: The substitution effect.CPI uses a fixed basket: if the price of beef doubles, CPI assumes you keep buying the same amount of beef. PCE uses a chain-weighted formula that allows the basket to shift: if beef doubles, PCE assumes some consumers switch to chicken, and adjusts the weights accordingly. This makes CPI structurally biased upward — it overstates the cost of living for consumers who adapt to price changes. The BLS addresses this partially with the Chained CPI (C-CPI-U), but headline CPI still uses the older formula.

The 2026 Divergence: A Tale of Two Cores

In normal times, the headline CPI-PCE gap hovers around 0.3–0.4 percentage points, with CPI higher. In 2026, the headline gap is roughly in that range: CPI at 4.2%, PCE at 3.8% (April, with May expected at 4.1%). But the core measures — which strip out volatile food and energy prices — tell a more unusual story.

Notice the reversal in the core measures: core CPI is 2.9% while core PCE is 3.3%. Normally, core CPI runs above core PCE. In 2026, the relationship has flipped. This is partly because shelter inflation, which dominates core CPI, is beginning to moderate as new lease rent growth slows. And partly because healthcare costs, which dominate core PCE, are accelerating: hospital costs are rising as deferred pandemic-era procedures are finally completed, Medicare Part D drug prices are being renegotiated, and insurance premiums have adjusted upward to reflect two years of elevated medical utilisation.

The result is that the inflation measure the Fed watches — core PCE — is stubbornly higher than the measure that appears in headlines. At 3.3%, core PCE is 65% above the Fed's 2% target. This arithmetic helps explain the hawkish shift in the June dot plot: the Fed revised its year-end PCE forecast from 2.7% (March projection) to 3.6% (June projection) — the largest single-meeting upward revision since the inflation surge began. Nine of eighteen FOMC officials now project at least one rate hike before year-end.

Why It Matters Beyond the United States

The choice of inflation measure might seem like an arcane methodological debate. It is not. The Fed's decision to hold rates at 3.50–3.75% — with dot-plot projections suggesting a possible hike to 4.00–4.25% later this year — is calibrated against PCE, not CPI. If the Fed targeted CPI instead, with its higher shelter weight pushing the headline reading to 4.2%, the argument for immediate rate hikes would be even stronger. Conversely, if it targeted core CPI at 2.9%, the argument for rate cuts would be plausible.

The choice of PCE has global consequences. US rates at 3.50–3.75% — the highest among G7 nations — keep the dollar strong, which compresses the dollar-denominated GDP of every country on earth. The yen at 160 per dollar, driven partly by the interest-rate differential with Japan, reduces Japan's GDP by over $600 billion relative to where it would be at 2021 exchange rates. The Indian rupee's depreciation, amplified by dollar strength, makes India's $4.15 trillion nominal GDP look smaller than its domestic economy warrants. Emerging-market central banks face an impossible choice: match US rates (crushing domestic growth) or let their currencies depreciate (importing inflation through dollar-priced commodities like oil and wheat).

All of this traces back to one technical decision made in 2000: that the Federal Reserve would judge inflation by PCE, not CPI. The 0.3–0.4 percentage-point average gap between the two measures, compounded through interest-rate decisions, exchange-rate movements, and GDP rankings, ripples outward until it touches every economy on the planet.

What to Watch: The May 2026 PCE Release

The Bureau of Economic Analysis is scheduled to release the May 2026 PCE data on June 25. The consensus forecast, compiled by Morningstar and Bloomberg, expects headline PCE to rise to approximately 4.1% year-over-year — up from 3.8% in April and the highest reading since April 2023. Core PCE is expected to hold at 3.3%, unchanged from April.

If the 4.1% figure confirms, it will mean that the Fed's preferred inflation measure is running at more than double its 2% target, despite rates already at a 20-year high. This is the data point that will determine whether the dot plot's hawkish tilt translates into an actual rate hike at the September meeting. The same release will include Q1 2026 GDP's final estimate, currently at 1.6% annualised — the combination of rising inflation and slowing growth that economists call stagflation.

Economists will also parse the monthly data: a month-over-month increase of 0.5% in headline PCE (versus 0.4% in April) would signal acceleration. A core monthly gain above 0.3% would reinforce the argument that inflation is re-broadening beyond energy into services — the stickiest component and the one that central bank rate increases are least effective at controlling.

The Deeper Question

The gap between CPI and PCE in 2026 is, at one level, a technical matter of basket weights and formula choices. At another level, it reveals something about what we mean by “inflation” — a concept that feels precise but is fundamentally a question of perspective. CPI answers: what does it feel like to be a consumer paying bills? PCE answers: what does it cost to sustain American consumption, including the parts someone else pays for?

In 2026, those two answers diverge in the core measures for the first time in years. If you experience inflation as a consumer paying rent and buying groceries, core inflation is 2.9% — elevated but not alarming. If you measure inflation as an economist tracking the full cost of American consumption including healthcare, it is 3.3% — stubbornly high and nowhere near the Fed's target. The Fed, by design, looks at the latter. Markets react to the former. And the resulting confusion — “Is inflation high or moderating?” — is not a failure of communication. It is a genuine ambiguity built into the measurement itself.

Neither CPI nor PCE is wrong. They measure different things, and in 2026 those things are moving in different directions. The one thing that is clear: at 3.8% headline and 3.3% core, the Fed's preferred measure is nowhere near its 2% target, and the American consumer is running out of savings to bridge the gap.

Explore the data behind this analysis: Inflation by Country • United States Economy • Healthcare Spending by Country • Country Comparison Tool.