The unemployment rate is calculated from a monthly survey of about 60,000 households, not from unemployment insurance claims or job postings

The U.S. Bureau of Labor Statistics (BLS) does not count people who file for unemployment benefits. Instead, it conducts a survey called the Current Population Survey (CPS) every month, asking households whether people are working, looking for work, or neither. The unemployment rate is the percentage of people actively looking for a job divided by the total number of people in the labor force — those working or actively searching.

This matters because it means the official unemployment rate can rise even when fewer people are filing for benefits, and it can fall even when job openings are shrinking. The survey captures a different population than the unemployment insurance system does, and the two numbers often move in different directions month to month.

Key Takeaways

  • The unemployment rate comes from a monthly household survey of 60,000 homes, not from unemployment insurance filings or job listings.
  • A person counts as unemployed only if they are without work, available to work, and actively looking — passive job searching does not count.
  • The labor force includes only people working or actively searching, so people who stop looking are no longer counted in either the numerator or denominator.
  • The BLS publishes the unemployment rate on the first Friday of each month for the previous month, along with separate rates for different age groups, races, and education levels.
  • The official rate (called U-3) differs from broader measures like U-6, which includes people working part-time who want full-time work and people who have looked for work recently but stopped.

Who counts as unemployed in the survey

The CPS asks households three things: Are you working? If not, have you looked for work in the past four weeks? If yes, are you available to start a job this week? Only people who answer no, yes, and yes are counted as unemployed.

"Looking for work" has a specific meaning. It includes sending out resumes, going to job interviews, contacting employers directly, placing or answering job advertisements, and registering with employment agencies. It does not include reading job postings, asking friends if they know of openings, or waiting to hear back from a previous process. The BLS counts only active, recent search behavior.

This is why someone who has been job hunting for months but took a week off to handle a family matter might not be counted as unemployed that month — they have to have searched in the past four weeks, and the survey happens in a specific week. Someone who is discouraged and has stopped looking altogether is not counted as unemployed; they fall out of the labor force entirely.

The labor force and why it shrinks or grows

The unemployment rate is not straightforward the number of jobless people divided by the total population. It is the number of unemployed people divided by the labor force — the sum of people who are working plus people who are actively looking for work.

When people stop looking for work, they leave the labor force. This can happen because they retire, return to school, take care of family members, become disabled, or lose hope of finding a job. During recessions, the labor force often shrinks because discouraged workers stop searching. During economic expansions, people re-enter the labor force because job prospects improve.

This creates a counterintuitive situation: the unemployment rate can fall even when fewer people are working, if enough people leave the labor force. Conversely, the rate can rise even when jobs are being added, if more people enter the labor force to search for those jobs. The BLS publishes the labor force participation rate separately so readers can see whether the unemployment rate is falling because jobs are being filled or because people have stopped looking.

How the survey is conducted and when results are released

The BLS contacts about 60,000 households each month in the Current Population Survey. The survey is conducted by the Census Bureau on behalf of the BLS. Households are selected to represent the U.S. population by region, city size, and other characteristics. The same households are interviewed for four consecutive months, then rotated out and replaced.

The survey happens in the week that includes the 12th of the month. Results are published on the first Friday of the following month. So the survey conducted in mid-January produces the unemployment rate released on the first Friday of February, which is called the January unemployment rate.

The BLS also publishes the survey results broken down by age, race, ethnicity, education level, industry, and length of unemployment. These breakdowns often tell a different story than the overall rate — for example, the national rate might be 4 percent while the rate for people without a high school diploma is 6 percent and the rate for college graduates is 2 percent.

The difference between U-3 and U-6 unemployment rates

The official unemployment rate reported in news headlines is called U-3. It counts people without work who are actively searching. But the BLS calculates five other unemployment rates, U-1 through U-6, that count different groups of people.

U-6 is the broadest measure. It includes U-3 (actively unemployed people) plus people working part-time who want full-time work, plus people who have looked for work in the past 12 months but are not currently looking. U-6 is always higher than U-3 because it includes more people. During recessions, the gap between U-3 and U-6 widens because more people move into part-time work or stop looking altogether.

U-1 through U-5 measure narrower groups: people unemployed for 15 weeks or more, people who left jobs versus people laid off, and other distinctions. The BLS publishes all six rates monthly so researchers and policymakers can see the full picture of labor market conditions. News reports focus on U-3 because it is the official rate, but U-6 often better captures how many people are struggling to find adequate work.

Why unemployment insurance claims and the unemployment rate are different numbers

Many people assume the unemployment rate comes from unemployment insurance filings, but the two are separate systems. Unemployment insurance is administered by state agencies. To receive benefits, you must have worked recently, lost your job through no fault of your own, and meet your state's earnings or hours requirements. The number of people filing for benefits depends on how many people lost jobs recently and how many have exhausted their benefits.

The unemployment rate, by contrast, counts anyone without work who is actively searching, regardless of whether they are receiving benefits. This includes people who quit their jobs, people whose benefits have run out, people entering the labor force for the first time, and people who do not meet their state's requirements for benefits. It excludes people receiving benefits who have stopped looking for work.

During the COVID-19 pandemic, unemployment insurance claims spiked to historic levels while the unemployment rate fell faster than expected, partly because many people received extended benefits and stopped looking for work, moving them out of the labor force. The two numbers can move in opposite directions because they measure different populations and different behaviors.

Seasonal adjustment and why the raw numbers are not what you hear

The BLS publishes two versions of the unemployment rate each month: the seasonally adjusted rate and the unadjusted rate. The unadjusted rate is what the survey actually found. The seasonally adjusted rate removes the effect of predictable seasonal patterns.

Some industries hire heavily at certain times of year. Retail hiring spikes before the winter holidays. Construction slows in winter. Schools hire in summer. These patterns happen every year, so the BLS uses historical data to estimate how much of the month-to-month change in unemployment is due to these seasonal swings versus actual changes in the job market. The seasonally adjusted rate removes that seasonal noise so readers can see the underlying trend.

News reports always cite the seasonally adjusted rate because it is easier to compare month to month. The unadjusted rate is published alongside it, but it is less useful for spotting economic trends because it includes the same seasonal swings every year. Both numbers are correct — they are just answering different questions.

Frequently Asked Questions

Does the unemployment rate include people who are retired or in school?

No. Retired people and full-time students are not counted as unemployed because they are not in the labor force — they are not working and not looking for work. If a student works part-time and is not looking for additional work, they are counted as employed, not unemployed. If they are looking for a job, they are counted as unemployed.

What if someone is working part-time but wants full-time work?

They are counted as employed in the U-3 rate, not unemployed. However, they are included in the U-6 rate, which captures underemployment. The BLS also publishes a separate statistic called "part-time for economic reasons" that counts people working part-time who want full-time work.

How accurate is the unemployment rate based on a survey of 60,000 homes?

The BLS publishes a margin of error with each monthly report. For the overall unemployment rate, the margin of error is typically around 0.2 percentage points, meaning the true rate could be 0.2 points higher or lower than reported. For smaller groups like specific age or race categories, the margin of error is larger.

Can the unemployment rate be zero?

No. Even in the strongest job markets, some people are between jobs or newly entering the labor force and searching. The lowest the U.S. unemployment rate has been in recent decades is around 3.4 percent, reached in 2023. A rate of zero would mean no one is looking for work, which does not happen in a functioning economy.

Why does the unemployment rate sometimes fall when the economy is weak?

The rate falls when the labor force shrinks faster than employment falls. This happens when discouraged workers stop looking for jobs. A falling unemployment rate during weak economic conditions usually signals that people have given up searching, not that the job market is improving. Checking the labor force participation rate and the number of jobs added that month provides the full picture.