INDPRO: US Industrial Production Index Monthly Data from FRED (1919–2026)

The INDPRO series, published by the Federal Reserve Board (G.17 release) and distributed through FRED, measures real US output from manufacturing, mining and utilities at monthly frequency since 1919. With more than 1,280 continuous monthly observations, INDPRO provides the deepest hard-data record of the physical US economy, with peaks and troughs that have aligned closely with NBER recession dating across every post-war cycle.

Dataset: US Industrial Production Index (1919–2026) · Updated —

Source: FRED series INDPRO · Federal Reserve — Industrial Production and Capacity Utilization (FRED: INDPRO)

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 INDPRO:

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

Direct CSV Access — Eco3min Structured Dataset

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

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

Using the Dataset in R

library(readr)

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

head(df)
summary(df$indpro_index)

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

Methodology

INDPRO is constructed by the Federal Reserve Board from a combination of direct physical output measures (kilowatt-hours generated, tons of steel produced, barrels refined) and production-worker hours weighted by their typical contribution to value-added. Roughly 312 individual series feed the aggregate index. The current base year is 2017=100, with weights updated every five years from the Census Bureau’s Annual Survey of Manufactures and the Economic Census.

The index is seasonally adjusted and published monthly via the G.17 release, typically around the 15th business day after the reference month. Each release revises the three most recent months as additional source data become available. Annual revisions in early winter rework the prior several years; benchmark revisions every five years can revise the historical profile back to 1972 or earlier.

Data Quality & Provider Notes

INDPRO carries one of the longest continuous monthly histories in the Federal Reserve catalog. The series has been reconstructed back to 1919, with progressively higher granularity and revision quality from 1947 onward. The Eco3min mirror pulls INDPRO from FRED on a daily cadence; new prints reflect within 24 hours of G.17 release.

• Release latency. The Federal Reserve G.17 release publishes INDPRO around the 15th business day of the month following the reference period (e.g., March data appear in mid-April). Revisions to the prior three months accompany each release.
• Revisions policy. Routine three-month revisions at each release; annual revisions in late winter or early spring; benchmark revisions every five years that can shift INDPRO level and growth several percentage points over historical windows. For real-time vintage analysis, use FRED ALFRED.
• Alternative sources. Bloomberg (IP CHNG Index or INDPRO Index), Refinitiv/LSEG, and Haver Analytics carry the same underlying Federal Reserve data. The Federal Reserve also publishes Capacity Utilization (TCU) alongside INDPRO in the G.17 release; the two series read together provide a fuller cyclical signal than INDPRO alone.
• Known gaps. None monthly since 1919. Pre-1919 historical estimates exist (NBER Macrohistory database, Miron-Romer) but use different methodology and are not part of INDPRO.

Before running cross-decade comparisons, confirm the rebase generation in use — the 2017=100 base differs from earlier 2012=100 and 2007=100 vintages, and historical YoY growth rates are unchanged by rebasing but absolute levels are not directly comparable across vintages without rescaling.

Common Pitfalls When Using INDPRO

INDPRO is a clean series with a long history, but its scope and timing create predictable interpretation traps.

1. Treating INDPRO as a GDP proxy. INDPRO covers the goods-producing economy (manufacturing, mining, utilities) which now represents roughly 20% of US GDP. Periods where INDPRO contracts while services-driven GDP expands have become routine since 2015; reading an INDPRO drawdown as a recession signal in isolation now generates more false positives than in the post-war era.
2. Conflating INDPRO with the ISM Manufacturing PMI. The PMI is a diffusion index based on survey responses; INDPRO is a measured output index. The PMI typically leads INDPRO by one to three months at turning points but can disagree directionally for several quarters. Treating them as substitutes loses the lead/lag signal.
3. Misreading rebased levels. The 2017=100 base means a current INDPRO value near 103 is not directly comparable to a 2007 value near 100 under the earlier 2007=100 base. The level relationship between vintages is preserved by linking, but mechanical level comparisons across rebases can mislead. YoY changes are insensitive to the base year.
4. Ignoring Capacity Utilization (TCU). INDPRO measures output level; TCU measures output relative to estimated capacity. The same INDPRO print can signal different cyclical positions depending on whether capacity has expanded (low TCU, slack) or contracted (high TCU, tight). Users reading INDPRO without TCU miss half the cycle picture.

Historical Regimes

1919–1929 — Roaring twenties expansion. INDPRO grew at an average pace near 4% per year, propelled by electrification, mass-production techniques and consumer durables (automobiles, appliances). The series peaked in mid-1929 before the equity-market and credit-cycle reversal that began the next regime.

1929–1939 — Great Depression and slow recovery. INDPRO collapsed by roughly 50% from the 1929 peak to the 1932 trough — the deepest contraction in the series history. Recovery was non-linear: a partial rebound through 1937 was interrupted by the 1937–38 recession before the war mobilization regime began.

1939–1944 — War mobilization surge. INDPRO more than doubled between 1939 and 1944 as production shifted to munitions, aircraft, and shipping. The series reached its highest growth rates on record. Post-war demobilization in 1945–46 produced a sharp but brief contraction.

1945–1973 — Post-war industrial era. INDPRO grew at an average pace near 4% per year through 1973, with short and shallow recessions in 1948–49, 1953–54, 1957–58, 1960–61 and 1969–70. US manufacturing dominated global output; the index served as a near-coincident proxy for GDP.

1973–1982 — Oil shocks and stagflation. Two oil shocks (1973, 1979) and the Volcker disinflation produced the deepest peacetime contractions of the post-war era. INDPRO fell roughly 10% from peak to trough in both 1973–75 and 1980–82. The US GDP growth rate mirrored the same cycle with lower amplitude.

1983–2007 — Globalization and offshoring. INDPRO continued to grow but its share of GDP declined sharply as services expanded and manufacturing offshored. Two recessions (1990–91, 2001) produced shallow INDPRO declines. The 2001 contraction was unusually mild by GDP standards but more visible in INDPRO due to the dot-com tech investment unwind.

2008–2009 — Great Recession. INDPRO contracted by roughly 17% from peak to trough, the deepest decline since 1945. Recovery to the prior peak took about five years, the slowest of the post-war record. The US federal debt to GDP ratio reached its post-war high during this period.

2010–2019 — Range-bound expansion. INDPRO grew modestly through 2014, then went largely sideways from 2015 to 2019 as the manufacturing share of the economy continued to shrink. A mid-cycle contraction in 2015–16 (commodity bust) did not produce a recession by GDP standards — illustrating the structural decoupling of INDPRO from headline GDP.

2020–2026 — Pandemic shock and supply-chain era. INDPRO dropped roughly 17% in March–April 2020 (matching the 2008–09 trough in magnitude but compressed into two months) and rebounded toward the prior peak by late 2021. The post-2022 trajectory has stayed near the 2018 peak, with supply-chain reshoring and CHIPS-Act investment supporting capacity expansion while higher rates restrained interest-sensitive sectors.

Related Macroeconomic Datasets

INDPRO is a coincident-to-slightly-lagging indicator of the goods cycle. Cross-referencing with leading surveys (PMI), output integrators (GDP) and confidence series isolates the timing and breadth of cyclical inflection points.

Macroeconomic Dataset Hub

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

Explore the Eco3min Dataset Hub

Sources

• Federal Reserve Board — Industrial Production and Capacity Utilization (G.17 release), FRED series INDPRO

Dataset Reference

Eco3min Research. (2026). US Industrial Production Index — INDPRO (1919–2026).<br> https://eco3min.fr/en/us-industrial-production-dataset/