The unemployment rate is a percentage, not a count of jobless people

The unemployment rate you see in news reports—say, 3.8% or 4.2%—is not the share of all people without jobs. It is the share of people actively looking for work who cannot find it. That distinction matters because millions of people without jobs are not counted in the rate at all: retirees, students, people caring for family members, and people who have stopped searching.

The U.S. Bureau of Labor Statistics (BLS) calculates this rate using data from two separate surveys. The first, called the Current Population Survey (CPS), asks about 60,000 households each month whether household members are working, looking for work, or neither. The second, the Current Employment Statistics (CES), collects payroll records from about 400,000 businesses to count how many jobs exist. The unemployment rate comes from the CPS data alone.

Key Takeaways

  • The unemployment rate measures only people actively searching for work who cannot find it, not all people without jobs.
  • The Bureau of Labor Statistics surveys 60,000 households monthly through the Current Population Survey to determine who counts as unemployed.
  • To be counted as unemployed, you must be without a job, available to work, and have taken a specific action to find work in the past four weeks.
  • The labor force itself shrinks or grows depending on how many people are actively looking for work, which means the unemployment rate can fall even if no new jobs are created.
  • Alternative measures like U-6 include discouraged workers and part-time workers seeking full-time jobs, and they are always higher than the headline rate.

Who counts as unemployed in the official rate

The BLS uses three conditions to count someone as unemployed. First, you must not have a job. Second, you must be available to work. Third, you must have taken a specific action to find work within the past four weeks. Those actions include submitting a job process, contacting an employer directly, attending a job interview, registering with a public or private employment agency, checking job listings, or placing or answering job advertisements.

This third condition is the one that excludes millions of people. Someone who wants a job but has not looked in the past four weeks—because they are discouraged, because they lack transportation, or because they have other obligations—is not counted as unemployed. They are counted as "not in the labor force." This category includes retirees, full-time students, people with disabilities who are not seeking work, and people who have given up searching.

The labor force itself is therefore not fixed. When more people start looking for work, the labor force grows and the unemployment rate can rise even if employers are hiring. When people stop looking, the labor force shrinks and the unemployment rate can fall even if jobs are disappearing. This is why a falling unemployment rate does not always mean the job market is improving.

How the Current Population Survey collects the data

Each month, the Census Bureau conducts the CPS on behalf of the BLS. The survey reaches about 60,000 households, selected to represent the U.S. population. Interviewers ask whether each household member age 16 and older worked last week, looked for work, or did neither. They ask follow-up questions about job search methods, reasons for not working, and hours worked.

The survey is conducted in the week that includes the 12th of each month, and results are released on the first Friday of the following month. The data is weighted to match the known population of the United States, so the answers from 60,000 households are scaled up to represent the entire country.

One limitation of the CPS is that it is a sample, not a count of every person. This means the numbers carry a margin of error. The BLS publishes confidence intervals alongside the headline rate, though these are rarely reported in news coverage. A change of 0.1 or 0.2 percentage points in the unemployment rate can fall within the margin of error and may not represent a real shift in the labor market.

The math: labor force, employment, and unemployment

The unemployment rate is calculated using three numbers: the labor force, the number of employed people, and the number of unemployed people. The formula is straightforward:

Unemployment Rate = (Unemployed / Labor Force) × 100

The labor force is the sum of employed people plus unemployed people. If 160 million people are employed and 6 million are unemployed, the labor force is 166 million. The unemployment rate would be (6 / 166) × 100, or 3.6%.

The labor force participation rate—a separate measure—is the share of the total population age 16 and older that is either working or actively looking for work. If the U.S. population age 16 and older is 260 million, and the labor force is 166 million, the participation rate is 63.8%. This rate has fallen over the past two decades, mainly because the population is aging and more people are retiring.

Why the headline rate differs from alternative unemployment measures

The unemployment rate reported in the news—officially called U-3—counts only people without jobs who are actively searching. But the BLS publishes five other measures, labeled U-1 through U-6, that count different groups. U-6, the broadest measure, includes people who have looked for work in the past 12 months but not in the past four weeks (discouraged workers), plus people working part-time who want full-time jobs. U-6 is always higher than U-3, sometimes by 1 to 2 percentage points or more.

During recessions, the gap between U-3 and U-6 widens because more people become discouraged and stop searching. During strong job markets, the gap narrows. Neither measure is "wrong"—they answer different questions. U-3 tells you what share of active job seekers cannot find work. U-6 tells you what share of the broader population is either jobless or underemployed. Policy makers and economists often look at both.

How state and local unemployment rates are calculated

Each state publishes its own unemployment rate, calculated the same way as the national rate but using state-level data. State rates come from the CPS, supplemented by state unemployment insurance claims data. Local unemployment rates—by county or metropolitan area—are published with a one-month lag because the sample sizes are smaller and require more time to process.

State and local rates can differ significantly from the national rate. A state with a large manufacturing sector may have a higher unemployment rate during a recession than the national average. A state with rapid population growth and strong tech hiring may have a lower rate. These differences reflect real variation in local labor markets, not errors in the calculation.

What the unemployment rate does and does not tell you

The unemployment rate is useful for tracking broad trends in the labor market. A rising rate over several months usually signals economic weakness. A falling rate usually signals job growth. But the rate alone does not tell you how long people have been unemployed, whether jobs are full-time or part-time, whether wages are rising or falling, or whether the jobs being created pay enough to support a household.

The rate also does not capture underemployment—people working part-time who want full-time work, or people in jobs far below their skill level. It does not count people who have stopped looking because they believe no jobs are available. And it does not distinguish between someone unemployed for two weeks and someone unemployed for two years, though the BLS publishes separate data on unemployment duration.

For a fuller picture of the labor market, economists look at the unemployment rate alongside job creation numbers, wage growth, labor force participation, and the duration of unemployment. News reports often mention these other measures, but the unemployment rate remains the single most-watched indicator because it is straightforward, published monthly, and has been calculated the same way for decades.

Frequently Asked Questions

Why can the unemployment rate fall when jobs are being lost?

The unemployment rate can fall if the labor force shrinks faster than employment falls. This happens when people stop looking for work—they move out of the "unemployed" category and into "not in the labor force." During the early months of the COVID-19 pandemic, the unemployment rate fell partly because many people stopped searching, even though total employment was still declining.

How is someone counted as unemployed if they have not worked in months?

They are not counted as unemployed unless they have taken a job search action in the past four weeks. Someone without work for six months who has not looked in the past month is counted as "not in the labor force," not unemployed. This is why the unemployment rate can understate the true number of people struggling to find work.

Does the unemployment rate include self-employed people?

Self-employed people are counted as employed if they work at least one hour per week for pay. They are not counted as unemployed unless they lose that self-employment income and begin actively searching for other work. Gig workers and contract workers are treated the same way.

Why is there a delay between when data is collected and when it is released?

The CPS is conducted in the week containing the 12th of each month, but the data must be processed, weighted, and reviewed before release. The BLS publishes the results on the first Friday of the following month. This one-month lag means the unemployment rate you read in February describes the labor market in January.

Can the unemployment rate be zero?

No. Even in the strongest job markets, some unemployment always exists because people are always transitioning between jobs, entering the labor force for the first time, or relocating. This baseline level of unemployment is called the natural rate or non-accelerating inflation rate of unemployment (NAIRU). Most economists estimate it between 3.5% and 4.5%, though estimates vary.