---
ticker: "USUNEMP"
title: "U.S. Unemployment Rate"
unit: "%"
frequency: "Monthly"
source: "Hussmanns, Mehran, and Verma (1990)"
release: "Updated Monthly"
category: "Labor Market"
country: "US"
language: "en"
canonical: "https://kred.dev/en/series/USUNEMP"
license: "https://creativecommons.org/licenses/by-nc-nd/4.0/"
latest_value: 4.10
latest_date: "2026-08-01"
first_date: "1948-01-01"
observations_total: 943
observations_shown: 120
---

# U.S. Unemployment Rate

## Overview

The share of the US labor force without work yet actively seeking it, the headline gauge of labor slack and of where the cycle stands.

## Key Figures

|  | Value | Date |
|---|---|---|
| Latest | 4.10 | 2026-08-01 |
| Change from previous | 0.00 | 2026-07-01 |
| Change over one year | -0.20 | 2025-08-01 |
| Highest on record | 14.80 | 2020-04-01 |
| Lowest on record | 2.50 | 1953-05-01 |
| Period covered | 1948-01-01 – 2026-08-01 |  |
| Observations | 943 |  |

## Recent observations

| Date | Value | Change |
|---|---|---|
| 2016-08-01 | 4.90 | +0.10 |
| 2016-09-01 | 5.00 | +0.10 |
| 2016-10-01 | 4.90 | -0.10 |
| 2016-11-01 | 4.70 | -0.20 |
| 2016-12-01 | 4.70 | 0.00 |
| 2017-01-01 | 4.70 | 0.00 |
| 2017-02-01 | 4.60 | -0.10 |
| 2017-03-01 | 4.40 | -0.20 |
| 2017-04-01 | 4.40 | 0.00 |
| 2017-05-01 | 4.40 | 0.00 |
| 2017-06-01 | 4.30 | -0.10 |
| 2017-07-01 | 4.30 | 0.00 |
| 2017-08-01 | 4.40 | +0.10 |
| 2017-09-01 | 4.30 | -0.10 |
| 2017-10-01 | 4.20 | -0.10 |
| 2017-11-01 | 4.20 | 0.00 |
| 2017-12-01 | 4.10 | -0.10 |
| 2018-01-01 | 4.00 | -0.10 |
| 2018-02-01 | 4.10 | +0.10 |
| 2018-03-01 | 4.00 | -0.10 |
| 2018-04-01 | 4.00 | 0.00 |
| 2018-05-01 | 3.80 | -0.20 |
| 2018-06-01 | 4.00 | +0.20 |
| 2018-07-01 | 3.80 | -0.20 |
| 2018-08-01 | 3.80 | 0.00 |
| 2018-09-01 | 3.70 | -0.10 |
| 2018-10-01 | 3.80 | +0.10 |
| 2018-11-01 | 3.80 | 0.00 |
| 2018-12-01 | 3.90 | +0.10 |
| 2019-01-01 | 4.00 | +0.10 |
| 2019-02-01 | 3.80 | -0.20 |
| 2019-03-01 | 3.80 | 0.00 |
| 2019-04-01 | 3.70 | -0.10 |
| 2019-05-01 | 3.60 | -0.10 |
| 2019-06-01 | 3.60 | 0.00 |
| 2019-07-01 | 3.70 | +0.10 |
| 2019-08-01 | 3.60 | -0.10 |
| 2019-09-01 | 3.50 | -0.10 |
| 2019-10-01 | 3.60 | +0.10 |
| 2019-11-01 | 3.60 | 0.00 |
| 2019-12-01 | 3.60 | 0.00 |
| 2020-01-01 | 3.60 | 0.00 |
| 2020-02-01 | 3.50 | -0.10 |
| 2020-03-01 | 4.40 | +0.90 |
| 2020-04-01 | 14.80 | +10.40 |
| 2020-05-01 | 13.20 | -1.60 |
| 2020-06-01 | 11.00 | -2.20 |
| 2020-07-01 | 10.20 | -0.80 |
| 2020-08-01 | 8.40 | -1.80 |
| 2020-09-01 | 7.80 | -0.60 |
| 2020-10-01 | 6.90 | -0.90 |
| 2020-11-01 | 6.70 | -0.20 |
| 2020-12-01 | 6.70 | 0.00 |
| 2021-01-01 | 6.40 | -0.30 |
| 2021-02-01 | 6.20 | -0.20 |
| 2021-03-01 | 6.10 | -0.10 |
| 2021-04-01 | 6.10 | 0.00 |
| 2021-05-01 | 5.80 | -0.30 |
| 2021-06-01 | 5.90 | +0.10 |
| 2021-07-01 | 5.40 | -0.50 |
| 2021-08-01 | 5.10 | -0.30 |
| 2021-09-01 | 4.70 | -0.40 |
| 2021-10-01 | 4.50 | -0.20 |
| 2021-11-01 | 4.10 | -0.40 |
| 2021-12-01 | 3.90 | -0.20 |
| 2022-01-01 | 4.00 | +0.10 |
| 2022-02-01 | 3.90 | -0.10 |
| 2022-03-01 | 3.70 | -0.20 |
| 2022-04-01 | 3.70 | 0.00 |
| 2022-05-01 | 3.60 | -0.10 |
| 2022-06-01 | 3.60 | 0.00 |
| 2022-07-01 | 3.50 | -0.10 |
| 2022-08-01 | 3.60 | +0.10 |
| 2022-09-01 | 3.50 | -0.10 |
| 2022-10-01 | 3.60 | +0.10 |
| 2022-11-01 | 3.60 | 0.00 |
| 2022-12-01 | 3.50 | -0.10 |
| 2023-01-01 | 3.50 | 0.00 |
| 2023-02-01 | 3.60 | +0.10 |
| 2023-03-01 | 3.50 | -0.10 |
| 2023-04-01 | 3.40 | -0.10 |
| 2023-05-01 | 3.60 | +0.20 |
| 2023-06-01 | 3.60 | 0.00 |
| 2023-07-01 | 3.50 | -0.10 |
| 2023-08-01 | 3.70 | +0.20 |
| 2023-09-01 | 3.70 | 0.00 |
| 2023-10-01 | 3.90 | +0.20 |
| 2023-11-01 | 3.70 | -0.20 |
| 2023-12-01 | 3.80 | +0.10 |
| 2024-01-01 | 3.70 | -0.10 |
| 2024-02-01 | 3.90 | +0.20 |
| 2024-03-01 | 3.90 | 0.00 |
| 2024-04-01 | 3.90 | 0.00 |
| 2024-05-01 | 3.90 | 0.00 |
| 2024-06-01 | 4.10 | +0.20 |
| 2024-07-01 | 4.20 | +0.10 |
| 2024-08-01 | 4.20 | 0.00 |
| 2024-09-01 | 4.10 | -0.10 |
| 2024-10-01 | 4.10 | 0.00 |
| 2024-11-01 | 4.20 | +0.10 |
| 2024-12-01 | 4.10 | -0.10 |
| 2025-01-01 | 4.00 | -0.10 |
| 2025-02-01 | 4.20 | +0.20 |
| 2025-03-01 | 4.20 | 0.00 |
| 2025-04-01 | 4.20 | 0.00 |
| 2025-05-01 | 4.30 | +0.10 |
| 2025-06-01 | 4.10 | -0.20 |
| 2025-07-01 | 4.30 | +0.20 |
| 2025-08-01 | 4.30 | 0.00 |
| 2025-09-01 | 4.40 | +0.10 |
| 2025-11-01 | 4.50 | +0.10 |
| 2025-12-01 | 4.40 | -0.10 |
| 2026-01-01 | 4.30 | -0.10 |
| 2026-02-01 | 4.40 | +0.10 |
| 2026-03-01 | 4.30 | -0.10 |
| 2026-04-01 | 4.30 | 0.00 |
| 2026-05-01 | 4.30 | 0.00 |
| 2026-06-01 | 4.20 | -0.10 |
| 2026-07-01 | 4.10 | -0.10 |
| 2026-08-01 | 4.10 | 0.00 |

## Definition

USUNEMP is the seasonally adjusted US unemployment rate recorded as published, the share of the economically active population that is without work yet available for and actively seeking work during the survey reference period, carried in percent without transformation. The rate is derived from a standard labor-force framework (Hussmanns, Mehran, and Verma 1990), which classifies each person of working age as employed, unemployed, or outside the labor force according to activity status, availability, and job-search criteria, and then divides the unemployed by the labor force.

Because it is a ratio rather than a headcount, numerator and denominator move together, so the rate can fall when discouraged searchers leave the labor force even with no gain in employment. A single monthly reading is therefore the net outcome of continuous flows of entry, exit, and unemployment-spell duration (Clark and Summers 1979).

Within the cyclical-indicator tradition the unemployment rate is one of the comoving series whose turning points cluster around aggregate reference cycles (Burns and Mitchell 1946). It is classified as lagging the cycle (Mitchell and Burns 1938), its turning points are identified through a reproducible algorithm (Bry and Boschan 1971), and it is placed within the diffusion- and composite-index apparatus (Moore 1961).

Labor aggregates of this kind are recast as observable shadows of a latent state of the economy (Stock and Watson 1989; Stock and Watson 1991), their measurement and timing properties are catalogued systematically (Zarnowitz 1992), and the value is treated as a monthly current-conditions signal (Shiskin 1961). Such atheoretical measurement carries meaning only against an explicit model of the labor market (Koopmans 1947).

## Methodology

KRED applies no transformation to USUNEMP and carries it exactly as constructed, neither rescaling, deflating, smoothing, nor annualizing it. Because the underlying series is already seasonally adjusted at source, the raw passthrough and the prior seasonal adjustment hold together without contradiction for the same monthly figure.

The measurement basis is a probability household survey in which each respondent's answers are translated into a discrete activity status for a fixed reference week and then tabulated under the operational definitions of the active population and of unemployment (Hussmanns, Mehran, and Verma 1990). The rate divides the unemployed by the sum of the employed and the unemployed, so numerator and denominator come from the same survey, and the flow accounting that links those statuses across months decomposes the change in the rate into entry, exit, and duration (Clark and Summers 1979).

Translating survey replies into a quantitative indicator belongs to the lineage of qualitative tendency surveys. The link between categorical survey replies and official output statistics was first modeled (Anderson 1952), the quantification of up-same-down responses was founded (Theil 1952), a probability method with indifference thresholds was added (Carlson and Parkin 1975), and competing balance and regression schemes were reviewed (Nardo 2003).

The cyclical reading of the resulting series follows a reproducible turning-point algorithm that dates its specific cycles (Bry and Boschan 1971) and a codified embedding in diffusion and composite indexes (Moore 1961). The single-index dynamic-factor representation is formalized (Stock and Watson 1991), as is the monthly real-time construction (Shiskin 1961). This measurement procedure nonetheless presupposes an economic model for its interpretation (Koopmans 1947).

## Applications in Economics

As a measure of labor-market slack, the unemployment rate anchors the reading of the business cycle. Its clustered turning points define the cycle as a comoving series (Burns and Mitchell 1946), and the rate is classified as a lagging indicator that confirms rather than anticipates a turn (Mitchell and Burns 1938).

A rising rate signals a deepening of cyclical weakness rather than an idiosyncratic shock, because labor aggregates are treated as noisy readings of the latent state of the economy (Stock and Watson 1989; Stock and Watson 1991).

The rate plays a standard role in composite cyclical assessment. Its place within diffusion- and composite-indicator systems is documented (Moore 1961; Zarnowitz 1992), turning-point dating separates cyclical from incidental movements (Bry and Boschan 1971), and its use as a current-conditions gauge is formalized as a monthly update (Shiskin 1961). A widely used recession gauge is derived from this same rate by taking the gap between its three-month moving average and the lowest such average of the preceding twelve months, and that gauge too is a descriptive transformation of the same series rather than a forecast.

An identical headline rate can carry different implications when the underlying flow mix differs. A stable surface value can mask shifting entry, exit, and duration flows that carry distinct welfare meaning (Clark and Summers 1979), and the population over which slack is measured is itself bounded by the operational definitions (Hussmanns, Mehran, and Verma 1990).

Labor-market conditions govern household income expectations, so the rate links to spending behavior (Katona 1951), and labor and confidence readings carry forward-looking content for consumption (Carroll, Fuhrer, and Wilcox 1994; Ludvigson 2004).

The indicator, however, informs but does not identify the mechanism (Koopmans 1947).

## Applications in Financial Markets

For markets the US unemployment rate is a monthly read on cyclical timing, and its lagging-indicator classification governs how a given print maps into expectations for the policy path (Mitchell and Burns 1938; Burns and Mitchell 1946).

Fixed-income and rate markets track it as a component of a latent-state estimate (Stock and Watson 1989; Stock and Watson 1991), and how much weight a surprise deserves is disciplined by the series' forecasting record (Zarnowitz 1992).

An unexpected change moves the expected trajectory of growth and inflation, because labor and confidence readings carry independent predictive content for household demand (Carroll, Fuhrer, and Wilcox 1994; Ludvigson 2004). Since expectations for the US short-rate path anchor global rates and exchange rates, the print transmits indirectly into the discount rates applied to Korean assets.

Each release is positioned within a composite and real-time signaling frame (Moore 1961; Shiskin 1961), phase transitions are flagged by turning-point dating (Bry and Boschan 1971), and a stable surface value must not be over-read against the flow decomposition (Clark and Summers 1979).

The labor signal connects to consumer attitudes that feed risk appetite (Katona 1951), and the survey definitions fix what is being counted (Hussmanns, Mehran, and Verma 1990). The same number nonetheless supports divergent trades depending on the model the trader imposes (Koopmans 1947).

## Statistical Tests

USUNEMP is a seasonally-adjusted-at-source unemployment rate, and its published level is the statistical object whose integration order the battery assesses. The sample is 940 monthly observations spanning 1948-01 to 2026-05, and the tests are fitted with a constant and trend.

The augmented Dickey-Fuller regression of Dickey and Fuller (1979), in the lag-augmented form of Said and Dickey (1984), rejects the unit-root null with a statistic of −3.9151 at p = 0.0116, and the nonparametric test of Phillips and Perron (1988) also rejects with a statistic of −4.0113 at p = 0.0085. The stationarity-null KPSS test of Kwiatkowski et al. (1992), however, likewise rejects trend stationarity with a statistic of 0.3065 at p < 0.01, so both opposing nulls are rejected, the three procedures do not reach honest agreement, and no clean order of integration is assigned. The GLS-detrended escalation of Elliott, Rothenberg, and Stock (1996), run only as confirmation of the augmented Dickey-Fuller leg, rejects with a statistic of −3.3399 at p = 0.0129, which corroborates that leg without resolving the ambiguous verdict, and the scope for unit-root tests to over-reject under near-unit moving-average errors is documented (Ng and Perron 2001).

Because the level is not a clean stationary object, the portmanteau and the break search are run on the first difference, since a level portmanteau reads mechanical near-unit autocorrelation and a level mean-break search spuriously segments the series (Perron 1989; Hamilton 2018; Bai and Perron 1998). The portmanteau statistic of Ljung and Box (1978), refining the form of Box and Pierce (1970), returns Q = 8.2465 at p = 0.7656 for lag 12 and Q = 12.7446 at p = 0.9701 for lag 24 on the differenced series, failing to reject the white-noise null, and the multiple-break procedure computed by the dynamic-programming algorithm of Bai and Perron (2003) finds no break in the mean of the differenced series either.

The seasonal battery is run with the interpretation that any absence of seasonality reflects the source adjustment rather than an inherent property. The seasonal-unit-root test of Hylleberg et al. (1990), with the monthly mechanics of Beaulieu and Miron (1993), rejects unit roots at the seasonal frequencies of the level with a statistic of 911.1556 at p = 0.000, the seasonal dummies fitted on the first difference are not jointly significant with F = 0.8728 at p = 0.5669, and the QS statistic on that first difference at seasonal lags 12 and 24 is 12.745 at p = 0.9701, detecting no residual seasonality. That reflects the seasonal adjustment already applied at source rather than an inherent absence of a seasonal pattern, consistent with the stationary-seasonality reading of Canova and Hansen (1995) and the seasonal-adjustment literature (Ghysels and Osborn 2001).

## Frequently Asked Questions

### What do the US unemployment rate and nonfarm payroll series report?

The United States unemployment rate and nonfarm payroll employment, carried as published without transformation.

### Can the US unemployment rate and payroll employment disagree?

They do. The unemployment rate comes from the household survey and payrolls from the establishment survey, with different samples and different definitions of employment. Divergence between them is common and is itself something to interpret rather than an error in either.

### Is the US unemployment rate seasonally adjusted?

Both are adjusted at source, and KRED applies no further adjustment.