CPI vs Core PCE

The Consumer Price Index (CPI) is published monthly by the Bureau of Labor Statistics. It measures the average change over time in prices paid by urban consumers for a market basket of goods and services. CPI is the headline inflation number covered by financial media, and it is the inflation measure used to adjust Social Security benefits, index tax brackets, and set TIPS (Treasury Inflation-Protected Securities) payouts.

The Personal Consumption Expenditures (PCE) price index is published monthly by the Bureau of Economic Analysis. It measures price changes in personal consumption expenditures across a broader definition of consumer spending that includes items paid by others on consumers' behalf, such as employer-provided health insurance and Medicare. PCE is the Federal Reserve's preferred inflation measure and the series the FOMC targets at 2 percent. The Fed specifically targets Core PCE, which excludes food and energy prices, to focus on underlying inflation trends.

CPI is limited to out-of-pocket spending by urban households. It excludes rural households, institutional populations, and spending paid by someone other than the consumer directly. Urban consumers represent approximately 93 percent of the US population, so the exclusion is small but real.

PCE captures a much broader scope: all personal consumption expenditures in the National Income and Product Accounts, regardless of who paid. This includes employer-paid health insurance premiums, Medicare benefits received by seniors, Medicaid benefits, and imputed financial services. Healthcare is the single largest area where the two measures differ, because employer and government healthcare spending is about three times larger than out-of-pocket healthcare spending. CPI weights healthcare at roughly 8 percent of the basket; PCE weights it at about 17 percent.

CPI uses a Laspeyres index formula with weights updated every two years. The Laspeyres formula holds the basket of goods fixed between updates, so if consumers substitute away from a good whose price is rising toward a cheaper alternative, the CPI does not capture that substitution. This is called substitution bias. The Boskin Commission famously estimated in 1996 that substitution bias alone accounted for roughly 0.4 percentage points of upward bias in the CPI relative to a true cost-of-living index.

PCE uses a Fisher Ideal chain-weighted formula with weights updated monthly or quarterly. The Fisher formula geometrically averages two indices (Laspeyres and Paasche) and captures substitution dynamics. When beef prices rise and consumers shift toward chicken, PCE's weights adjust to reflect the shift, while CPI's do not until the next weight reconstitution. This is the single largest methodological reason CPI tends to run higher than PCE across time.

Since 2000, annual CPI inflation has averaged approximately 0.4 percentage points higher than PCE inflation. That spread is not constant: it has been as high as 0.8 percentage points during periods of rapid consumer substitution (2008-2009, 2022) and as low as 0.1 percentage points during stable-price periods (2014-2015, 2019).

The sources of the spread are roughly: 0.4 pp from substitution bias (formula effect, CPI higher), 0.3 pp from scope differences with housing and healthcare weighting (CPI higher), minus 0.3 pp from other methodological differences including price-change measurement and small-item coverage (PCE slightly higher). Net, CPI runs approximately 0.4 pp higher on the typical year. In 2026 as of early readings, CPI is approximately 0.26 pp above Core PCE, near the historical lower bound of the typical spread.

The Federal Reserve formally adopted PCE as its preferred inflation measure in February 2000, shifting from its previous focus on CPI. The FOMC cited three reasons in its 2000 statement. First, PCE incorporates substitution effects through the Fisher formula, giving a more accurate picture of cost-of-living changes. Second, PCE has broader coverage of goods and services actually consumed, including items paid on behalf of households. Third, historical PCE data can be revised as better source data becomes available, so the entire time series remains consistent with current methodology, while CPI is fixed at publication.

The Fed targets Core PCE (excluding food and energy) rather than headline PCE because food and energy prices are volatile, often driven by global supply shocks, and do not necessarily reflect underlying inflation trends that monetary policy can meaningfully address. Powell and other FOMC members consistently reference Core PCE in speeches and the Summary of Economic Projections, making it the single most important inflation number for monetary policy decisions.

When CPI rises faster than Core PCE, the divergence typically reflects food and energy price shocks (which are in headline CPI but excluded from Core PCE), or rapid shifts in consumer behavior that the PCE formula captures but CPI does not. The 2022 inflation peak is a clear example: CPI peaked at 9.1 percent in June 2022 while Core PCE peaked at approximately 5.5 percent in March 2022, a spread of 3.6 percentage points at the extreme.

When Core PCE rises faster than CPI, the divergence typically reflects healthcare or services price increases that are weighted more heavily in PCE than in CPI. This has been less common but occurred briefly in 2014-2015 and in some months of 2024. When the two move in unison, inflation is broad-based and the policy signal is clearer. The current April 2026 reading shows them moving approximately in line, with the spread (CPI minus Core PCE) near 0.26 percentage points, close to the long-run average and suggesting underlying inflation is broadly consistent across both measures.

CPI is released on the second Tuesday or Wednesday of each month, at 8:30 AM Eastern, covering the prior month's prices. It is the earlier of the two releases and receives the larger market reaction because it is the first look at prior-month inflation. Trading in Treasuries, dollar futures, and equity index futures typically moves within seconds of the release if the print is meaningfully different from consensus.

PCE is released toward the end of the following month, typically four to five weeks after the month it covers. By then the market has already digested CPI, so PCE releases are usually less market-moving unless they diverge materially from what CPI-derived PCE nowcasts predicted. The Cleveland Fed's Inflation Nowcasting tool produces real-time PCE estimates updated after CPI, and large deviations from the Cleveland nowcast can move bond yields 5 to 10 basis points on the PCE print.

Both CPI and PCE have headline and core versions. Headline includes food and energy; core excludes them. The reason to exclude food and energy is that their prices are volatile and driven largely by global supply factors (OPEC, weather, geopolitical conflicts) that are outside the control of domestic monetary policy. In 2022 the gap between headline and core CPI peaked at 3 percentage points as oil and food prices surged following Russia's invasion of Ukraine. By 2024 that gap had closed to less than 0.5 percentage points.

The Fed focuses on core measures because removing volatile components reveals the underlying trend that monetary policy can influence. However, consumers experience headline inflation, so when food and energy are persistently high, it can feed into wage demands and eventually into core inflation through the services channel. Core measures lead in identifying inflation turning points; headline measures lead in setting consumer and political expectations about inflation.

The post-COVID inflation spike was the most severe in 40 years. CPI peaked at 9.1 percent in June 2022 and Core PCE peaked at 5.5 percent in March 2022. Both declined through 2023 as energy prices fell, supply chains normalized, and Fed rate hikes began to bite. By early 2024 CPI had reached 3.1 percent and Core PCE 2.8 percent, both still above the Fed target but clearly trending toward normalization.

Progress slowed in 2024-2025 as services inflation proved stickier than goods. CPI stalled in the 3.0-3.5 percent range through late 2024 while Core PCE stalled in the 2.7-3.0 percent range. The Fed nonetheless began cutting rates in September 2024 (100 basis points cut by year-end), accepting that achieving the 2 percent target would take longer than the initial disinflation phase suggested. As of March 2026, CPI sits at 3.26 percent and Core PCE at 3.0 percent, both still above target but with the 2026 Iran war creating renewed upside risk through oil prices.

The primary question is whether oil-price inflation from the 2026 Hormuz disruption feeds into core services inflation through second-round effects (wage demands, freight costs, services pricing). If it does, both CPI and Core PCE could reverse higher and force the Fed to pause or reverse rate cuts. If the inflation remains contained to headline measures (food, energy), Core PCE can continue drifting down toward target even while CPI is noisy.

Secondary signals to watch include the Cleveland Fed Trimmed-Mean PCE (which strips out the largest monthly outliers and typically provides the cleanest read on underlying trend), shelter CPI (which remains elevated at approximately 4.7 percent and is the single largest contributor to core inflation), and the Atlanta Fed Wage Growth Tracker (which shows wage pressure relative to the price target). The CPI-Core PCE spread itself is worth watching: a widening above 0.6 percentage points would indicate that substitution dynamics are accelerating, which the Fed typically interprets as a sign that consumers are successfully adapting to price pressures rather than a sign of uncontrolled inflation.