The unemployment rate: U-3, U-6 and labor force participation

The unemployment rate is the share of the labor force that has no job, is available to work, and has looked for work recently. It is the most quoted measure of how much slack there is in the labor market, and it appears in the Federal Reserve's statements, in election debates and in every monthly The jobs report: how to read nonfarm payrolls and the two surveys. In September 2026 it was 4.2%.

The headline rate, known as U-3, is one of six measures the Bureau of Labor Statistics publishes. The broadest, U-6, adds people who want a job but have stopped looking and people working part time who want full-time work; it was 7.6% in the same month. Reading the two together, and alongside the labor force participation rate, tells you much more than the headline alone.

This guide explains how the rate is measured, what each of the six measures counts, how participation changes the picture, and the traps in reading a single month's number. For a one-line definition, see Unemployment rate.

How the unemployment rate is measured

Each month the Census Bureau, on behalf of BLS, interviews about 60,000 households for the Current Population Survey. Every member aged 16 or older is classified by what they did in the calendar week that includes the 12th of the month:

• Employed: did any paid work, worked in their own business or farm, did at least 15 hours of unpaid work in a family business, or had a job but was temporarily absent.
• Unemployed: had no job, was available for work, and made a specific effort to find work in the four weeks ending with that week. People on temporary layoff who expect to be recalled count as unemployed even if they are not searching.
• Not in the labor force: everyone else, including retirees, students, people caring for family and people who have given up looking.

The labor force is the employed plus the unemployed. Receiving unemployment insurance has nothing to do with the classification: plenty of unemployed people get no benefits, and the count of benefit claims is a separate series, covered in Jobless claims: reading the weekly unemployment insurance report.

U-1 to U-6: the six measures

BLS publishes six measures of labor underutilization in Table A-15 of the Employment Situation. They are built so each one is broader than the last, except U-1 and U-2, which are narrower slices of U-3.

Three definitions make the table readable:

• Marginally attached people are not working and have not looked in the past four weeks, but want a job, are available for one, and have looked at some point in the past 12 months.
• Discouraged workers are the marginally attached who give a job-market reason for not looking, such as believing no work is available.
• Part time for economic reasons means employed people who want full-time work but are working part time because their hours were cut or they could only find a part-time job. They count as employed in U-3 but as underused in U-6.

The gap between U-6 and U-3, 3.4 points in September 2026, is mostly that last group plus the marginally attached. It widens in downturns, when employers cut hours before they cut jobs and more people give up searching.

U-3 and U-6 over three decades

The chart shows the two measures moving together but with very different amplitudes.

• Peaks. After the 2007–2009 recession, U-3 reached 10.0% in October 2009 while U-6 reached 17.2% in December 2009 and again in April 2010. In April 2020, U-3 hit 14.8% and U-6 22.9%, the highest readings since U-6 starts in 1994. The gap between them was 8.1 points that month, the widest in the series.
• Troughs. U-3 fell to 3.4% in April 2023, and U-6 bottomed at 6.6% in December 2022 and again in April 2023, the lowest in the series. The narrowest gap, 2.9 points, came in October 2000.
• Recent months. U-3 rose from 3.4% in April 2023 to 4.4% in September 2025 and 4.5% in November 2025, then eased to 4.2% by September 2026. U-6 went from 8.1% in September 2025 to 7.6% in September 2026.

The bottom pane is the participation rate. It peaked at 67.3% in January 2000, was 66.4% in January 2007, fell to 62.4% by September 2015, and was 63.3% in February 2020 before dropping to 60.1% in April 2020. It stood at 61.8% in September 2026. Because retirees are counted outside the labor force, an older population pulls participation down even when jobs are plentiful, so economists often watch the participation or employment rate of people aged 25 to 54 as a cleaner gauge.

How to read the rate

Look at why it moved

A rise in the unemployment rate can come from job losses, from new entrants who are still searching, or from people returning to the labor force after time away. A fall can come from hiring or from people dropping out. The monthly release breaks unemployment down by reason (job losers, job leavers, re-entrants and new entrants) and reports participation, so the cause is usually visible in the same report.

Watch the trend, not the tenth

The household survey is small relative to the payroll survey, so the rate wobbles. BLS estimates that, with the rate near 6%, a monthly change needs to be about 0.3 percentage point to be statistically significant at 90% confidence. A move from 4.1% to 4.2%, as in September 2026, is well within that band. Rules that average several months, such as the The Sahm rule: a recession signal from the unemployment rate, exist precisely to filter that noise.

Why markets care

The unemployment rate is one half of the Fed's dual mandate, so a sustained rise tends to pull expected interest rates down, and a sustained fall tends to push them up. Because unemployment usually rises fast at the start of recessions and falls slowly in recoveries, a turn upward from a low is one of the more reliable, if late, signs that the business cycle has changed. The The business cycle: expansions, peaks, recessions and troughs guide puts that pattern in context.

Pitfalls and limitations

• Discouraged workers vanish from U-3. Anyone who stops searching leaves the labor force and stops counting as unemployed. U-4 through U-6 exist to catch them.
• Part-time work hides slack. Someone working 10 hours who wants 40 counts as fully employed in U-3.
• Classification errors in unusual times. In April 2020, BLS reported that many people absent from work for "other reasons" should have been recorded as unemployed on temporary layoff. Had they been, the unadjusted unemployment rate would have been almost 5 points higher than the 14.7% first published.
• Missing months. BLS did not collect household survey data for October 2025 because federal appropriations lapsed from October 1 through November 12, 2025, so there is no official October 2025 rate. Year-over-year comparisons and moving averages across that month need care.
• Revisions are rare but real. The seasonally adjusted rate is revised once a year when new seasonal factors arrive, so a reading can change by a tenth: July 2024 was first published at 4.3% and stands at 4.2% in the current data.
• National rates hide dispersion. Unemployment differs widely by age, education, region and industry; the national average can stay steady while one group's rate rises.

How the rate compares with other labor measures

How Finabase does it

Finabase does not carry unemployment data, a jobs calendar or any macro data page. The chart on this page is ours, built from the public BLS series on FRED as downloaded on October 30, 2026.

What the app shows is the market's response. The Treasury page lists yields from 1 month to 30 years, where a change in the jobs picture shows up first in the 2- and 10-year yields. The Sector page shows how the 11 S&P 500 sectors trade on the day, which is where a shift in growth expectations often appears, and the economy topic in Discover gathers web coverage of each release.

Frequently asked questions

What is the difference between U-3 and U-6?

U-3 counts only people with no job who looked for one in the past four weeks. U-6 adds people who want work but have stopped looking (the marginally attached) and people working part time who want full-time hours. In September 2026, U-3 was 4.2% and U-6 was 7.6%.

Does the unemployment rate count people on unemployment benefits?

No. The rate comes from a household survey and depends only on whether a person is jobless, available and searching. Benefit receipt is irrelevant; weekly claims for benefits are a separate Labor Department series.

Why can the unemployment rate fall when the economy is weak?

Because people who stop searching leave the labor force and are no longer counted as unemployed. In the illustrative example above, 1 million people giving up cuts the rate from 4.2% to 3.6% with no new jobs. Checking participation and U-6 shows whether that is happening.

What was the lowest US unemployment rate?

In the BLS monthly series starting in 1948, the lowest reading is 2.5% in May 1953. The lowest since 1970 was 3.4% in April 2023 in the current data.

Is there an unemployment rate for October 2025?

No. The household survey was not conducted in October 2025 because of the lapse in federal appropriations from October 1 to November 12, 2025, and BLS did not collect the data afterward.

Summary

The unemployment rate is the share of the labor force that is jobless, available and searching, measured by a monthly survey of about 60,000 households. The headline U-3 is one of six BLS measures; U-6 adds discouraged workers, other marginally attached people and involuntary part-timers, and in September 2026 stood at 7.6% against U-3's 4.2%. Because people who stop looking drop out of the count, read the rate with the participation rate, give monthly moves of a tenth little weight, and remember the missing October 2025 observation in any long calculation.