PCEPILFE: Core US PCE Inflation Index Excluding Food and Energy — Monthly Since 1959

PCEPILFE is the Federal Reserve Bank of St. Louis identifier for the Core PCE Price Index — Personal Consumption Expenditures excluding food and energy, computed by the Bureau of Economic Analysis. Since 2000, PCEPILFE has been the Federal Reserve’s preferred inflation gauge and the basis for its 2% policy target, making it arguably the most consequential inflation number in the world. Unlike the CPI, PCEPILFE uses chain-weighted dynamic consumption weights and a broader spending basket that includes employer-paid health insurance and other imputed expenditures. Why this metric carries that weight is detailed in the complete guide to inflation.

Dataset: US Core PCE Inflation (1959–2026) · Updated 2026-06-01

Source: FRED series PCEPILFE · Bureau of Economic Analysis (BEA) via FRED

Macro Takeaway

Dataset Overview

Dataset Variables

The CSV and Excel files contain the following columns.

Column names match the CSV headers exactly.

Download the Complete Dataset

The full dataset is available in CSV and Excel formats.

FRED Direct CSV Access

The underlying data is available from FRED under series code PCEPILFE:

https://fred.stlouisfed.org/graph/fredgraph.csv?id=PCEPILFE

Direct CSV Access — Eco3min Structured Dataset

https://eco3min.fr/dataset/us-core-pce.csv

This URL returns the complete dataset in CSV format. It can be used directly in pandas, R, curl, or any data tool.

Using the Dataset in Python

import pandas as pd

url = "https://eco3min.fr/dataset/us-core-pce.csv"
df = pd.read_csv(url, parse_dates=["date"])

print(df.head())
print(df["core_pce_index"].describe())

Using the Dataset in R

library(readr)

url <- "https://eco3min.fr/dataset/us-core-pce.csv"
df <- read_csv(url)

head(df)
summary(df$core_pce_index)

Both examples load the dataset directly from the URL — no download or API key required.

Methodology

PCEPILFE is the Core Personal Consumption Expenditures Price Index, less food and energy, compiled by the Bureau of Economic Analysis as part of the monthly Personal Income and Outlays release. Unlike the CPI, the PCE deflator uses chain-weighting — basket weights update with each release based on observed consumption patterns — which captures consumer substitution in real time. The basket itself is also broader: PCE incorporates imputed consumption (such as employer-paid health insurance and financial services consumed through net interest margins) that CPI excludes.

The reference base period is currently 2017 = 100 following the comprehensive update of the National Income and Product Accounts. The methodology differs from CPI in three structural ways: dynamic vs fixed weights, broader vs narrower scope, and Fisher index vs Laspeyres aggregation. These three differences together explain most of the historical CPILFESL–PCEPILFE wedge.

This dataset is updated monthly via automated pull from the FRED API, typically within 24 hours of the BEA release. The accompanying year-over-year column is computed as core_pce_yoy = (PCEPILFE[t] / PCEPILFE[t-12] − 1) × 100.

Data Quality & Provider Notes

PCEPILFE is published by the BEA with a documented release calendar and incorporated into FRED within minutes of release. The series carries a higher revision risk than CPI because PCE is itself a derived measure constructed from multiple source inputs, including the CPI itself. Eco3min mirrors FRED with a daily pull.

• Release latency. The BEA publishes the Personal Income and Outlays report at 08:30 ET, typically on the last business day of the month for data from two months prior (one-month lag). FRED updates PCEPILFE within minutes of the release.
• Revisions policy. PCEPILFE is revised regularly. Each monthly release revises the prior two months, and the annual NIPA update (each July) revises several years back. This is fundamentally different from CPI, which is not revised after initial publication. Real-time analysis therefore should use ALFRED vintage data rather than the current FRED snapshot.
• Alternative sources. The same data is available directly from BEA (Table 2.4.4U for the core deflator) and through Bloomberg, Refinitiv, and Haver Analytics. The Dallas Fed Trimmed Mean PCE (FRED: PCETRIM12M680SFRBDAL) and the Cleveland Fed Median PCE provide alternative core measures with different filtering rules.
• Known gaps. Monthly series with no gaps since January 1959. Methodological changes (chain-weighting introduced in 1996, comprehensive NIPA updates roughly every five years) create structural breaks that are not flagged in the FRED file but matter for cross-vintage comparisons.

For Fed reaction function work, the relevant vintage is the one available to the FOMC at decision time — which is typically the latest release as of two business days before the meeting. ALFRED provides exactly this view.

Common Pitfalls When Using PCEPILFE

PCEPILFE is the most policy-relevant inflation measure in the US, but its construction differs enough from CPI that several recurring misinterpretations distort analysis.

1. Confusing core PCE (PCEPILFE) with core CPI (CPILFESL). The two series cover overlapping but distinct baskets, use different weighting schemes (chain-weighted vs biennially-updated fixed weights), and typically differ by 30–50 basis points. Treating them as interchangeable produces persistent bias in any model that maps inflation to policy or to TIPS pricing. See our analysis of the PCE vs CPI gap.
2. Reading PCEPILFE level when the analytical target is the YoY rate. The FRED series PCEPILFE is the price index, not the year-over-year change. The Eco3min dataset adds the precomputed YoY column (core_pce_yoy), but analyses that pull PCEPILFE directly from FRED and plot the level can mislead readers who expect an inflation rate.
3. Ignoring revisions and using current vintage for historical reaction functions. Because PCEPILFE is revised every month and again every July, the values available to the FOMC at any past meeting differ from what FRED shows today. Reproducing historical Fed decisions requires ALFRED vintage data — using the current FRED snapshot biases all real-time analyses.
4. Confusing PCEPILFE with PCEC96 or trimmed-mean PCE. PCEPILFE is the core price index. PCEC96 is real personal consumption (a quantity, not a price), and the Dallas Fed Trimmed Mean PCE is a separate filter-based core measure. Each series serves a different analytical purpose and they are not substitutes.

Historical Regimes

1959–1965 — Pre-acceleration baseline. PCEPILFE averaged 1.5–2% under the gold-anchored Bretton Woods regime. The series begins in January 1959, just before the inflation regime change that defined the next two decades.

1966–1982 — Great Inflation. PCEPILFE accelerated from 2% to a peak above 10% in 1980, reaching the inflation regime that ultimately forced the Volcker tightening (Fed Funds above 20% — see FEDFUNDS). Core PCE peaked slightly below core CPI in absolute terms during this episode, reflecting the chain-weighting effect that captured early consumer substitution toward less-inflated categories.

1983–2007 — Great Moderation. PCEPILFE stabilized in the 1.5–3% range as Fed credibility anchored expectations. The structural gap with core CPI stabilized around 30–50 basis points, a wedge that informed the Fed’s eventual shift from CPI-centric communication to PCE-centric communication.

2008–2020 — Persistent undershoot. PCEPILFE averaged 1.5% across the decade, undershooting the Fed’s newly explicit 2% target (formalized in 2012). The cycle low of 0.95% in April 2010 prompted multiple rounds of quantitative easing and ultimately the 2020 shift to Average Inflation Targeting, which allowed PCEPILFE to overshoot 2% to compensate for prior shortfalls.

2021–2022 — Post-pandemic overshoot. PCEPILFE reached 5.6% in February 2022 — the highest core PCE reading since 1983. The gap with core CPI compressed sharply as goods inflation hit both indices similarly, but reopened as services inflation drove the persistence phase.

2023–2024 — Sticky disinflation. PCEPILFE declined from above 5% toward 3% but plateaued, with services components remaining elevated even as goods PCEPILFE inflation collapsed. The disinflation path drove the Fed’s rate cycle visible in the FEDFUNDS series.

2025–2026 — Convergence phase. PCEPILFE continues to converge toward the 2% target. The relationship with 5-year breakeven inflation and with services-vs-goods components defines the dominant analytical questions for this phase.

Related Macroeconomic Datasets

Core PCE is the metric the Fed uses to anchor rate policy. The following datasets contextualize PCEPILFE against the broader PCE deflator, the alternative CPI-based core measure, the policy rate it influences, and the market-implied inflation expectations that compete with it.

Macroeconomic Dataset Hub

This dataset is part of the Eco3min macro-financial data repository.

Explore the Eco3min Dataset Hub

Sources

• Bureau of Economic Analysis — Personal Consumption Expenditures Price Index, Less Food and Energy
• Federal Reserve Bank of St. Louis — FRED series PCEPILFE

Dataset Reference

Eco3min Research. (2026). PCEPILFE: US Core PCE Inflation (1959–2026).<br> https://eco3min.fr/en/us-core-pce-inflation-dataset/