The unemployment rate is built from a monthly survey, not from benefit records

The unemployment rate you see in the news does not come from counting people who filed for unemployment insurance. Instead, the U.S. Bureau of Labor Statistics conducts a monthly survey called the Current Population Survey, or CPS. They call about 60,000 households and ask whether people in those households are working, looking for work, or neither. The answers to those calls become the unemployment rate.

This matters because it means the official rate can go up even when fewer people are filing for benefits, and it can go down even when more people need help. The survey counts people who want work but have stopped looking, and it misses people who never filed for unemployment insurance in the first place. Understanding how the number is built helps you see what it actually measures and what it leaves out.

Key Takeaways

  • The unemployment rate comes from a monthly phone survey of 60,000 households, not from unemployment insurance claims or government benefit records.
  • To be counted as unemployed, a person must have no job, have looked for work in the past four weeks, and be ready to start work when ready.
  • People who stopped looking for work are not counted as unemployed, even if they want a job — this is called "discouraged workers" and it lowers the official rate.
  • The Bureau of Labor Statistics releases the unemployment rate on the first Friday of each month, covering the previous month's data.
  • Different unemployment rates exist (U-3 through U-6) that count different groups, and the official rate reported in news is always U-3.

Who gets counted as unemployed

The Current Population Survey asks three things to decide if someone is unemployed. First, did you have a job last week? If yes, you are employed and the survey moves on. If no, the interviewer asks: did you look for work in the past four weeks? This means actively searching — sending resumes, going to interviews, contacting employers, registering with a temp agency, or checking job boards. Passive activities like reading want ads do not count.

Third, if you did look for work, are you ready to start a job this week if one were offered? This filters out people who are looking but cannot actually begin work due to school, health, or other commitments. Only people who answer yes to all three questions become part of the unemployment count.

The survey also counts your labor force status — whether you are in the labor force at all. The labor force includes people with jobs and people actively looking. It does not include retirees, full-time students not seeking work, people with disabilities not seeking work, or anyone else not in the job market. This is why the unemployment rate can seem low even when many adults are not working — they are straightforward not counted in the denominator.

Why people stop being counted as unemployed

Once you stop looking for work, you leave the labor force and the unemployment rate no longer counts you. This happens for many reasons. You might be discouraged after months of rejection. You might have returned to school. You might have a health condition that makes work impossible right now. You might be caring for a family member. The survey does not distinguish between these situations — once you stop actively searching, you are out.

This creates what economists call the discouraged worker effect. During recessions, when jobs are scarce and competition is fierce, many people stop looking. The unemployment rate can actually fall even as the job market worsens, because people are leaving the labor force faster than they are finding work. This is one reason why the official unemployment rate sometimes feels disconnected from what people experience in their communities.

The Bureau of Labor Statistics tracks these people separately in other unemployment measures (U-4, U-5, and U-6), which include discouraged workers and people marginally attached to the labor force. But the number reported in news headlines is always U-3, the official rate, which does not include them.

How the survey becomes a national number

The Bureau of Labor Statistics contacts the same 60,000 households each month, though the sample rotates so that no household stays in the survey longer than four years. The households are selected to be representative of the entire U.S. population by region, age, race, and other characteristics. Interviewers conduct the survey by phone or in person, usually in the first two weeks of the month, asking about the previous week's work status.

Once the data is collected, the Bureau weights the responses to account for the fact that the sample is not a perfect mirror of the country. A household's answers are multiplied by a factor that represents how many similar households exist nationwide. This weighting process is where regional differences, population shifts, and demographic changes get factored in. The final number — the unemployment rate — is the total number of unemployed people divided by the total labor force, expressed as a percentage.

The Bureau releases this number on the first Friday of each month at 8:30 a.m. Eastern time. The data covers the previous month, so the January unemployment rate, released in early February, describes what happened in January. This one-month lag means the rate you read in the news is already several weeks old.

What the unemployment rate does not measure

The official rate misses several groups that affect the real job market. Underemployed workers — people working part-time who want full-time work — are counted as employed, even though they are not getting the hours or income they need. Self-employed people are counted as employed if they have any income, regardless of whether the business is sustainable. Gig workers and people in the informal economy are included only if they report the work to the survey.

The rate also does not capture people who have never entered the labor force. Someone who has never worked and is not looking is straightforward not part of the calculation. This means the unemployment rate can obscure long-term workforce withdrawal, particularly among men without college degrees and among people with disabilities.

Additionally, the survey relies on self-reporting. People may misremember whether they looked for work, or they may not report informal job searching. The survey also cannot capture people who are homeless or living in institutions, so it systematically undercounts unemployment among the most vulnerable populations.

The difference between U-3 and other unemployment measures

The Bureau of Labor Statistics publishes six different unemployment rates, labeled U-1 through U-6. Each one answers a slightly different question about the job market. U-3, the official rate reported in news, counts people with no job who looked for work in the past four weeks and are ready to start when ready. This is the rate you see in headlines.

U-4 adds discouraged workers — people who want work but stopped looking because they believe no jobs are available for them. U-5 adds all marginally attached workers, meaning people who want work and have looked in the past year but not in the past month. U-6, sometimes called the "underemployment rate," adds both marginally attached workers and people working part-time involuntarily. U-6 is typically two to three percentage points higher than U-3.

The Bureau publishes all six rates each month, but news outlets and government officials almost always report U-3. This is partly because it is the official definition agreed upon internationally, and partly because it is the lowest number and therefore the most optimistic. If you want a fuller picture of the job market, looking at U-5 or U-6 alongside U-3 gives you a sense of how many people are struggling to find adequate work.

How unemployment rates vary by state and demographic group

The Bureau of Labor Statistics publishes unemployment rates for each state, and for demographic groups including age, race, gender, and education level. State rates vary significantly because local economies are different. A state with a major industry downturn will have a higher rate than the national average. Demographic rates show persistent gaps — unemployment for Black workers is typically about twice the rate for white workers, and unemployment for teenagers is much higher than for adults.

These breakdowns are published monthly alongside the national rate, though state data takes longer to process and is released with a longer lag. If you are looking for unemployment information specific to your region or demographic group, the Bureau's website publishes tables with this detail. However, smaller demographic groups have smaller sample sizes in the survey, so their rates are less stable month to month and should be read as trends rather than precise numbers.

Frequently Asked Questions

Why does the unemployment rate sometimes go down when the economy is struggling?

The rate falls when people stop looking for work faster than they find jobs. During recessions, discouraged workers leave the labor force, which shrinks the denominator of the unemployment calculation. The rate can improve on paper even as the job market worsens, because fewer people are counted as unemployed — they are straightforward no longer in the labor force.

Is the unemployment rate the same as the number of people receiving unemployment benefits?

No. The unemployment rate comes from a survey and counts people actively looking for work. Unemployment insurance claims come from a separate system and count people who filed for benefits. Many unemployed people do not file for benefits, and some people receiving benefits are not counted as unemployed because they stopped looking for work.

How accurate is the unemployment rate?

The survey of 60,000 households has a margin of error, typically around 0.2 percentage points. This means the true rate could be slightly higher or lower than reported. The Bureau publishes confidence intervals with each release. The bigger limitation is not statistical error but conceptual — the rate does not count discouraged workers or underemployed people, so it may not reflect how many people are struggling in the job market.

When is the unemployment rate released, and where can I find it?

The Bureau of Labor Statistics releases the unemployment rate on the first Friday of each month at 8:30 a.m. Eastern time. The data covers the previous month. You can find the current and historical rates on the Bureau's website at bls.gov, or through news outlets that report the release.

Does the unemployment rate include self-employed people?

Self-employed people are counted as employed if they report any income from self-employment, even if the business is very small or unstable. They are not counted as unemployed unless they report having no self-employment income and are actively looking for work.