Home › Guides › PCE vs Core PCE: The Fed's Preferred Inflation Measure
The Federal Reserve's preferred inflation measure. When you hear 'the Fed's 2% target,' this is what they mean.
PCE stands for Personal Consumption Expenditures price index. It's another inflation measure, similar in spirit to CPI but published by a different government agency (the Bureau of Economic Analysis, inside the Commerce Department) using a different methodology. PCE is the inflation measure the Federal Reserve has officially used as its target since 2012. When you hear that the Fed has a "2% inflation target," the 2% applies specifically to PCE - not to CPI.
The Federal Reserve's preferred inflation measure. When you hear 'the Fed's 2% target,' this is what they mean.
PCE stands for Personal Consumption Expenditures price index. It's another inflation measure, similar in spirit to CPI but published by a different government agency (the Bureau of Economic Analysis, inside the Commerce Department) using a different methodology. PCE is the inflation measure the Federal Reserve has officially used as its target since 2012. When you hear that the Fed has a "2% inflation target," the 2% applies specifically to PCE - not to CPI.
PCE differs from CPI in a few important ways. First, its formula is more flexible - it captures the fact that when one product gets expensive, consumers buy more of a cheaper substitute. CPI's formula assumes consumers keep buying the same basket regardless of price changes; PCE's formula adjusts as relative prices shift. Second, PCE has a broader scope - it includes things you don't pay for directly, like employer-paid health insurance and many forms of medical care paid for by government programs. Third, PCE basket weights are updated more frequently (essentially every quarter) compared to CPI's once-a-year cadence (annual since January 2023, previously every two years).
The practical result of these differences: PCE usually runs about 0.3 to 0.5 percentage points below CPI on a year-over-year basis. So if CPI is 3.5%, PCE is typically 3.0% to 3.2%. That gap is structural, not random. It matters for policy because the Fed's 2% target on PCE is roughly equivalent to a 2.3–2.5% target on CPI - which means people watching CPI need to keep the gap in mind when judging whether the Fed is meeting its goal.
Core PCE - PCE with food and energy stripped out - is the most-watched single number in Federal Reserve policy discussions. Each FOMC meeting's policy statement, each Powell press conference, each Fed speech references Core PCE as the cleanest read on underlying inflation trend. It's published monthly by BEA, typically the last business day of the month at 8:30 AM Eastern, with a one-month lag.
A recent example: in early 2022 (February) Core PCE peaked around 5.5–5.6% year-over-year (vs Core CPI's roughly 6.6% peak). It then disinflated through 2023 and 2024 toward the 2% target, with the path repeatedly stalling around 3% as service-sector inflation proved stickier than goods inflation. Watching that disinflation path month by month was the central drama of Federal Reserve policy in 2023–25.
BEA's Fisher chain-weighted price index; headline PCE is the Fed's official 2% target (since 2012), with core PCE the operational near-term gauge.
PCE is published monthly by the Bureau of Economic Analysis (BEA) as part of the Personal Income and Outlays release. Release timing is the last business day of each month at 8:30 AM Eastern, with the data referring to the previous month (so the May release reports April data). The release also includes Personal Income, Disposable Personal Income, Personal Consumption Expenditures (nominal and real), and the Personal Saving Rate - see the Savings Methodology explainer.
The headline PCE price index is FRED ticker PCEPI; Core PCE (excluding food and energy) is PCEPILFE. "LFE" in the FRED ticker stands for "Less Food and Energy." PCE inflation is typically reported as the year-over-year percent change in PCEPI or PCEPILFE.
The key methodological difference from CPI is the formula. CPI uses a modified Laspeyres formula, which holds the basket roughly fixed and captures only within-category substitution. PCE uses a Fisher ideal formula, which is the geometric mean of Laspeyres and Paasche indices and captures consumer substitution across categories as relative prices change. The Fisher formula is theoretically superior because it doesn't assume consumers keep buying the same basket when prices change - empirically, consumers do substitute toward cheaper alternatives, so PCE captures that behavior.
The basket weights for PCE come from BEA's National Income and Product Accounts (NIPA), updated quarterly. CPI's weights come from the Consumer Expenditure Survey (CEX), updated annually since January 2023 (previously every two years). PCE's quarterly weight updates mean the basket adapts more quickly to changing consumption patterns. This matters during regime changes - during COVID, PCE adapted to the shift to durables and away from services faster than CPI did.
Scope differences: PCE includes employer-paid health insurance (CPI doesn't), Medicare and Medicaid (CPI excludes government-paid medical), certain financial services (e.g., bank-account services valued at the spread between deposit interest and lending rates), and items consumed on households' behalf (e.g., the imputed value of nonprofit-provided services). CPI is narrower - it measures only what households pay out-of-pocket. The expanded PCE scope is one reason PCE inflation is structurally different from CPI inflation.
The Fed's 2% target on PCE inflation has been formal policy since the January 2012 Statement on Longer-Run Goals and Monetary Policy Strategy. Before 2012 the Fed targeted price stability without an explicit numerical target - the 2% target was the formal articulation of what "price stability" meant. The 2% target applies specifically to headline PCE on a longer-run average basis. Core PCE is the operational measure the Fed focuses on for near-term policy because it filters short-term noise.
The August 2020 framework revision (the "Average Inflation Targeting" or AIT framework) modified the 2% target - periods of inflation above 2% could be allowed to compensate for periods below 2%, so that average inflation over time would tend toward 2%. The AIT framework was the institutional response to the post-GFC decade of persistent below-target inflation, and it informed the Fed's relatively patient approach to allowing 2021–22 inflation to overshoot before tightening aggressively in 2022. The AIT era ended in August 2025, when the Fed's framework review concluded with a revised Statement on Longer-Run Goals that dropped the averaging language while reaffirming the 2% headline-PCE target.
What moves PCE: ultimately the same drivers as CPI (energy, food, shelter, core services, supply shocks, monetary policy, inflation expectations) but with different magnitudes because of the scope and weight differences. PCE energy weight (~5%) is smaller than CPI (~7%); PCE medical-care weight is much larger because of the employer/government coverage inclusion; PCE housing weight is much smaller (roughly 15–18% of core PCE vs 40%+ of core CPI) because PCE's broader scope dilutes shelter's share.
PCEPI / PCEPILFE; Fisher chain weights updated quarterly; the 2020 AIT framework was retired in the August 2025 framework review.
PCE construction overview: BEA computes PCE prices by combining detailed source data from numerous price indices (CPI components, PPI for some categories, BLS Employer Costs for Employee Compensation, Medicare and Medicaid reimbursement rates) using a Fisher ideal formula chained on quarterly NIPA expenditure weights. The result is a Fisher chain index - the standard formula for national-accounts deflators.
FRED tickers: PCEPI (headline) and PCEPILFE (core); month-over-month percent changes are obtained by applying FRED's percent-change transformation to those series rather than via separate tickers. For sub-aggregates: DGDSRG3M086SBEA (PCE goods price index) and DSERRG3M086SBEA (PCE services price index). The BEA's underlying NIPA tables (Table 2.3.4 "Price Indexes for Personal Consumption Expenditures by Major Type of Product" - monthly version Table 2.4.4 - and Table 2.4.4U at the detailed-product level) provide finer disaggregation than FRED.
PCE vs CPI gap decomposition: historically PCE runs ~0.3–0.5 pp below CPI on year-over-year inflation. McCully-Moyer-Stewart (2007, BEA Survey of Current Business) decomposition separates formula effects (Fisher vs Laspeyres + chain weights), weight effects (BEA's quarterly NIPA weighting captures recent consumption shifts faster than CPI's CEX-based weights, updated annually since 2023 and biennially before that), and scope effects (medical-care coverage, financial services valuation, OER vs imputed shelter differences). The components partially offset in some periods, and the relative contribution of each shifts by year and economic regime.
Medical care: PCE medical care is the single largest scope difference. CPI medical care captures only the out-of-pocket portion of medical-care prices. PCE captures the full societal cost - employer-paid health insurance premiums (priced through BLS ECEC compensation surveys), Medicare/Medicaid reimbursements, hospital costs to providers, pharmaceutical costs to wholesalers. The PCE medical-care weight is roughly 17% vs CPI's roughly 8%. Different inflation rates by component create persistent PCE-vs-CPI divergence in medical care specifically.
Housing: BEA prices PCE housing services using BLS's CPI rent and owners'-equivalent-rent indexes as source data, so the two indexes' housing inflation rates track closely. PCE housing weight is much smaller than CPI's - roughly 15–18% of core PCE versus 40%+ of core CPI - which makes hot shelter inflation a major source of CPI-PCE divergence. Census and American Community Survey housing-stock and rent data determine the PCE housing weight, not an independent price imputation; both indexes inherit the 12–18 month lag of the BLS rental sample discussed in the CPI explainer.
Month-over-month variability: PCE M/M tends to be slightly less volatile than CPI M/M because of the Fisher-chain-weighting smoothing and the broader scope. M/M Core PCE is the FOMC's primary near-term inflation indicator - successive months below the 0.2% pace would point to disinflation back toward the 2% annualized target.
Average Inflation Targeting (AIT) framework: announced by Chair Powell at Jackson Hole in August 2020, codified in the Federal Reserve's revised Statement on Longer-Run Goals (August 27, 2020). Operational implication: periods of inflation above 2% can be allowed to compensate for periods below 2%, with the goal of achieving inflation averaging 2% over time. The framework gave the FOMC latitude to be patient in 2021–22 as inflation rose, on the theory that the prior decade of below-target inflation justified some overshoot. AIT did not survive the 2022–23 inflation episode: the 2024–25 framework review concluded on August 22, 2025 with a revised Statement on Longer-Run Goals that dropped the averaging makeup strategy and the "shortfalls" framing while reaffirming the 2% headline-PCE target.
Longer-run inflation expectations: the Fed monitors several survey-based measures (Michigan, NY Fed SCE, SPF) and market-based measures (TIPS breakevens) alongside realized PCE. The Fed staff economic outlook (Tealbook A materials, declassified after 5 years) includes these expectation measures as inputs to policy advice.
Real-time data issues: PCE data is revised. The first estimate published in the Personal Income and Outlays release is the "advance" estimate; subsequent monthly releases include revisions as new source data becomes available. Annual NIPA updates (released in late September since 2023, previously late July) can change historical PCE levels - the 2023 comprehensive update, released September 28, 2023, revised core PCE inflation over 2017Q4–2022Q4 up by about 0.1 pp. Revisions can shift the policy narrative around the disinflation pace, though usually modestly.
Research caution: comparing real-time-vintage PCE to current-vintage PCE for the same period requires using ALFRED (FRED's archival vintage database) to retrieve historical first releases. Modeling policy decisions in real time requires the vintage that was available to the FOMC at the meeting; post-revision data can produce hindsight bias.
The Fed's 2% target applies to year-over-year headline PCE on a longer-run average basis; near-term policy reactions are typically framed around Core PCE (the operationally-relevant measure). The distinction matters for press-conference parsing: when Chair Powell says "inflation is too high," the operational measure is usually Core PCE, but the target is technically headline.
FRB/US PCE component disaggregation; supercore (services ex-housing) is the wage-driven persistent piece.
Supercore PCE - core services excluding housing - has been the Fed's most-emphasized analytical decomposition since Chair Powell's November 2022 "Inflation and the Labor Market" speech at the Hutchins Center at Brookings. Powell explicitly identified supercore (also called "non-housing core services" or "core services ex-housing") as the most policy-relevant slice because of its tight link to wage growth and its historical stickiness in disinflation episodes.
The standard decomposition: total PCE = core PCE + food + energy. Core PCE = core goods + housing services + non-housing core services (supercore). Core goods (PCE: ~25% of core basket) is sensitive to supply shocks, tariffs, dollar strength, semiconductor cycle. Housing services (~17% of core) is driven by OER and tenant rent, lagged versus market rents. Supercore (~55–57% of core) is wage-driven and includes financial services, transportation services, recreation services, healthcare services, education services, food services and accommodations.
The supercore-wages link: empirical work (Bobeica, Ciccarelli & Vansteenkiste at the ECB; multiple Federal Reserve staff papers) shows supercore inflation is closely correlated with unit labor cost growth in the services sector. When wage growth is running 5–6%, supercore inflation tends to run 4–5%; sustained 2% supercore requires wage growth in the 3.5–4.5% range (assuming 1–1.5% productivity growth). The 2022–24 disinflation challenge was that wage growth (measured by Average Hourly Earnings, the Employment Cost Index, or the Atlanta Fed Wage Growth Tracker) decelerated more slowly than goods inflation, leaving supercore elevated long after goods-PCE turned negative.
The Fed staff's published PCE breakdown alongside FOMC briefing materials (Tealbook A and Tealbook B, declassified 5 years after the relevant meeting) shows the supercore decomposition along with cyclical/acyclical decompositions, wage-share decompositions, and forecast paths. The Tealbook discussions of inflation in 2022–24 are public via the FRASER archive (with the 5-year lag).
Annual NIPA revisions and policy implications: BEA's annual update - comprehensive roughly every five years - now lands in late September and can revise historical PCE. The 2023 comprehensive update (released September 28, 2023) revised core PCE inflation over 2017Q4–2022Q4 up by about 0.1 pp. The 2024 annual update (released September 26, 2024) further refined the 2022–24 picture. Each revision can shift the policy narrative - the Fed's 2024 communication around "inflation is moving back toward 2%" depends partly on whether you're using pre-revision or post-revision data.
Cyclical-vs-acyclical decomposition for inflation: Mahedy and Shapiro (Federal Reserve Bank of San Francisco, 2017), extended in Adam Shapiro's ongoing SF Fed series, split core PCE into a part responsive to labor-market tightening and a part driven by external factors; Stock and Watson's "cyclically sensitive inflation" work is a related but distinct construct. The cyclical component tracks employment-cost-index-related dynamics; the acyclical component captures supply shocks and trend factors. The decomposition is the analytical workhorse for asking "is the remaining inflation gap something monetary policy can close, or is it structural?"
FRB/US model: the Federal Reserve Board's flagship macroeconometric model has detailed PCE-component pricing equations (for goods, housing, supercore, food, energy) calibrated on historical NIPA data. Forecasts from FRB/US (publicly available via the FRB/US website) provide the model-based reference for Fed staff inflation projections. The FRB/US PCE forecast architecture is documented in Federal Reserve Board working papers (Brayton-Tinsley series, multiple authors).
International comparison: euro-area HICP (Harmonized Index of Consumer Prices) and UK CPI are the foreign analogs to U.S. CPI; the euro-area uses HICP excluding food and energy as the operational core. The ECB aimed for HICP inflation "below, but close to, 2%" from 2003 and adopted a symmetric 2% HICP target in its 2021 strategy review. Cross-country inflation comparisons need careful methodology alignment - HICP excludes owner-occupied housing (a major methodological gap that the ECB has been working to close for years).
Research caveats for academic use: (1) use ALFRED-vintage data to study real-time policy decisions; (2) distinguish first-release PCE from revised PCE; (3) cross-reference NIPA revisions when working with historical periods; (4) be careful with seasonal adjustment vintage - BEA's X-13 seasonal factors are revised periodically; (5) decomposition-component weights shift with each annual revision, complicating multi-year decomposition continuity.
The Fed's 2024–25 monetary-policy framework review concluded on August 22, 2025 with a revised Statement on Longer-Run Goals - the biggest institutional event for PCE-based policy since 2020. The revision dropped the average-inflation-targeting makeup strategy and the "shortfalls" framing adopted in 2020, returning to a flexible inflation-targeting approach closer to the original 2012 statement. The 2% headline-PCE target itself was reaffirmed; only the framework for achieving it changed.
For a comprehensive practitioner reference, the BEA NIPA Handbook (Chapter 5 for PCE-specific methodology), the Federal Reserve's Statement on Longer-Run Goals and Monetary Policy Strategy (current version August 2025), and Federal Reserve staff working papers on inflation analysis are the canonical sources.