UNRATE: US Civilian Unemployment Rate Monthly Data from FRED (1948–2026)

The UNRATE series, published monthly by FRED from the US Bureau of Labor Statistics, tracks the civilian unemployment rate (U-3 headline measure) in the United States since January 1948 — over 930 monthly observations. Released on the first Friday of each month alongside nonfarm payrolls as part of the BLS Employment Situation report, UNRATE represents one half of the Federal Reserve’s dual mandate and is the single most cited labor-market metric in US monetary-policy debates.

Dataset: US Unemployment Rate (1948–2026) · Updated 2026-06-01

Source: FRED series UNRATE · Federal Reserve Bank of St. Louis

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 UNRATE:

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

Direct CSV Access — Eco3min Structured Dataset

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

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

Using the Dataset in R

library(readr)

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

head(df)
summary(df$unrate)

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

Methodology

The UNRATE series is the U-3 measure of labor underutilization, derived from the BLS Current Population Survey (CPS) — a monthly survey of approximately 60,000 households conducted jointly with the US Census Bureau. The CPS reference week is the calendar week containing the 12th of each month. A person is classified as unemployed if they were not employed during the reference week, were available for work, and had actively searched for work at some point in the prior four weeks. Those out of the labor force entirely (discouraged workers, retirees, full-time students not seeking work) are excluded from both the numerator and denominator.

BLS publishes six alternative measures (U-1 through U-6) of labor underutilization; UNRATE refers exclusively to U-3, the official headline rate. The series is seasonally adjusted using X-13ARIMA-SEATS, with seasonal factors updated annually each February.

This dataset is updated monthly via automated pull from the FRED API.

Data Quality & Provider Notes

The UNRATE series is methodologically stable across most of its 78-year history, with one significant exception during the 2020 pandemic that the BLS itself flagged in real time.

• Release latency. BLS publishes the Employment Situation report on the first Friday of the following month at 8:30am ET — a 3–4 week lag. FRED mirrors the release within minutes. The Eco3min pipeline pulls the FRED feed on a monthly cadence.
• Revisions policy. UNRATE itself is generally not revised in subsequent monthly releases; only the seasonally adjusted figures may shift slightly during the annual benchmark revision in February, when updated seasonal factors are applied to the prior five years of data. This is structurally different from the wide-revision PAYEMS series.
• Alternative sources. The same data is available directly from the BLS Employment Situation release. International comparisons with Eurostat or OECD harmonised unemployment rates use somewhat different definitions; cross-country charts should rely on the OECD harmonised series rather than mixing national headlines.
• Known gaps. During the April and May 2020 releases, BLS acknowledged a classification error: workers absent from work for “other reasons” who should have been classified as unemployed on temporary layoff were not. Correctly classified, the April 2020 UNRATE would have been approximately 5 percentage points higher than the published 14.7%. BLS chose not to revise but noted the error transparently.

For analytical work, the convention is to read UNRATE in conjunction with the labor-force participation rate (FRED series CIVPART) and the broader U-6 measure (U6RATE), since each can independently distort the cyclical signal.

Common Pitfalls When Using UNRATE

The UNRATE series is the most-cited US labor metric and also one of the most-misread.

1. Treating U-3 (UNRATE) as the only measure of labor slack. The BLS publishes six measures of underutilization. U-6 — which includes part-time workers seeking full-time work and “marginally attached” workers no longer actively searching — typically runs 3–5 percentage points above U-3 and captures slack that UNRATE misses. The gap between U-3 and U-6 is itself a cyclical indicator.
2. Ignoring labor-force participation effects. UNRATE is a ratio. The denominator — the labor force — can shrink when discouraged workers exit job-seeking, mechanically lowering UNRATE without any improvement in employment. The US labor-force participation rate fell from 66% in 2008 to 62.5% in 2015, contributing to the headline UNRATE decline of that period independently of actual job creation.
3. Confusing UNRATE level with the Sahm Rule trigger. The Sahm Rule is a rate-of-change indicator: the 3-month moving average of UNRATE rising 0.5 percentage point above its trailing 12-month low. A UNRATE level of 4.0% can be either benign (stable near multi-decade low) or recessionary (rising from 3.4%) — the level alone does not separate the two. The Sahm reformulation exists precisely to address this.
4. Comparing UNRATE across decades without demographic adjustment. The age and gender composition of the US labor force has shifted substantially since 1948. Adjusting UNRATE for demographic composition (the BLS “experimental” series and academic prime-age unemployment rates do this) yields a noticeably different historical comparison than the headline series — a 5% UNRATE in 1980 with a different age mix is not directly comparable to 5% in 2026.

Historical Regimes

1948–1969 — Postwar prosperity and Korean War cycles. UNRATE cycled between 2.5% (May–June 1953, the all-time low) and 7.5% (1958 recession). The mid-1960s saw sustained readings below 4% during the Vietnam-era fiscal expansion.

1970–1982 — Stagflation era. Three recessions pushed UNRATE successively higher: 6% in 1971, 9% in 1975, and 10.8% in November and December 1982 — the post-WWII high until the 2020 pandemic. The 10.8% peak coincided with the Volcker disinflation and Fed funds rate readings above 15%.

1983–1989 — Reagan recovery. UNRATE fell from 10.4% in early 1983 to 5.0% by March 1989, a 5.4-percentage-point decline over six years. The pace of decline averaged roughly 1 percentage point per year — the steepest sustained drop in the postwar era.

1991–2000 — Long expansion. Following the brief 1990–1991 recession (UNRATE peak 7.8% in June 1992), the series fell steadily to 3.8% in April 2000 — the lowest reading in 30 years and one consistent with sustained 4–5% real wage growth during the late-1990s productivity boom.

2008–2010 — Global Financial Crisis. UNRATE peaked at 10.0% in October 2009, the highest reading since 1983. The recovery was protracted: the series did not return below 5% until late 2015, six years after the GFC trough, and labor-force participation fell sharply during the period — masking part of the underlying labor weakness. The level-versus-cycle distinction this illustrates is the subject of how UNRATE behaves as a business-cycle signal.

2014–2019 — Slow recovery to multi-decade lows. UNRATE fell from 6.7% in early 2014 to 3.5% in September 2019 — the lowest reading since December 1969. The decline was nearly symmetrical with the Reagan-era recovery in pace.

2020 — Pandemic shock. UNRATE jumped from 3.5% in February 2020 to 14.7% in April 2020 — the highest published reading since the series began in 1948, and the largest two-month change on record. Recovery was rapid: UNRATE returned below 5% by mid-2021 and below 4% by early 2022, a pace unprecedented in modern US labor-market history.

2023–2026 — Post-pandemic tightness and the ambiguous Sahm trigger. UNRATE reached 3.4% in January and April 2023 — the lowest reading since May 1969. A subsequent gradual rise toward 4% triggered the Sahm Rule in mid-2024, but without the labor-flow signals (rising layoffs in ICSA, contracting payrolls) that historically accompanied past Sahm triggers — the most ambiguous case on record, addressed in our study on NFP revisions and recession bias.

Related Macroeconomic Datasets

UNRATE is the household-survey-based stock measure of US labor underutilization. The datasets below cover the complementary establishment-survey, labor-flow and activity measures that the analytical reading of UNRATE depends on.

Macroeconomic Dataset Hub

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

Explore the Eco3min Dataset Hub

Sources

• Federal Reserve Bank of St. Louis — FRED database
• US Bureau of Labor Statistics — Current Population Survey (CPS), Employment Situation report

Dataset Reference

Eco3min Research. (2026). US Civilian Unemployment Rate — UNRATE (1948–2026).<br> https://eco3min.fr/en/us-unemployment-rate-dataset/